|
Home
/ FAQ
/ News Classifieds / Events |
Audio Asylum Thread Printer |
Get a view of an entire thread on one page |
198.54.202.234
cPlay The open source high-end audio player using ASIO May 2008 cPlay delivers high quality audio playback using ASIO 2. Playback is achieved using any ASIO compatible soundcard. cPlay is a minimalist audio player using the latest high quality SRC resampler (Best Sinc SNR 145.68db or 121.33db) or SoX (VHQ or HQ). cPlay’s design offers state-of-the-art ASIO-only playback and caters for touch screen users. Installation, setup and use is easy. cPlay is built in c/c++ and operates on Windows XP SP2 Professional (32 bit).
FEATURES
- Resampling is sourced from LibSampleRate (version 0.1.5) and SoX 14.2.0 under GNU GPL license. LibSampleRate is aka SRC (Secret Rabbit Code) and supersedes the version as used in foobar2000. Best Sinc converter now offers a SNR of 145.68db or 121.33db (versus 97db). SoX VHQ offers better than 170db SNR. Resampling is bypassed when input rate matches output.
- Supports Steinberg’s excellent ASIO 2 and is backward compatible to prior ASIO versions (as required by ASIO drivers).
- Offers high quality 64 bit double precision digital volume control (in 0.5db steps). This can be bypassed.
- Playback is achieved through .cue, .wav or .flac files. cPlay loads entire .wav or .flac (decoded) into RAM before starting. Playback is done directly from cPlay’s internal buffer. Cue playback requires .cue files as created by EAC (single or multi file standard).
- Ensures efficient CPU resource utilization allowing for low specification processors or high levels of upsampling. This means CPU’s can be underclocked / undervolted.
- Supports up to 3 ASIO soundcards with each having up to 100 output channels.
- Advanced optimizations are applied (if available from ASIO driver) during playback.
- Best results achieved when using cMP (i.e. cMP²). This allows for low level Windows optimizations. Use cMP release 1.0 final or later – this allows for bypassing RAM load in cMP (set RAM Load in cMP Settings to No) otherwise wav file is RAM loaded twice. cPlay allows for both svchost and lsass to be suspended during playback thus reducing the Windows footprint. Only exception is EMU’s ASIO driver which requires both (svchost and lsass) to be operational. Set cMP’s ‘Optimize’ setting to ‘Critical’.
- Full remote control is achieved with cMP: offering volume control, track navigation, next/previous and stop/eject via (wireless) mouse.
USER MANUAL
Visit cMP² website (http://cplay.sourceforge.net) for more details and setup.
Screen Shot
GETTING STARTEDDownload cPlay’s installer from sourceforge here (1.3MB). Installation and startup is straight forward.
If you don’t have an ASIO compatible soundcard, use ASIO4ALL. Note that ASIO4ALL does not support channel mapping (use default) and can only handle up to 48k sample rate.
Your feedback will help guide cPlay’s future development. Source code (4.1MB) is available via email.
Edits: 05/05/08 07/12/08 09/03/08 05/28/09 07/10/09 09/08/09
From what I understand cMP uses AutoHotkey to setup/create its environment.
I cannot find any info on the commands they use and how the screens of cMP are manipulated.
I want to add some functions to cMP using AutoHotkey (1 is switching off the monitor with remote). For that I need to know how cMP uses AUtoHotkey, which commands it uses etcetera.
Where can I find that info?
See post in cMP thread.
Thanks for the answer. This is the AutoHotkey script for the remote, I was under the impression that cMP at startup runs a script as well to change/hide backgrounds and screens? Maybe I am wrong.
Anyway, I am trying to switch off the screen using something like this:
^1::
KeyWait, CTRL
KeyWait, 1
state := state ? : 1 ;toggle state
Gui,Color,0x000000
Gui,-Caption
WinMinimizeAll
Gui,Show, NA w%A_ScreenWidth% h%A_ScreenHeight%
If state
SendMessage, 0x112, 0xF170, 2,, Program Manager
Return
It works like a charm outside cMP, it switches off the monitor with ^1 and with ^1 starts again. I've also tried this with WinMinimizeAll, same thing.
The moment I start cMP and play a song thru Foobar the command ^1 does nothing. If I use a "set focus script, (setting focus to Foobar)it works but ignores the standard background of Windows which is aparently still hyding under the cMP menu?
I just don't get it to work and my latest idea is that if cMP uses similar AutoHotkey commands to setup its environment they might disturb eacht other?
I am fairly new to AutoHotkey so maybe I am just doing something wrong?
If you want to keep the cicsRemote functions, perhaps you could use cics' script, then add your own extra hotkeys, and name it cicsRemote.ahk, compile it with the autohotkey compiler, copy it into the cics Memory Player directory, replacing the original cicsRemote.exe. I made my own crude scripts to do what I wanted (not turning the monitor off though) and did this and it works fine.
Thank you for making this cool utility. But I hit some difficulties when I tried to use cplay on an AMD system. My secondary computer is an AMD Phenom II X4 which supports SSE3, SSE4A. But any version higher than SSE2 will not start on my AMD system, and even though SSE2 version is running, it comes with cracking noise. My primary computer uses Core 2 Q6600 and it works fine with SSSE3 version. Both computer runs on Windows 7 X64.
I searched your older posts and found that you are using Intel Compiler, which could be the reason for the problem. Is there any chance for you to give GCC compiler a try, please?
Thank you very much.
Yes, Intel compiler is used (no plans for GCC). The fact that SSE2 operates and others not suggests an AMD processor error. It may also be that AMD reports instruction sets differently which causes cPlay to exit.
Phenom's TLB bug is an issue and requires a BIOS workaround. This could explain your playback issues.
Same issue here though with an AMD X2 Brisbane that reports as SSE3 compatible. I can only run the sse2 version.
Just reporting, not complaining.
Thank you for the explanation.
My CPU is Phenom II and the chipset is 785G, both roll out after 2009 Q3, I don't think they still have a bug discovered in 2007, though.
Anyway, the AMD system is located in my workplace with a certain level of background noise, using cplay on it is a waste, but I just got addicted to listen everything using cplay XD
Refer to ReadMe.txt for important details & specifications.
Change Log (2.0b32 release):
- Low level improvements in DSP and ASIO
cPlay documentation can be found at the new cMP² website .
Please REMOVE previous versions before installing cPlay 2.0b32. Normal (SSE2), SSE3 B9, SSSE3 B9 and SSE4 B9 versions are available. Running cPlay on a CPU not supporting the required instruction set will cause cPlay to exit immediately. For best results, uncheck "Timer" setting and test different "Buffer" sizes.
I started at b31. Now b32 seems to have fixed my only real gripe. I loved b31.
However, Foobar seemengly had it beat in depth of stage and "air". B1 brought improved clarity and detail. The only problem, it compressed the imaging, mainly front to back. It also constrained the width somewhat. With my Magnepan MMGs, depth & overall imaging is terrific via either Foobar (E-mu 0404 USB) or my Denon 3910 player (the Denon is better).
This seems to have been "fixed" in B2. The only reason I am asking is that the configuration of my system has temporarily changed from the time I mostly listened to b31. I still had b31 on briefly after changing the system and did not notice this:
-- when I switched to B2, the immediate impression was that nothing good was lost
-- but now the stage got as good as with Foobar
-- PLUS the definition and placement of instruments became way more solid.
If no one has heard this type of improvement from b31 to b32 I'll have to suspect that the mushrooms I had with dinner yesterday were "loaded".
hey J,
You are right. There is a big difference and I myself heard similar things. Such differences undoubtedly are system dependent.
Did you enable AWE or are you using cplay "straight out of the box?"
Now this is what I am talking about!
Great job. Very musical yet detailed.
Its weird because the image is bigger and with 31 it was smaller and more precise. That precision though seemed to make things less warm. The bigger image on 32 seems to make it warmer.
Is that the trade off? Laser precision vs. musicality??
No tradeoff involved, just a reduction in playback jitter ☺
ASIO change was significant and together with lowest latency (Lynx does 32 samples), results will be impressive (esp mids).
hey cics,
now that my cPlay> Jack> allocator VST> lynxTWO B setup crashed again (due to a new lynx driver installation) my hope for an alternative to the direct ASIO output of cPlay, enters my mind again.
the impossibility of using the JACK within cMP2 mode is still a bummer, so i wonder what could get you to look at this ?
winamp, foobar and (I think) J.River too have the ability to output to a VST client.....
kindest regards
.
I agree. Adding such support would give cplay a whole new world of usefulness and capability.
Thanks in advance.
.
Hello Y.N. I am listening to one of those B&W audio CD's> lossless> ultralinear. I have been burning in B32 for 48 hours. The fret work is palpable the decay is excellent. Thank you again, T.
No worries!
Hi, Please, tell me, can you planning to add support for Monkey Audio files?
.
Another improvement: a bit warmer and more real-sounding and engaging. Ongoing appreciation and thanks to you, cics.
Thank you very much cics! It's always exciting to test new version... As this is now late night in Hong Kong but I just can't wait, so I use AOBCT PC -> AES16e -> Mytek -> Beyer DT990 headphone to listen to wave files. Again SSSE4, upsample 192 SOX Linear 95.
The musicality is back, just after a few songs I can't stop... This is a good version! I will listen further and provide you an update tomorrow.
On the other hand, just an opinion that for every time you release a new version, if you can provide some further details on exactly what you have changed in coding, then we can provide further opinion on what "parameters" we like so for your reference. But anyway as you know I am always a supporter of cplay. Cheers!
Another thing that would like to bring up is that due to the "coarse" of high frequency in B31, I was forced to look for solution and came across the method below... DIY shielding of the sound card using aluminum foil... sounds crazy but it works! It is cheap and effective on high frequency, worth a try:
http://www.overclock.net/sound-cards-computer-audio/571718-how-make-emi-shielding-your-sound.html
To further reduce high frequency distortion, you may also try applying binder clips on edges of your sound card. I still have a fan in my computer and another one in the power supply, but I think anyone with a cmp would still notice an improvement even if you don't have any fans. The force of the compressed spring will dampen mechanical vibrations that although not easily detected, will transmit into and interfere with your electrical signal. Large ones and tiny ones all fit in different places and I have them on everything.. from equipment shelves and interconnects to RAM and sata cables. It's not a pretty tweak, but it is cheap...the more you use the more they show their effectiveness.
Thank you very much texastea006! I know we are a brunch of weird audiophiles... When I was holding my al-foil-wrapped Lynx AES16e walking around in the house, my sister saw it and just couldn't stop laughing... you can imagine how silly it looks.
Anyway, I understand what you mean and I may use cable tie/binder clip to tighten up the cards and cables inside. THanks again for your advise!
http://en.wikipedia.org/wiki/Binder_clip
Two key areas were affected:
- TLBs or Translation Lookaside Buffers when using SRC
- ASIO - this area has had biggest improvement
Overall, B32 is a refinement on the extensive work done in B31. I wasn't planning to release B32 so soon but SQ improvement is significant and felt it warranted a release.
From now I have to be more careful before making comment on any new version, as cplay's resolution is so high that any flaws or noises in my PC will be revealed by cplay mercilessly. Recall that I thought cplay B31 was coarse at high frequency. However, after inspecting my own PC, wrapping up cables and sound card using aluminum foil, the high frequency resumes normal, strings becomes crispy clear without grains.
Thanks again cics :)
THank you cics. Would you suggest using 64bit Win7 instead of XP? As someone claims 64bit Win 7 uses floating point Sound system, but my understanding is that ASIO bypasses everything of Windows Audio. So does it matter?
I would avoid 64 bit OS as it adds more hardware complexity (64 address lines vs 32+4). Main reason for switching to 64 bit is using more than 4GB RAM. This is useful when using large capacity database servers.
.
SQ improvement is indeed significant. The top end haze is again reduced, or in the case of b32, removed, resulting in:
-- better high frequencies
-- more solid mid frequencies
-- more precise imaging, and
-- wider and more 'substantial' sound stage
Thanks again cics. The cMP/cPlay combination means a lot to my listening life.
I didn’t gave my feedback on version B31 because I couldn’t make my mind up on B31. It had more accuray, details and more soundstage than B30, but somehow the scanpeak tweeters in my K&H O300 didn’t like the high’s from B31. B31 as a whole sounded “tensed & stressed”. Version B31 in my setup would be best described as: B30 ‘on steroids’ Or ‘on speed’. Good and fast but somehow not very pleasant to listen to.
But since there were also many very good reviews of version B31, saying the high’s where very smooth and a pleasure to listen too, version B31 left me puzzled about my gear, which I have carefully chosen for their reputation being very neutral and uncoloured gear. I choose to go back too version B30 and occasionally even used the players included in Wavelab lite and in Audition 3 again.
But my worry’s are over now. The unpleasantries of B31 are gone! But all the quality’s of the B31 are still there (more detail, more open, more stage) and are now a real pleasure to listen too. High’s are even more detailed and open than in B31. But probably the most progress has been made in the mid section. B32 has very very nice sounding and very accurate mids. Hmmmm ….. I really, really love the mids in version B32 !!
The D-tour via version B31 was probably a necessity to reach version B32. I’m very pleased with B32. B32 sounds very, very good in my setup. I think it will be tough job to better B32 again.
Thank you cics for B32 !
fully AOB optimized cMP2 PC -> Lynx AES16 -> XLR AES/EBU -> Lavry Black DA10 -> XLR Mogami Gold -> Klein & Hummel O300
Good afternoon,
and many compliments for the development of Cplay.
I have a small issue, related to my HTPC soundcard and the proprietory ASIO drivers that come with it.
In Cplay, the proprietory ASIO 2.0 drivers are visible, but I cannot use them to play music, as an error message comes out telling me that they are 24bit ASIO drivers, while Cplay only accepts 32bit ASIO.
As a solution, I'm using ASIO4ALL, but I'd like to know if this situation is widespread or specific to my setup, and if there's any chance of an alternative solution allowing me to use the original ASIO drivers of my (quite nice) soundcard.
Thank you for the patience...
Ciao!
tranfa
I had exactly the same problem. I've been using winamp and foobar with Asio.
Yes, that's exactely what I'm doing as well...but I tend to like CPlay better for FLAC reproduction....
Should it have a Convolution utility inbuilt, in order to apply DRC filters, it would be the perfect application for me.
Hi carcass93,
thank you for your frank reply.
Tranfa
It is not unique to your setup.Aside from using ASIO4ALL, which is a compromise sound quality-wise, you have 2 options:
1. Give up on ASUS, and replace it with another soundcard;
2. Give up on cPlay, and replace it with another media player.As I understand, cics (developer of cPlay) is not interested in making 24-bit version. By the same token, I'm fairly sure ASUS is not going to create 32-bit version of their driver.
Edits: 11/04/09
If I build the cMP to spec. What complications am I adding to use a Blu Ray capable drive and video card? Does that completely defeat the purpose?
Whatever sounds good...do that.
Probably. Minimizing video is a decent part of this project and you would be going the other way.
Though there ARE some tips for home theater in the build and you can take what parts you want. There are some inmates who have used the principles and made a HT rig and are pretty happy once they got it figured out.
...No, the real killer feature of Windows 7 is scalability. Simply put, Windows 7 does a better job of taking advantage of the available hardware resources than its predecessors. This scalability edge manifests itself in the form of better performance under complex, multiprocess, multithreaded workloads.
Given the same number of CPU cores, Windows 7 runs circles around both Windows Vista and Windows XP. In fact, the results aren't even close: In one multiprocess workflow test, Windows 7 outpaced Windows XP by 250 percent -- this on an eight-core (dual quad-core Xeon) HP Z800 workstation.
This is Windows 7's killer feature. It means that, as customers invest in new PC hardware, they'll be better positioned to reap the improvements in CPU, memory, and chip set performance by deploying Windows 7. It also means that sticking with Windows XP -- ostensibly because it is less bloated and performs better -- is a fool's errand...
Ok to any cplay/cmp uisers out there on Windows 7: do you notice any cpu efficiency running music files? More specifically does cpu % usage drop with Windows 7 versus XP?
Hi all,Summer in The Netherlands is at it’s end, so it’s time too pick up my indoor hobby again. Which means its time for: Power Supply improvements: part 2.
As already stated by other inmates (Gstew, Reylands, Bertel, Theob, carcass93, Rayban, Sondale, fmak and many others) improving the power supply is a very worthwhile and very rewarding exercise.
Optimizations of the Operation System, Kernel and the BIOS all together do have a nice effect on sound quality, but I was really amazed what BIG IMPACT (!) a linear PSU on the P4-pin had on sound quality in my setup.Note: my setup is a fully optimized PC as per Cics hard- and software recommendations. So my findings are based on a setup like this: see: http://cplay.sourceforge.net. With only 2 exceptions: I do not under-volt the FSB nor RAM . I over-volt both with + 0,2 V.
Improving the power supply to the P4 pin with a linear PSU turned out to be surprisingly easy (and also relatively cheap), Click on picture to see photo's in Picasa webalbum:
![]()
cMP2 project: Optimizing power with linear PSU to P4 power connector’
So I now would like to try optimizing the power to the P24 connector.Just before summer (in june) more knowledgeable inmates who already tried to improve the power on the P24 (Bertel, Gstew and others) gave me a lot of suggestions and possible directions.
-> Bertel pointed out the importance of choosing good quality low-ESR smoothing caps but also and the use of low resistance wires (!).
“ to keep power supply as clean and quick as possible to keep the rise time between 0 and 1 ("no power" and "power", changing ever so quickly) at a minimum! IMHO this can best be done with caps sized the "right" way, plus cables from the caps to the board which are short, have a lot of internal conductors (that brings speed, e.g. I use AWG 6 which is 133*AWG27) and a very low internal resistance ( <.5Ohm/1000ft, just to be safe - that's in the region of an average PCB's traces) “
-> Gstew suggested use a tweaked Pico PSU (extra smoothing caps) on the P24 pin.So that’s what I did (wires and picoPSU) last week, with unexpected and very surprising results !!
* using LOW RESISTANCE internal power wires to P4 and P24.
Although Bertel already had pointed out to me the importance of using low resistance wires, I still wired everything together with standard computer wires (18AWG wire). Why? because its easy and I’m lazy. Standard computer 12 v power wires are constructed of 18 AWG and can simply be bought everywhere and they have already all the right plugs and connectors attached and in place. Regular wire to distribute power inside pc’s is: 18AWG = 6,5 ohm/1000ft. Bertel suggested to use 6AWG which is: 0,4 ohm/1000ft. http://www.interfacebus.com/Copper_Wire_AWG_SIze.htmlUntill last Sunday. A suddenly cancelled social appointment left me with nothing much to do on a rainy Sunday afternoon. With electronic stores being closed on Sunday in The Netherlands, I had no other wire at hand than the type of wire that is used to wire the electricity in housing (with solid 2 mm copper conductor).
I braided 2 solid copper wires together until I had 1 meter of braided solid copper cable. With still 1 cm standard 18AWG computer wire left at the P4 plug I soldered the braided solid core copper wire to the P4 plug (a P4 plug has 2 black 18AWG wires + 2 yellow 18AWG wires)
I saw no other way than too solder the 2 mm solid copper wire to the P4 plug than trough leaving 1 cm of the original 18AWG standard wire attached to the P4 plug. The best solution would be to solder the low resistance wire directly to the metal-connectors inside the plastic plug. But I don't manage / don't know how, to get the metal-connectors out of the plastic plug.Click on picture too see photo's of the solderingproces of low resistance multi-stranded wire to the P4 plug in the picasa webalbum.
![]()
cMP2 project: low resistance wire between linear PSU’s and P4 / P24 power connector
I also braided 1 meter of 2 times 2 mm solid copper wire to also connect the picoPSU with this braided wire to the linear PSU that is feeding the picoPSU.Working with this braided 1 meter 2 x 2 mm solid coper wire is a hassle. The braided wire is not flexible and one has to bend the braided solid copper wire in place all the way from the linear PSU, through the pc case, to the P4 and P24 connectors. Of course it’s much easier to use multi-stranded wire (6AWG or similar).
I didn’t expect much of this construction because there still was a bottle neck of 1 cm 18 AWG standard computer wire attached to the P4 plug. So I expected the whole wrestle with the stiff braided cable would probably not be worth the trouble.
But WHOW… I couldn’t be more wrong! It made big and massive impact!!
Bass, mids, treble everything changed dramatically. (I’m not writing this for effect purposes) Especially drum kicks, plugged strings, or any type of ‘explosive sounds’, where very much more heard and present. A drum kick, now sounds like a real drum kick !!!
The changes where that big, that I first thought, that there was something wrong. That I only added some kind of reverb like distortion. But after playing more and more music and tracks I started to realize that this was no distortion at all. This is probably how the music was recorded and mastered.
* Optimizing power to the 12Volt rail of the P24 with linear PSU + picoPSU.
- Chosing the right picoPSU model.
It turned out that not al picoPSU are made by the same concept. Some picoPSU models leave the 12 V line untouched and pas it straight on too the MoBo, where other picoPSU models don’t leave the 12 V line untouched and regulate the 12 V line.So shit happens and I first bought the wrong picoPSU model. A PW-200-M w/P4-ATX. This model regulates the 12 V line which results in almost no sound quality improvement. So it’s not worth the trouble to change a ‘smoothing capped’ ANTEC ATX PSU for an PW-200-M w/P4-ATX.
By now I have ordered an other picoPSU model and installed it. The model picoPSU-150-XT leaves the 12 v line untouched and passes the 12 V line straight on too the MoBo. Now there is a very nice improvement in sound quality !
However a very nice improvement, it’s not as much as with a linear PSU on the P4 pin.
In comparison: a picoPSU-150-XT on the P24 gives about 1/3 of the improvement of a linear PSU on the P4 pin. (So improving the 5 Volt and the 3,3 volt power also needs to be done. Especialy the 5 v line. But thats for PSU improvements: part 3, because I haven’t made up my mind yet on how to do it the easy way). But although only 1/3, it’s still a worthwhile and recommended improvement !So chosing the right picoPSU model (which leaves the 12v line untouched) is important ! Installing a picoPSU model that regulates the 12V line, it’s not worth the trouble IMHO.
![]()
cMP2 project: Optimizing power with linear PSU + picoPSU to 12V rail on P24 power connector’
* First question on internal power wires.
Are there any other inmates who swapped out the standard PC power wire type (18 AWG) and replaced these internal pc-power-wires for low resistance wires like Bertel suggests?
(AWG 6 wire or lower resistance)
Or, are there any inmates willing to experiment with low resistance type of wire between their linear PSU and the P4 pin and/or their picoPSU ? And report back there findings ?* Second question on internal power wires.
If this would be analog audio, I could easily understand what is happening and understand the effect. But what is happening here ? Is this really the effect Bertel pointed at? “ keep the rise time between 0 and 1 ("no power" and "power", changing ever so quickly) at a minimum! ”Any comments, suggestions, feedback is very much appreciated. Or even better: please try this at home and report findings back here at the forum !
Power supply improvements part 3, will on the P24 5 volt and 3,5 volt lines but I haven’t made up my mind yet on how do to this the easy way. Probably there is no easy way.
Note:
I don’t encourage anyone too use solid core 2 mm copper wires. It’s a hassle. I just had nothing else lying around at that time. But I do encourage to try multi-stranded 6AWG wire (or any other low resistance wire) between the linear PSU and the P4 / P24 connecter and report back the findings to this forum.
fully AOB optimized cMP2 PC -> Lynx AES16 -> XLR AES/EBU -> Lavry Black DA10 -> XLR Mogami Gold -> Klein & Hummel O300
Edits: 11/02/09 11/02/09 11/02/09 11/03/09 11/03/09 11/08/09 11/09/09 11/11/09 11/11/09 11/11/09 11/11/09 11/11/09 11/11/09 11/15/09
I think I'm almost ready to take this same journey. All this information is very exciting to me. Do you have any pictures you can share?
Also, I'm interested to know, do you have just one linear PSU powering the P4 and the pico or are you using two separate linear PSUs? Also, which brand/model linear PSU are you using? Or did you build it yourself?
Thanks for sharing!
Hi Edward,
Since you asked I made some picture to share via Google Picasa webalbums this week, for your and other inmates convenience .
See added links to Picasa web photo albums in original post.
fully AOB optimized cMP2 PC -> Lynx AES16 -> XLR AES/EBU -> Lavry Black DA10 -> XLR Mogami Gold -> Klein & Hummel O300
Thanks for the pix! So then I take it you are going from the PSU straight to the P4 & P24? And not using any "smoothing" capacitors in between? Have you found that with the linear PSU, the caps are not needed?
Also, I'm intrigued about using lower resistance (higher gauge) wiring. I'm confused though. Correct me if I misunderstood, but did you not say you braided two 12 gauge solid core wires together? In the photo it appears to be a multi-stranded wire. I'm not an expert on these matters nor an electrician, but I am a fan of the concept of solid core cables. So I'm wondering if it would be sufficient to use 10 gauge solid core copper (and try to go directly to the P4 connector without using the 18 gauge stranded wires).
Thanks again for your contributions and information.
Hi Edward,Yes from the linear PSU the wires go straight to the P4 / straight to the picoPSU on the P24.
I’m also not an expert nor trained / skilled in electronics, so we both should be careful not to become: two blind people, who are leading each others way. :-)
But my understanding is that smoothing caps are already employed inside a linear PSU after the rectifier bridge and or Voltage regulator. I don’t know if extra caps outside the linear PSU betters things extra.I’m not trained /skilled in audio electronics and/or digital electronics. So I can’t say that I’m a fan of solid core cables or stranded wire. I just don’t know which one is better for a given situation/application. I just wanted to try Bertels suggestion of using wires with a much lower resistance than the standard used 18AWG wire. At that time (Sunday afternoon) I only had solid (1,5 mm core) copper wire lying around as used in the electricity system in walls in housing. I braided 2 strands of this solid core copper wire together, to get 1 metre cable.
Since the 1 meter cable of 2 braided solid copper wires is really stiff and a hassle to work with, this week I bought some 2,5 mm (outside diameter including isolation) multi-stranded copper wire and made again: a P4 extension cable of 4 times 2,5mm multi stranded copper wire. Soldering of these 4 multi-stranded 2,5 mm copper wires to the four 1cm 18AWG wires at the P4 plug, is what you see on the pixs.
Both wires types (solid copper and multi-stranded) sound the same as used for P4 extension cable / power cable to the picoPSU at the P24. But 2,5 mm multi-stranded copper wire is much more easy to work with.I haven’t received any feedback from other inmates yet on sound quality when replacing 18AWG wire for a wire type with a much lower resistance as Bertel suggests (6AWG or similar).
I hope other inmates do want to try this and report their findings to the forum. Because – also very much to my surprise ! - it had a very, very big impact on sound quality in my setup.As for your question: “So I'm wondering if it would be sufficient to use 10 gauge solid core copper (and try to go directly to the P4 connector without using the 18 gauge stranded wires)”, I can’t answer your question as I don’t know that.
fully AOB optimized cMP2 PC -> Lynx AES16 -> XLR AES/EBU -> Lavry Black DA10 -> XLR Mogami Gold -> Klein & Hummel O300
Edits: 11/09/09
Hi Edward,I bought a simple Voltcraft 12V/2A inear PSU (40,- euro’s) at www.conrad.nl (part number 510172). I have not enough knowledge nor skills to build my own.
Voltcraft is their home-brand and is produced in Chechoslowakia. But one can buy this type of simple linear PSU’s everywhere. For what is worth: the technical specification shows a max ripple of 2,5 mVolts. But I don’t know if there are any other distortions present on the 12Volt output.As Reylands already suggested it really very easy and simple to connect a linear PSU to the P4 socket if one uses a P4 extension cable. http://www.atxpowersupplies.com/p4-extension-cable-adapter.php . Just cut off one connector and attach that side of the extension cable too the linear PSU and put the other end into the P4 connector on the MoBo. That’s all. This will take you no longer than max 15 minutes. And there you go !
If one P4 extension cable is to short (they are about 14 inch / 36 cm which usually is to short), just use 2 or 3 and connect them together to get the desired length.
But using these handy ready made P4 extension-cables comes at a sonic price as I discovered this weekend. They are constructed with AWG 18 wire. And as inmate Bertel already suggested to me, using low resistance wire (for instance AWG 6 or so) might be far better. And Bertel was o so right !!
If one uses low resistance wire as Bertel suggests, than there’s again a big sound quality improvement.I think many inmates did the same thing like I did, and wired everything together with standard AWG18 wire. So they are not getting the maximum sound quality improvement possible. That’s why I decided to post my findings and encourage other inmates to try the same and report back there findings on sound quality.
Right now I have 2 linear PSU’s outside my pc-case to power the cMP2 PC and 1 standard ‘dirty’ ANTEC Earthwatts ATX PSU mounted inside the PC case. The ANTEC now takes care of the job that originally was indented for the Granite PSU’s.
I don’t like the Granite solution to feed the USB-port, HHD and DVD-drive. When using the Granites one has to bring 230 AC wires inside the PC-case to power these Granites. I don’t like that in terms of noise, hum, RF-polution, etc.
On the other hand I don’t worry very much about ‘horror stories’ on how ‘noisy’ it supposably is inside a PC case. A PC entirely runs on 12/5/3,3 volt DC, so I don’t worry at all about ‘RF-noise’ inside my PC-case. I just don’t like 230 AC wires and unshielded 230 AC -> 12V DC converters inside my case.Having 2 linear PSU’s was not a deliberated choice, it just went this way because I first optimized the power on the P4 pin. I found this to be so rewarding that I decided to also give the P24 a try. So I had to buy second linear PSU 12V/5Amp to feed a picoPSU on the P24 (Using a picoPSU model that leaves the 12V line untouched !)
But when using 1 linear PSU it will have to be a big one because it will have to provide 7 or 8 amp total. May be there is also some benefit from using 2 separate linear PSU’s. They cannot influence each other, ect.
If you buy a linear PSU be sure to buy one with low ripple specs and passive cooling (without a noisy cooling fan).
Furthermore I do suggest that you first carefully read the ‘Advanced’ section chapter 12.b on http://cplay.sourceforge.net/ and their subsequent links to other inmates here on the forum. Some of those inmates are far more knowledgeable than I am. I just read all their recommendations and suggestions and than try something in practice and report back my findings to the forum.
If you don’t mind an extra PSU box beside you PC-case, than a linear PSU on the P4 is a must do. When using P4 extension-cables as suggested by Reylands, you are done within 15 minutes.
It even gets much (!) better, when you don’t use the ready made P4 extension cables (with AWG 18 wire), but when you use low resistance wire (something like AWG 6 wire as Bertel suggests) .Succes and happy listening.
fully AOB optimized cMP2 PC -> Lynx AES16 -> XLR AES/EBU -> Lavry Black DA10 -> XLR Mogami Gold -> Klein & Hummel O300
Edits: 11/03/09
Vellerman's PS1503SB (for ~$90) looks very impressive with RMS ripple specifications of ≤1mV and ≤3mA!
![]()
According to manual: for stable voltage supply you need to set current output to Max (3A), turn on and adjust voltage to 12V. Anybody tried this?
PS - I've added a link to your excellent post for Hybrid P24 power supply (Advanced section).
Hi Cics,* using one linear PSU (instead of two)
When I first red your suggestion: “how about 1 linear PSU of 12/3A ? ”, I thought: well a 3Amp linear PSU is not enough. Other inmates on this forum reported that the current draw on the 5 volt P24 line alone was between 4 and 6Amps. If one would ad this up to the also reported:
- 0,5 Amps for the 3,3 Volt
- 0,5 Amps for the 12 Volt
And for the P4:
- 0,4 / 0,5 amp during boot (0,3 – 0,4 while playing music)
That would make a total of:
- P24: 6A + 0,5A + 0,5A
- P4: 0,5A
Total = 7,5 A.
Say 8 Amps to be on the save side.But it’s seems that 4-6 Amps trough the 5 Volt rail on the P24 pin (as other inmates have reported) isn’t correct.
Since I now use 2 linear PSU’s, of which one is feeding the picoPSU at the P24 pin, it’s has become very easy to measure the current flow through the 12 Volt line that is powering the picoPSU. So that’s what I did this afternoon, after reading your suggestions too use only one 3 Amp linear PSU. My measurements gave the following readings:
The current flow on the 12 Volt DC line towards the picoPSU is:
- while booting up: between 1,5 and 1,7 with peaks to 2,0 amps
- while starting XP: between: 1,75 and 1,85 with peaks to 2,0 amps
- when Xp is at rest: 1,75 Amps
- when starting a program: 1,8 – 1.85 Amps
- While playing music: 1,76 – 1,79 AmpsSo when using only one linear PSU, the unit would have to deliver:
- P24: max 2 amp
- P4: 0,5 amp
makes a total of: 2,5 Amp.
So your suggestion of a linear PSU at 12V/3 Amp would indeed be enough.
I indeed would like only one linear PSU for space and energy consumption reasons. But also because only than both the P24 and the P4 will get exactly the same voltage. And there will also be no a voltage difference between the 2 black minus poles of each PSU and the ground.
On the 12V/2Amp linear PSU the voltage between the black minus pole and the grnd always was: 0 Volt.
But on the 12V/5Amp linear PSU the voltage between the black minus pole and the grnd always was: -0,003 / -,021 volt. No matter what I did, I couldn’t bring it also back to also: 0 volt.
If this where analog audio, one wouldn’t want this difference between two PSU’s. I don’t know if this also matters when powering digital data processing, but this didn’t give me ‘ease of mind’.I now connect both lines feeding the P4 and feeding the P24 to one linear PSU as you suggested. The 12V/5Amp one. It works fine. The safety-circuit hasn’t shut it down. Not when switching on the PC (run-in currents), booting up or when playing music. On first ear, I doesn't make any differences in SQ, But I will do some carefull listening on that in near future.
* low resistance wire and linear PSU’s with better specs.
Unstill now I haven’t had any feedback from inmates who also changed from 18AWG wire to a wire with much less resistance and how this improves sound quality in their setup.
That’s a pitty, because the sound quality in my setup improved really, very, very much !!
And also the MoBo liked it very much :-) I now can cold boot without any failure at only 130 mhz FSB speed. And warm boot at even 125 mhz FSB speed !If low resistance wires have such a (real big !) impact on sound quality (in my setup) than I probably should have paid much more attention to linear PSU specs, such as very low Voltage ripple and very low Amp ripple, when I ordered one. I more or less bought the first low cost linear PSU, with reasonable low Voltage ripple, I came across.
I think the Velleman linear PSU you suggest, looks like a good buy. It’s only € 57.50 euro and has much better specs than the 12v/2A one that I bought for € 40,-. And it’s looks better too! Thankx for your suggestion. Although I won’t buy another linear PSU in the coming months. All funds will be directed to my upcoming 1 month holiday in New Zealand and Australia in February 2010.
I hope other inmates will soon report their findings on linear PSU’s with better specs and swapping their internal 18AWG power wires for low resistance ones.
fully AOB optimized cMP2 PC -> Lynx AES16 -> XLR AES/EBU -> Lavry Black DA10 -> XLR Mogami Gold -> Klein & Hummel O300
Edits: 11/08/09 11/09/09
Thanks again for the info. And it brings up another question. On your measurements of 2 amp (P24) and 0,5 amp (P4), I assume this is based on underclocking and undervolting? Does it make any difference if the CPU is run at full speed with default volt?
For example, if I get a PSU at 3A, but then experiment with lowering FSB and voltage further and then go too low to the point it will not boot - then I will have to reset the CMOS/BIOS which will boot at defaults. Will 3 amps still be enough at default speeds?
Hi Edward,Good point. your assumption is correct. 3 Amps is not enough with the BIOS set at: ‘Load optimized Defaults’ :-(
I connected the multi-meter again and loaded BIOS settings: ‘optimized Defaults’. Then I repeated the booting process several times. The highest peak I saw on the multi-meter while the BIOS was starting the PC was a short peak of 3,44 Amps. One of these booting sessions I recorded on camera, which can be seen here:
During this boot up session you can see a short peak of 3,33 Amps in the 18 / 19 second (is fase where the BIOS is starting the PC). Another short peak of 3,44 Amps can be seen later on when (a normal full version of) XP starts up.So a linear PSU of 3 Amps is only okay if it handles / tolerates short peaks of 3,5 Amps.
Too be on the safe side a Linear PSU which has to feed both the P4 and P24 should be cappable of supplying: 4 Amps.I also was curious so what the maximum Amp need of the system was. So I played a HD-movie (1920 x 1080) in a normal XP version with Windows Media Player Classic Home Theatre Version (1 Gb RAM installed). The PC than needs 3,75 / 3,85 amps for playing a high-definition movie
I also played the same HD movie in Windows 7 with Windows Media Player 11. In Win7 the PC needs 3,90 / 4.00 for playing a High Definition movie.So a linear PSU capable of providing 4 Amps is enough for the BIOS to start the system with setting 'Optimized Defaults'. It's also (just) enough for the system to play a high definition movie.
* PC system:
-> Gigabyte GA-G31M-ES2L, 1 RAM HyperX, E7300 Intel Processor.
* PSU’s:
-> the linear PSU 12V/5A powers: P4 = processor. P24 = MoBo, Lynx AES16 digital audio interface and integrated VGA on MoBO (1920x1080).
-> Antec 430 ATX PSU powers: USB, HDD and DVD.
fully AOB optimized cMP2 PC -> Lynx AES16 -> XLR AES/EBU -> Lavry Black DA10 -> XLR Mogami Gold -> Klein & Hummel O300
Edits: 11/12/09
nt
Will 3 amps still be enough at default speeds?
I posted some measurements with the Gigabyte G21 board back in April. They were buried deep in one of the cMP^2 threads but the key bit was:When playing flac-format music data without upsampling with the CPU clock at 140MHz, VID at 0.75000, PSU at precisely 12v, EIST off, 512 MB Kingston "ValueRAM" and a "Full Monty" cMP2 setup, the motherboard and CPU typically draw about 1630 ma.If you reset to defaults, the current draw will get quite close to 3 amps. Running flat out 24/7 with a 3-amp unit is cutting things a bit fine but, given you're going to be resetting fairly quickly, my guess is it'll be just fine.
HTH
Dave
I have problem displaying multi-byte Chinese characters in cMP and cPlay; though the characters are displaying quite nicely in MS notepad when I edited the cue sheet. I believe a friend of mine has no such problem in his setup so I think may be I am lacking the appropriate Chinese font file. Can someone help me with this problem and what font is cMP and cPlay using?
Unicode / DBCS support has to be added...
Thanks for the info; and I finally got it after a few trail and errors.
This is what I'd done in addition to having Asian Language added to Widows XP:
1. Save the cue sheets with Chinese characters as ANSI files. There is a warning when one tries to save files with Chinese characters to an ANSI file, but one has to just ignore it.
2. Go to Control Panel> Regional and Language Options> Advanced, and set the "Language for Non-Unicode Program" to the language one needs--in my case "Chinese (PRC)"
Then everything just works perfect.
I already had added Asian language support in XP. Chinese character input and display works in the OS (file and directory names ext.) as well as applications like Notepad. In cMP and cPlay, the Chinese character in the cue sheets, after edited in through Notepad, are display as garbage. May be I don't have the required Chinese font file? What is the font used in cMP and cPlay?
I am newbie and haven't tried to test this but someone on head-fi has stumbled upon an issue with cPlay's resampler on his Windows 7 64-bit computer. Is the resampler being used improperly or is this just a bug from using windows 7?
To quote: There appears to be no change in volume below -80 db & the output is very dirty below -60db. there appears to be no dither used to clean it up as it sounds just like the early 14 to 16 bit ladder converters with early digital recordings. Massive distortion in the lower 2-3 bits. This seems to me to mean that there is some bit truncation occuring in the software of the Cplay program. This is true no matter the settings.
http://www.head-fi.org/forums/f46/cmp-cplay-media-player-444784/index6.html#post6094470
Thank you.
cPlay outputs 32 bit audio to the soundcard driver. On startup, if the driver doesn't support 32 bit, cPlay reports an error and will not allow for any playback.
Driver converts 32bit format down to 24 (or 16) either at sw or hw level and may apply dithering if needed.
Based on the problem experienced, that driver needs to be either reinstalled or may have a bug. There could also be an issue with 64bit Windows - here, cPlay must be run in 32bit mode.
cics, I have not been able to get cPlay to play .wav files that I have ripped from DVD-A's using DVD-A explorer although i have been successful in playing high rez .wav files ripped from DVD-V's.
Is there something different about the DVD-A .wav files that is not compatible with cPlay?
Thanks,
Jack
Look at "Diagnostics" messages. There should be a reason given for why the wav file was rejected.
cics, the error message I am geting from these wav files is:
Wav file is NOT supported.
Only 16, 24 or 32 bit (PCM or IEEE754 Float) uncompressed stereo (at any sample rate) is supported.
You can use SoX to check your file. Download "MS-Windows executable", unzip folder and copy the 2 dll files (cygwin1 & cygomp-1) to c:\windows. Command from a command prompt is:
c:\mysoxfldr\sox.exe --i "c:\my long name folder\My Wav.wav"
Output would be something like this:
C:\> sox --i "c:\music\01 Track01.wav"
Input File : 'c:\music\01 Track01.wav'
Channels : 2
Sample Rate : 44100
Precision : 16-bit
Duration : 00:02:39.73 = 7044240 samples = 11980 CDDA sectors
File Size : 28.2M
Bit Rate : 1.41M
Sample Encoding: 16-bit Signed Integer PCM
C:\>
It may be different wav format and SoX may help in converting it back to the old wav format.
Try following:sox "MyTrack01.wav" "NewTrack01.wavpcm"
rename "NewTrack01.wavpcm" "NewTrack01.wav"
Test cPlay with "NewTrack01.wav".
In cMP Settings, set RAM Load to "No" and both Suspend svchost & lssas to "No". What soundcard are you using?
Thanks cics. I"ll try your suggestion but I am pretty certian that the files I ripped from the DVD-A's were the stereo track. I will follow up if my SOX output confirms this.
Another problem I am experiencing id that when I try to use cPlay while in CMP-2 it appears that CMP/cPlay recognizes and loads all of my .wav and .cue files but I cannot gey them to play. I have no problem playing the .cue files when nCMP is not loaded when I select hem directly from my music file directly so I know it's not the files themselves but perhaps the path to them?
I seem to have a similar problem when I create a playlist with cPlaylisteditor. I cannot get the saved file tp play in cPlay.
Thanks again for any help you may be able to lend.
I have a question that might seem stupid to those with more than my beginner’s knowledge. I apologize.
My question concerns upsampling in cPlay. Do I need this, given that my DAC upsamples? Is there any potential downside to upsampling in cPlay and then again in the DAC?
Thanks to all for the information. I know that my ears ought to be able to tell me what works and what doesn't, but sometimes it's good to get some outside verification to save on some of those terrible moments of self-doubt.
I can’t disable the upsampling in my DAC, so I’ll just continue to experiment with the setting in cPlay.
This is not a stupid question at all. And the answer is not so cut and dry. It depends on what sample rate your DAC upsamples to and how good it is. Some people prefer the upsampling in cPlay and others think their DAC upsampling sounds better (or no upsampling at all). The *opinions* on this subject will run the gamut.
The simplest answer I can give you is to try it both ways and see which sounds best to you.
Hi DeDe
Just to continue Edward's "it depends" answer... I run the output of cMP/cPlay from the Juli@ card via S/PDIF into a Benchmark DAC1. The DAC1 resamples ALL incoming datastreams to 110 KHz (this being the SR that the engineers feel is the sweet-spot of performance for the DAC chip they use). There is no user option to alter this behaviour - it's the internal architecture.
Curiously (or maybe not...) The change in upsampling with cPlay is readily apparent in my system, from no-upsampling, to 2x SR to 4x SR. Although the DAC1 converts it "down" to 110 KHz, I find setting cPlay to output at 192 KHz gives me the most pleasing (and seemingly accurate/ revealing sound).
Other's preference will certainly vary... so the "correct" way would be as Edward says - whatever you like best. Figure that out, and enjoy your music guilt-free !
That's not a Toy... IT'S A TOOL !!
I find this hard to understand. I too output to a dac1 digital aes3. the dac1 input is 24/96, how do you input at 192. When I output above 96 the dac does not input the signal, Lynx L22/Cplay advanced no power mods. T.
No worries!
@tsearay: I connect to DAC1 using S/PDIF. Though mine is original issue DAC1, it has been factory upgraded to full 192 KHz capability.
Does the Lynx22 send 192 KHz in single-wire AES mode, or dual-wire mode ? It must be single-wire AES for DAC1 to work at 192 KHz on XLR input.
Hope this helps.
Cheers,
Grant
That's not a Toy... IT'S A TOOL !!
AH. You dont output higher than 96k on the Lynx.The Juli@ though can do 192.
And have you tried the analog outs of your lynx? That WILL handle 192k.
Edits: 10/20/09
What leads you to say this? My lynx aes shows an output setting of single wire 192 as well as dual wire 192. I have no way to test it presently, but it appears it is in its ken.
Yeah, for the AES, but we are talking about the L22.
Look at the spec sheet for the AES 16 it DOES have support for 192, but not the L22. Look under Digital I/O in the lynxs ( :) )below
http://www.lynxstudio.com/pop/product_file.asp?i=25
and for the l22
http://www.lynxstudio.com/pop/product_file.asp?i=24
Your right, I missed which Lynx you were talking about and hadn't realized they differed in this respect.
I wasn't sure myself!
Looks like the 16 came later and supports 192. Thats a great thing for those who have the cards.
Ah! Thank you! you are correct the digital out on the Lynx is limited to 96. I have tried the balanced out but prefer the digital to DAC. I have been struggling with "to upsample or not to upsample" to the dac, myself. I get a lower latency at 96 than 41 but really hard to tell if there is a benefit in SQ. AES/XLR throughput to balanced amplification. T.
No worries!
I thought it was worthy of a semi-celebration.
Heh-heh... Thank You Rick for the kind thought. Let us find a suitable container and pour a half-glass...
Cheers,
Grant
That's not a Toy... IT'S A TOOL !!
New cPLAY user that has only used versions 2.0b30 and 2.0b31 so far. On both these version cPlay does not fit in the maximized window (it is just cut a little short) nor does it fill the full width and height of the monitor.
I have used a 17" and a 19" LCD at differant resoutions with no prevail. It really isn't bothersome but was wondering if others have had the problem and if there is a possible fix.
This is on a allout cMP² minus ps mods.
Brad
to fix that:
In cicsMemoryPlayer.pth, set CUE_PLAYER to start with %M (Maximised) instead of %N (Normal). It helps only when video resolution is 1024x768.
i must be using that resolution because it worked, though I dont remember choosing that resolution.
Basically it made the bottom of the player longer so it filled the screen.
I must have something turned off that won't allow me to open the .pth extension???
Brad
d
Duh me thanks, that worked but only with resolution at 1024 x 768 as you said.
Brad
.
Had no luck with dpi setting
Brad
What resolution do you want to use? Even if it is more than 1024x768, you should be able to get close to filling the screen with the right dpi, try with %N if %M isn't doing it, and maybe you can make up the difference with adjustments to display size on the monitor itself. Or make it a bit smaller and have a nice coloured border (desktop colour).
If you are working with Process Explorer or Explorer, or any embedded apps, and you are using mouse exclusively, it is useful to not have fullscreen, so you can easily click between them and cPlay, and then not forget and leave them running while you listen to music.
It looks great at 1024 x 768 with M changed as Dawnrazor recommended and the desktop set to black using it as a border since it doesn't fill the entire monitor as you recommended. Also hiding icons behind cPlay and taskbar set to auto-hide.
Thanks guys
It also may be graphics-card related.I have my cMP2 in a Zalman case and use the "itzy-bitzy "embedded" monitor". I had this issue with the two GA-EG45M-UD2H motherboards I've used in varying degrees... they differed in how much of the window was off the monitor.
OTOH, no issue with either the GA-G31M-S2Ls or GA-G31M-ES2L I've used.
All were as fully optimized standard cMP2 setups as we know how to do today.
Greg in Mississippi
P.S. Ryelands, sorry, have had your cable for about a month now, but too busy to do anything more than take it out, look at it, say 'yup, that's a pretty cable', and put it back. SORRY!
Edits: 10/15/09
[I] was wondering if others have had the problem . . .
We all have the same problem - except for those who use the itzy-bitzy "embedded" monitor. (The display resolution is fixed for that format and is not scalable.)
. . . and if there is a possible fix.
Not so far. As you say, it's not a deal-breaker - just tell your friends that all the best people call Mozart's 41st symphony "The Jup".
Dave
Thanks
Just looking for opinions of what has worked best for you.
Thanks, Brad
If the files are already tagged (downloaded from Linn, HDtracks, etc), I simply use the foobar cuesheet creator plugin. For those that aren't, I recommend Recursive Cuesheet Creator by alrjordan. It's smart and very easy to use.
"We should no more let numbers define audio quality than we would let chemical analysis be the arbiter of fine wines." N.P.
Edits: 10/11/09
Added foobar cuesheet creator to components and it worked great, Thanks
Brad
.
"We should no more let numbers define audio quality than we would let chemical analysis be the arbiter of fine wines." N.P.
I should note that I used this post as a guide.
Brad
Please note that cMP/cPlay have gone through substantial improvements, and the incompatibility "[n]one of these multiple track cue sheets can be read by cPlay" observed by seger no longer exists.
"We should no more let numbers define audio quality than we would let chemical analysis be the arbiter of fine wines." N.P.
Refer to ReadMe.txt for important details & specifications.
Change Log (2.0b31 release):
- DSP refinements
- Threads optimisation for use with cMP's "Realtime" Optimise setting. Recommendation of "Critical" setting remains however significant efficiency gains makes "Realtime" a strong option.
This allows for complex decoding (i.e. FLAC) to occur without affecting playback in cMP. Depending on soundcard ASIO driver, "Realtime" setting could be better.
- SRC efficiency improvements reducing CPU load and L2/L3 cache footprint
- Minor refinements
- Added support for "Tiny" DSP Buffer Size. Previous buffer options remains the same with "Auto" using either "Small", "Medium", or "Large", i.e. "Tiny" must be manually selected.
"Tiny" buffer option reduces L2/L3 cache footprint enabling lower specification CPUs to be used with resampling. The 2MB L2/L3 cache requirement remains to produce highest quality 192k output.
cPlay documentation can be found at the new cMP² website . cPlay 2.0b31 offers significant SRC processing improvements and remains the reference resampler. Those not using SRC should re-evaluate this. "Tiny" buffer setting is also highly recommended (especially in combination with SRC).
Please REMOVE previous versions before installing cPlay 2.0b31. Normal (SSE2), SSE3 B9, SSSE3 B9 and SSE4 B9 versions are available. Running cPlay on a CPU not supporting the required instruction set will cause cPlay to exit immediately. For best results, uncheck "Timer" setting and test different "Buffer" sizes.
Putting things into perspective
Changes are readily observed with each new release. What follows is a technical explanation of why this is so. Modeling (Periodic) Jitter Theory for a wide range of J pp using a frequency sweep of the audible range and beyond yields the following:
![]()
Practical implications are profound as such Jitter Distortion damages sound quality:
- Sideband distortion rapidly increases with input frequency, i.e. high frequency sounds are most affected. This is seen in the rapid rate of distortion increase as frequency rises. Critical harmonic sound information in this HF band (delivered by tweeters) is distorted. This has detrimental effects such as veiling, poor transients and poor tonal decay.
- Most important insight is the extent to which J pp must be reduced.
- We see that for high J pp (150..350ps and beyond) an audio designer is not well rewarded for Jitter reduction efforts, i.e. despite large scale drop in J pp , distortion levels remain stubbornly high. Modern DACs (including some soundcards) have an excellent noise floor of -118db or better. This implies that jitter distortion arising from HF input (above 5kHz) will be well above such noise floor. For example, an input tone of 10kHz and J pp at 150ps would yield sideband distortion of -112.6dbFS. Ouch!
Interestingly this explains why some reviewers unexpectantly express disappointment with front-ends offering ~200ps J pp performance. That is, the noise floor is polluted with excessive sideband distortion.
This may also explain why some vendors have become disillusioned with Jitter and have abandoned it's role as a major source of audible distortion. Sadly, others have yet to understand how to measure such distortion (which can only be done correctly at the analogue outputs).
- Implications for poor setups with J pp above 150ps suggest sound quality improvements will not be easily experienced. Perhaps this explains why some object to for example RAM size, quality & setup to have sound quality impact.
- Much to our relief, distortion drops rapidly below 150ps J pp . As can be seen, for each 50ps drop, distortion reduces in ever increasing drops. That is, we see an exponential decay. Consider for example 50ps J pp , any improvements here will have factors more benefit than with same improvement at higher J pp . This suggests high quality setups like a fully specified cMP (whose measured jitter performance is 51ps J pp RSS using foobar) would be significantly better at revealing improvements.
- For the ideal case, jitter performance results in sideband distortion for all input frequencies below the DAC's noise floor. A noise floor of -130dbFS, requires jitter performance below 10ps J pp ! Intel's new 32nm Nehalem platform is a good step forward.
Rarely does theory and practice meet so beautifully. The efforts of Julian Dunn who laid the foundations for understanding Jitter and its challenges is truly remarkable.
Can you add .APE (monkey Audio) files support, may be in new versions?
cPlay 2.0b31 Released ,plese!
I just got around to installing 31 today.
It is amazing how different versions can change the sound pretty radically.
mixed is a word I would use like a previous poster did.
SOme things like the resolution is dramatically improved and you can hear more into the recording. That comment about a light being turned on is so true.
The soundstage is a bit weird on my rig with 31. The center is so prominent now that it kind of excludes the other areas of the soundstage., but the center does sound really in the room.
Warmth is missing compared to 30. I miss that. And as I was listening to 30 and thinking how I missed some resolution, and now that I have it with 31, I miss the warmth.
I just put it on tiny and that is what I listened to. But maybe that isnt the best setting for my cpu?
Are there any settings that can get me "30.5"?
I am using a latency setting of 32 with tiny and src 145db set at 192k.
And do take this with a grain of salt. I just got my cmp2 up and running a month or so ago and now I have new amps and in the middle have changed speaker positions a bit. SO my memory could be off, but I did listen with 30 and then install 31 and listen.
If I had to pick it would be a tough choice. I think i lean toward 31 but on some recordings I miss 30.
Thanks
Sometimes lowest is not "best". On my system, latency setting of 48 sounds the most powerful and "precise", but falls short in terms of "warmth" and treble smoothness. I found that 128 sounds the most pleasant, natural, and the bass is better than any other source I've heard in my system.
Incidentally, latency setting of 128 is also what cics recommended for 192k. He said, "for 192k output, latencies to 128 samples (ASIO Hz of 1500) will be excellent. This assumes there's no "dead bands" in PLL, i.e. PLL is inoperable for certain jitter frequencies."
My other settings are: Tiny, SRC 145db, 192k, AWE, Timer Off
"We should no more let numbers define audio quality than we would let chemical analysis be the arbiter of fine wines." N.P.
It just was more musical and 31 often left me uninvolved even though I was hearing more.
But I never tried 128. The cplay diagnostics reported that the Lynx liked 64 so I tried that and 32 and tiny, small, and medium. It just never sounded as good as 30. Perhaps I should give 128 a try.
I'm not sure if it's a glitch, but when I switch between different latency settings (i.e. 48 and 64), the cPlay diagnostics *always* report my current setting as "preferred". Hmm...
"We should no more let numbers define audio quality than we would let chemical analysis be the arbiter of fine wines." N.P.
Hey, I was on 32 and it reported 64 was preferred so I tried 64. So I dont think I have that issue.
she needs to make up her mind!
Quick update, I now prefer the 48 sample setting after remaking my I2S cable with RG179 coax. Digital audio works in mysterious ways...
"We should no more let numbers define audio quality than we would let chemical analysis be the arbiter of fine wines." N.P.
SOunds like!
I tried 128 with tiny, and that does seem to be much better. Not sure i will stay with 31 but right now it is pretty good.
Glad you like it!
In one of his posts, theo stated that "my system has many degrees of adjusting tonal balance ... [so] when I make changes I always prefer to go the way of better dynamics then adjust the tonal balance accordingly."
I happen to have very similar beliefs when it comes to choices of equipment/playback software. I believe that certain qualities, such as warmth, tonal balance, even better soundstaging, can be had through relatively simple tweaks. On the other hand, attributes like detail, dynamics, and low noise floor are more difficult to obtain. Oh, and I also happen to prefer to do things the easier way!
"We should no more let numbers define audio quality than we would let chemical analysis be the arbiter of fine wines." N.P.
This combo at src/192 is nothing short of amazing, I just love it. Imaging, dynamics and a sense that a light is turned on at the back of the soundstage---so much more to hear but smooth, very smooth. Congratulations.But I have some questions regarding some anomalies. I get zero metallics but I do get an occasional what I call a 'static-ee' noise which goes away via going to another track or changing sample rate. Others get this as well. I got the same noise when I was running juli@ analogue outs. If metallics were a 10 as an issue the static-ee is a 3: far fewer and easier to fix but annoying just the same and does not require a reboot. So upon throwing this around in my head I believe it may have the same root cause as metallics but since the system is no longer using juli@ drivers it is just manifested as
different. But assuming this is correct here are some more observations: when I get the static’s my cpu usage is typically 50 % with spikes up and down. But since Sunday when you asked me to run 48/tiny I have gotten zero static’s so I thought it was the combination of 48 / tiny parameters that was doing it (and maybe it is) but I also checked cpu usage and I'm back down to 25%. This same phenomena occurred back in June when I was running spdif out to the Benchmark: for a wonderful week I was getting no metallics while getting 25% (when prior to this I was getting 50% cpu usage and lots of metallics). So I think there is still an issue with juli@ (albeit minor with I2S out) that is somehow associated with cpu usage jumping around from 50% to 25%. This am I had to reboot after being on for 3 days and upon reboot I got 25% cpu usage but with spikes up to 45-50%. After being on continuously for a while the spikes started to settle down in the 35-40% level but come back to a higher level after not running cplay for a while but with the pc still continously on.Edit: my mistake on the cpu usage bit, I found out my pc reset unbeknownst to me my bios hcc to auto and I got clock speed of 1600 when I reset back to 840 cpu usage goes back up to 45-50%. When I reboot I normally sense a reset of bios when It takes long to boot up, I missed it. I hate computers. I love cplay/cmp.
Not to throw cold water on my glowing review of cplay 31 @ tiny/48 (and it is nothing short of magnificent) what in your opinion could be happening here with the noise issues and cpu usage which seems to be correlated. Look if I had to live with this situation as it is now I would be happy for a long time. I'm just trying to understand if juli@ is fatally flawed and whether I should pursue another sound card. I was tempted by the Lynx but w/o anyway to run battery power or I2S out I fear the Lynx may be too limiting now that I have experienced LiFePo4 batteries and I2S out.
Edits: 10/20/09
Which type of cable(s) are you using for the I2S? I currently use Cat5e for my I2S -> DAC connection. While I can't complain about the sound quality, it certainly couldn't hurt to get "proper" cabling. I'd love to hear your thoughts on this.
Thank you in advance!
"We should no more let numbers define audio quality than we would let chemical analysis be the arbiter of fine wines." N.P.
It is 75 ohm coax with teflon dielectric. I believe it is bg 179 (or something like that). It costs about $40 for 4 feet of it. Its very delicate but fairly easy to solder right onto juli@ pins.
Thank you! Sounds like the cable (or something similiar) in the link below. Can you please confirm?
Oh, and did you ground the braided shield or make twisted pairs with the cable?
"We should no more let numbers define audio quality than we would let chemical analysis be the arbiter of fine wines." N.P.
yes it was rg not bg 179. I used the cable as a coax: soldered the center connector to juli@ pins & grounded the shield to a juli@ pin on the 10 pin connector. I twisted all the shields together and fastened a wire from the twisted bunch of 3 shields to a ground pin on juli@.
Thank you theo.I remade my I2S cable yesterday following your instructions, and it sounds good so far. Overall, it sounds cleaner, more detailed and resolving. However, the presentation also becomes a bit bright and forward, making vocals less enjoyable. I wonder if this is due to the lack of burn in. Does my experience with the RG179 coax sound familiar to you?
"We should no more let numbers define audio quality than we would let chemical analysis be the arbiter of fine wines." N.P.
Edits: 10/28/09 10/28/09
Not sure...I was breaking in a new sound card as well as the I2S cables. Give it a few days then pass judgment. But yes it did mellow out eventually.
Edits: 10/29/09 10/29/09
Thanks, I think it already improved noticeably during the past 24 hours (I left the system on the entire time). By the way, the 30 AWG conductor of the RG179 seems *really* fragile. I used some hot melt glue on my Juli@ to ensure I don't tear the I2S cables off by accident. Ugly, but it works!
"We should no more let numbers define audio quality than we would let chemical analysis be the arbiter of fine wines." N.P.
Edits: 10/29/09
why not solder?
Glue on top of solder. I'll post a pic when I have time.
"We should no more let numbers define audio quality than we would let chemical analysis be the arbiter of fine wines." N.P.
Edits: 10/29/09
I was thinking of using a usb cable for I2S, just a thought :).
Brad
After intensive listening and adjusting (in a tri-amped system one has many tonal balance options) I have currently settled on medium buffer setting with juli@ panel @ 64 latency. I also needed to adjust subwoofer and midbass levels several iterations to get it right. With 31 I get great balance at low levels but at FFFF levels the highs seem to come forward and be prominent (not edgy) . Edit: ...Subsequent listening and I find this phenomena has diminished...I can understand why many have mixed emotions re 31 because w/o the adjustments I have, I may not have been able to dial in an acceptable tonal balance. I am now listening to cmp set to Realtime (rather than Critical) and will evaluate this based on your comments Cics. But what really does this do versus the Critical setting? My memory is that Critical was optimum, is that still correct for juli@?
Edit...Critical is better...
Edits: 10/17/09 10/17/09
Optimise set to "Critical" is the recommendation. "Realtime" can work with a minimal ASIO driver - one needs to test this. I prefer Critical with my Lynx.
It seems that you are changing too many things at once (and you have a new Juli@). Sound will change with more burn-in. Try this: in cPlay, set buffer to "Tiny" and set Juli@'s latency to 48 samples - gives this at least 48 hours.
You are right I believe my juli and my teflon coax (for I2s) are breaking in. I had been listening to 64 latency and tiny buffer for last several hours and its great(I mean absolutely MAGNIFICENT). Right now per your suggestion I switched to 48 latency and will evaluate this for awhile. When I loaded 31 I could not run 48 latency with tiny buffer (it stuttered) so I believe it was breakin all along.
But so far after several days the sound is simply the best I have ever heard in my system. The anomaly with the highs (dynamics shifts back & forth) was simply a breakin phenomena. The imaging is nothing short of superb. Later I will comment on the tiny/48 combination with cplay 31.
I have been down for 2 weeks. Events had overtaken me. In an attempt to implement I2s I hooked up my LiFePo4's backwards and fried juli@. Took me two weeks to get a new juli@. In the interim I lost my 96 year old mother...lots of stuff to take care of. I got my juli@ and just now loaded cplay 31 (still with spdif into Buffalo)...wow and I'm not even to 140 host clock control nor 3 3 3 5 on the memory. When I'm down for a while I have to creep back to optimum hcc and timings. But nevertheless 31 is great. Can't describe it yet but it sounds very very natural. Lesson learnt for me and hopefully for others KEEP POLARITY CORRECT ON BATTERY SUPPLY. More on 31 later, Initially very very natural
Hey Theob,
Sorry to hear about your mother. That is sad to hear, and kind of puts this hobby into perspective.
Thanks for keeping us updated and for the tips on the Juli@. One of the nice things about that card is that it is such a great performer for the money, so accidents are recoverable.
One of the reason i kind of have to run my lynx stock...cant afford to screw it up!
Thank you for your kind words. I appreciate it. I can't afford to screw up the Juli@ BUT I did, hopefully nobody else will.
For those of you who have followed my issues (& I had a lot) I could not get spdif to work on my juli@ after a dreaded metallic. So necessity becomes the mother of invention...I had my 75 ohm teflon coax to try I2s but 3 weeks of trial and error I could not get the mechanics to work. But with help from inmates (bertel and wackybytacky) I finally got it to work. Big shout out to Sonics without his i2s sketch I could not have corrected my incorrect hookup to Buf32. Alfred and Robert you are the man (er men)!!! Anyway I2s from cplay 31 is a significant step up in soundspace (wider and deeper) separation of instruments, fantastic (ie bigger) dynamics and lo and behold no metallics (omg what did it take me ---10 months from identification to solution). I'm running 192 with buffer set to auto and juli@ latency set to 64 (tried 48: too dry, tried 128: not enough air, 64 just right). Anyway 31 is unbelievably natural, open and dynamic. I'm not sure I am set on all the parameter setting but I am so glad to be back into 192 music! Thank you cics for 31: a masterpiece!
ps Robert: I think I am not going to try those 22 ohm resistors to damp the I2S signals...just a little gun shy now. I'll wait till you measure juli@ output impedance.
For those of you with Buf32s be carefull and learn from my mistakes if you decide to go I2s. Yes it was worth it but it took 3 weeks. The journey may not be for the faint of heart.
Sorry Theob. I was looking at things relative to the price of my Lynx.
If it is any consolation I think most of us have done some boneheaded things to destroy audio gear.
Why within the last year I toasted my Lynx card by wrapping it with tinfoil. Who would have thought it would have overheated? The damn pci slots end up making things upside down anyhow...
Fortunately Lynx replaced the op amps for a reasonable fee.
Forgive me, but I think I remember that we tracked down your metallics to your juli@. Do I have that right?
Yes the metallics are caused by juli@ drivers (required when using spdif out). However when going I2S no drivers are required and so no metallics.
I used to have a professional sound card (not Lynx, but the RME AES-32) in my cMP as well.However, ever since I found the I2S of the Juli@ outperformed my RME's AES, I sold the RME and never looked back. Then again, I didn't find the stock SPDIF out of the Juli@ impressive at all. The RME was significantly better than the Juli@ unmodded.
"We should no more let numbers define audio quality than we would let chemical analysis be the arbiter of fine wines." N.P.
Edits: 10/14/09
Hey FP,
I am using the dacs on the Lynx as I will soon be triamping and need 6 dacs basically. The Lynx is fantastic for the money.
Actually I ran out to my external dac and while it was pretty good, I thought the Lynx was everybit as good, probably a bit better given the price difference.
Are you doing DSP crossover inside the PC? If so, what software are you using for that? Or does the Lynx software allow you to do that? Pardon my ignorance. I am running six channels active but do the crossover in a modified Rane RPM 26z. That is fed by a Mac/iTunes/Amarra/Offramp but am reading this thread to see if I want to try doing it with cPlay/Lynx instead.
Hey B,
The plan is to use the thuneau Allocator (www.thuneau.com) with cmp/cplay.
The hold up has been the funds for new identical amps. I get them tommorrow and I can then start in earnest.
Play-mate and Bibo01 are ahead of me in implementing the Allocator with the Lynx. Hopefully I'll catch up soon enough.
Nice Job! With a couple of mentions of mixed reviews and 2.0b30 being so good, I was not sure what to expect. I was initially pleased to see that 31 clearly had not taken a step back in any regard. It seemed at first that there was an increase in detail and clarity. CMP2 has likely gone well past the ability of my speakers to resolve all that it is putting out, but as I listened over a couple of sessions it is clear that while it is a subtle difference, I am hearing much deeper into the music. It is again time to re listen to my entire collection. For my system, SOX has been the clear winner, but with this release the gap has closed to the point that I can easily flip between SOX and SRC with no loss of air and detail and enjoy the subtle advantages of each on given track. Its now closer to flipping phase. I run on a laptop with a celeron with very small cache, and Tiny buffer really works well.
Again thank you for sharing this great audio achievement.
I have the new release for some hours running and I am amazed, that after the last, great version again some subtancial qualities are obvious. Again more details. As stated by someone already, lyrics are easy to follow. Voices in the background are clearly separated. I am used now, to let the system run in for several hours, before I give a statement. Thanks to sics for this version. Within a few days I will have my buffalo-dac running and can not wait for the outcome. Walter
Tarkowsky wrote:
As stated by someone already, lyrics are easy to follow.
You're absolutely right - but it may not always be a good thing.
My wife was amused to discover last night (over 40 years after first buying the record!) that the lyric for Chuck Berry's "Tulane" has one word in a line that is not what the usual sources say it is. They usually give it as "Get me out this funky jail". They're wrong.
Worse, this morning, I was playing "Purcell in The Ale House" (an album of 17th Century "catches" and "rounds", i.e. light songs for private performance, often by young men, in "noble" houses).
I knew HP wasn't among the angels whose praises he so eloquently rang in his music but I wasn't prepared to hear staid English choral group Pro Cantone Antiqua singing a hitherto-indecipherable line (no lyrics in the sleeve note) to a song called "Once, twice, thrice I Julia tried, the scornful puss as oft denied" was in fact "Kiss my arse" sung repeatedly in the most delicate harmony one could wish for.
Whatever next?
Cics, thank you for your clarification re latency. It did become clearer under closer study. b31 has reduced latency to 55 using tiny with the L22. My impression of b31 is somewhat mixed, with the greatest regard to that which I can comprehend but never achieve by myself. The instrumentation is very clearly balanced but does not have the room interaction of b30. Base music has lost some of the string decay, which is to say slightly smeared. For example Patricia Barbers Mythologies. I have switched back and forth several times and am still burning in b31. Thank you again, T.
No worries!
congratulations again, cics....!
there is a remarkable difference in this release :
its very obvious that the sound is drier and less resonant. especially on multiple brass arrangements there is clearly more precision. maybe a tad less glossy and shiny, but equally transparent.
beautiful !
I so adored b30. But after 3 weeks in-depth listening, I had two reservations (SOX, HQ, 48K, Small Buffer) :
-- there was a haze in the high frequency that could inject extra space in some recordings. The resulting 3D effect could be quite attractive but ultimately felt false. For vocals that were mixed way at the back of the sound stage, the images are fuzzy.
-- for some recordings (I am thinking of 'Astral Weeks' as an example, the Japanese remastered version) the emotional impact of the instrumentation was not quite there when compared to the LP version (again Japanese vinyl version). The attack of instruments such as congo, high hat, guitar etc. were soft and I couldn't feel the human behind the playing. Please note that the emotional content of the instruments is a big thing in 'Astral Weeks'
With b31, I now switch to SRC (tiny buffer, 48K, 121db) and the above too 'problems' are mostly gone! The high end haze is no longer there and the emotions are mostly back in 'Astral Weeks". And the bass is now more solid.
I find myself listening to the music rather than the hifi, finally, after all these years in the digital world.
Thanks Mr. cics.
In the past year (plus) with the many cPlay releases, the difference in SQ has ranged between subtle to a little more noticeable (and I'm a "Sharpener"), but (I may be exaggerating to say this) but wow, the difference between Tiny and Small is about the most dramatic difference I've heard. And I haven't even compared it to Medium and Large yet.Dawnrazor, you'll probably appreciate this: I started to listen to "Loveless" by "My Bloody Valentine" and was floored at how I could finally hear the lyrics in "When You Sleep" (without it being smeared by all the guitar distortion).
A usually "dark" (pardon the pun) sounding "Masquerade" by "Clan Of Xymox" now sounds more intelligible, and a usually "bright" sounding "Four Calendar Cafe" by "Cocteau Twins" now sounds less edgy.
Edits: 10/08/09
edward wrote:. . .the difference between Tiny and Small is about the most dramatic difference I've heard.
Judging by your list of favourite music, there's not much risk of our getting in each other's way in a music store but, having just experimented with these settings, you're certainly not far out - the difference is very marked.
(What I haven't done is compare v31 set at Tiny with earlier versions which, of course, lack the setting.)
Meanwhile, it's worth visiting the link below.
Dave
Edits: 10/11/09
Hey Edward,
I unfortunately havent had any time with 31 on the big rig yet. I was really digging 30 after my recent cmp2 build.
Hard to imagine it being better but cics always amazes.
I can tell you that over the years recordings that i thought were crap have actually turned out to be some of the better ones as my stereo has improved.
I can totally appreciate your comments and hope soon to share in them.
Thanks for posting this and I'll see if I have similar results.
There is a line by the Church in "Destination" on their starfish album that I have always been unable to hear. I hope this version reveals it, but I just think that might not be intelligible any how. We'll see.
D
Well, Dawnrazor, I know you and I share the same musical taste, and some of those 80s recordings that we are talking about are just dreadful. This past year has been rewarding though, as I have picked up "remastered" versions of Dead Can Dance, The Cure, Tears For Fears, Yaz, Joy Division and there are also remasters of Skinny Puppy and Cocteau Twins out there and Bauhaus is releasing a remastered version of Mask (and hopefully the rest of the catalog will follow). The Yaz box came with a DVD that included 24/48 versions which sound awesome. In fact all these remastered versions I mentioned sound awesome. I'm hoping that now that all these artists are doing remasters, that they digitally mastered at 24/192 and that these will some day be available. And I do thank NIN for paving the way, and offering "The Slip" as a free download (in 24/96 no less), but his music is so distorted anyway, that having the 24/96 version doesn't really offer you that much more.
But here's hoping that more of the 80s discs follow suit. (And I know my wife will pay good money if and when Duran Duran's entire catalog is remastered and/or available as hi-rez).
E,
That is awesome info. i had no idea all these remasters were around. i might have to buy some music for the first time in a LONG time. (OK I did buy that Sisters of Mercy "First And Last And Always" release that was supposed to be like the original vinyl release that was somehow different than the original cd release. But it wasn't remastered.)
Thanks a bunch for sharing.
FWIW your wife is better than mine in this case. When she saw me adding the cure cds to my shopping cart at amazon she said something like "you already have those cds, who cares if it is remastered. Its the same old song anyhow..."
I was nice enough to have added the new Michael Buble cd she has been dropping hints on, so I said, "Then why do you want that Michael Buble cd? Its mostly just covers of the same old songs..."
My wife is the perfect wife in almost every way. Audio is her one flaw...
Ironically I have enough pcs lying around that I was thinking of making a rip box to see if there were better ways to rip than I did years ago. Now I will have even more reason to go to the trouble!
Thanks again!
.
.
Hi Cics,
Thank you very much for the b31 update!
My feeling for this version is quite mixed: This version's accuracy and sound stage is the best thus far. Sound stage is so accurate that I can feel the singer standing in front of me. More micro details are revealed when playing back complex classical (orchestra) music. The other advantage to me is richer bass but not boom (well I only use JM Lab micro Utopia so other experts who have floor standing speaker will provide more relevant comments). However, the sound is "too accurate" that musicality is affected, especially when compared to b30. Therefore this version is good for classical but a bit coarse for female vocal. Therefore, I need to check "small buffer" in stead of "tiny" in order to smooth out the sound a bit. Other settings remain the same (upsample 192, VHQ, Linear 95, no alias). I still miss the female vocal in b30.
Thanks again for the good work!
Try adjusting juli@ panel latency. As I adjusted from 48-> 128-> 64 I found the impact on dry or etched to detailed and liquid very profound. I prefer 64 but you may like another setting.
You should try SRC, as cics recommended. Having experimented extensively with SoX settings in the past, I have a strong feeling that SoX tends to outperform SRC in certain areas, but the latter is simply more pleasing and musical overall.I will report back after I have the chance to properly evaluate the b31 release (new DAC still burning in).
"We should no more let numbers define audio quality than we would let chemical analysis be the arbiter of fine wines." N.P.
Edits: 10/08/09
Cic's wrote: Refined setups will readily reveal sound quality changes with changing latency. Its best to use lowest stable latency. A good programmer would argue that latency is a non-issue for playback. "Just set it to highest level as we get less context switches which is more efficient...". This is not correct for best SQ (Sound Quality).
After months of playing around with Cplay/cMP/Juli@ settings i have to conclude that in my config high buffer settings sounds better. When i choose small buffer in cplay or the smallest latency at the juli@ panel, the sound looses musicality en smoothness. Maybe its a bit more refined, but less enjoyable.
What is your experience with latency settings?
For 192k output, latencies to 128 samples (ASIO Hz of 1500) will be excellent. This assumes there's no "dead bands" in PLL, i.e. PLL is inoperable for certain jitter frequencies.
B31 will bring important SRC processing improvements. Would be very interesting to see how it works in your setup.
Does this apply to all cmp users, or Lynx card owner only? Do you recommend setting 128 samples @ 192k for everyone?
Thank you!
"We should no more let numbers define audio quality than we would let chemical analysis be the arbiter of fine wines." N.P.
It's not specific to Lynx cards but directly relates to output rate. At 192k, latency of 128 gives 1500 ASIO buffer loads likewise, at 96k, latency of 64 samples gives 1500 ASIO Hz. The idea is to ensure this periodic "interference" does so above the PLL cut-off.
I find with Lynx cards, its best to disable its Synchrolock feature (digital PLL for external clock sources).
Thank you for the explanation. Forgive my newbieness, but what about those of us using I2S over SPDIF? Should we go for the lowest latency possible, or stick with 128 samples like you suggested?
"We should no more let numbers define audio quality than we would let chemical analysis be the arbiter of fine wines." N.P.
You need to test which latency (32..128) is best - this depends entirely on the PLL performance specific to your equipment.
Thanks, I'll do so when I get the chance.
"We should no more let numbers define audio quality than we would let chemical analysis be the arbiter of fine wines." N.P.
Edits: 10/08/09
I am having some confusion with setting latency. I have the Lynx L22 sound card using digital out to a dac. When I review the Cplay's Guide for ASIO drivers it indicates an ideal setting for the Juli@ of 48 in "Buffer details" and "Latency". How exactly does one set latency. My buffer details are( min 32 max 1024 preferred 32 granularity -1), Latency is (input97, output 72). Is there a way to adjust these settings. My PC is a Cplay dedicated modified with the exception of advanced mods, Gigabyte ES2L mobo. T.
No worries!
Your setup "(min 32 max 1024 preferred 32 granularity -1)" is perfect. The key item to look for is "preferred" which is set to 32 - lowest for Lynx.
Latency is adjusted using cPlay Settings > ASIO Settings > Lynx Mixer settings .
I noticed this at DIYAUDIO in the commercial section:
http://www.diyaudio.com/forums/showthread.php?t=151630
It seems to be on the right track. First thing one thinks about is powering it properly.
I personally would prefer for conversion to be done outside of the computer but one cannot be too sure of their assumptions!
Looks interesting. Although there are a few points I'm unclear about.-The driver appears to be a WDM driver with a patch mixer that will route other driver interfaces (ASIO) to their driver. Is it just me, or does this sound like the same thing as ASIO4ALL? IOW, it does not sound like a true ASIO driver. Or maybe I'm just mis-reading the ESI description. I'd like to know if it's truly ASIO (and 32-bit).
Edit: Doh! I just realized this is from the makers of the Juli@ So drivers should be fine, right?
-Also, they claim supporting up to 24/192 however I've seen other products make this claim and then not support 24/176. I'd like to confirm that they indeed support all rates (44.1, 48, 88.2, 96, 176.4, 192).
Can anyone answer these questions?
Edits: 09/27/09
> The driver appears to be a WDM driver with a patch mixer that will route
> other driver interfaces (ASIO) to their driver. Is it just me, or does
> this sound like the same thing as ASIO4ALL?
> IOW, it does not sound like a true ASIO driver.
> Or maybe I'm just mis-reading the ESI description.
> I'd like to know if it's truly ASIO (and 32-bit).
You are just reading things into the description that aren't there. The ASIO driver is integrated with the WDM driver. if I had written the drivers, that is what I'd have done.
You just select ASIO output from your player. You do not need to patch anything with DirectWire. ASIO output to the AudioTrak Prodigy HD2 has been flawless for me in a Win XP system with J. River media Center 11/12/13/14.
Outputting at 32 bits from J. River MC seems to work.
> -Also, they claim supporting up to 24/192 however I've seen other
> products make this claim and then not support 24/176. I'd like to
> confirm that they indeed support all rates
> (44.1, 48, 88.2, 96, 176.4, 192).
I played material in J. River MC at various rates.
44.1/16 - works
88.2/24 - fails (error message from JRMC.)
96/24 - works
176.4/24 - works
192/24 - works
There are several threads on op-amp choice on the head-fi computer audio forum.
http://www.head-fi.org/forums/f46/
Bill
Thanks for setting me straight on my ASIO confusion Bill. I guess I just never heard it described as integrated with the WDM driver before. (Or I'm just not paying close enough attention).
I still find it curious that 88.2 doesn't work and yet 176.4 does. (Or is this more common than I realize - and I'm just being dense today).
That is weird. SO far I only have direct experience with 2 cards, Lynx and M-audio Revolution. Both could handle anything between 44 and 192.
I am able to output at 24 bits / 88.2 KHz to the ESI Juli@ and it plays.
ESI makes both the Juli@ and the AudioTrak Prodigy HD2. The console apps look similar and the driver files have similar distribution of function. I haven't a clue why 88.2 doesn't work for the AudioTrak card.
I will point out that supporting rates higher than 48 KHz for playback was mostly a checklist feature with little use for consumer audio applications until recently. Both Juli@ and the AudioTrak cards were designed several years ago.
Right now, there is just one 88.2 download I have any interest in acquiring (Mackerras and the Scottish Chamber Orchestra playing Mozart symphonies 35-41). Linn sells this set as 44.1 Flac downloads or as 88.2 downloads for considerably more money. I'm not keen about paying more for a higher quality format that will be converted to a different rate to be played on my office system. The effect has been to keep me from buying these performances in either format.
Bill
Found this review:
http://ixbtlabs.com/articles2/multimedia/prodigy-hd2-gold-page1.html
Thanks for the link.
Odd that under ASIO it shows 88.2KHz as "Not Supported", yet 176.4 IS supported. And it does not support "Output Ready". But it looks like it is ASIO 32-bit.
Maybe I am missing something here, but that doesnt seem like a card that fits into cics principles, especially cmp2.
The lowest latency supported I saw was 256. Wouldnt we want a card that did 32?
Are we still talking about the plain AudioTrak Prodigy HD2 card?
> The lowest latency supported I saw was 256.
> Wouldnt we want a card that did 32?
I normally use a setting of 256 for latency. I tried 48 just now and it worked fine. That is the lowest setting in the Prodigy control panel.
I use the drivers written for this card and supplied by ESI/AudioTrak. It is possible to use generic Via drivers but that might require re-flashing the firmware on the card.
The fancier versions of this card may come with different firmware and work only with the Via drivers.
Bill
Hey Bill,
Yeah, I think that is the one...whatever was in the link.
Anyhow, it is good to know you can go lower. Maybe 48 is ok?
I just set my Lynx to the lowest setting which is 32 and that works great.
I guess the easy solution is still not there.
Lots of compromises that are not very attractive.
Oh, well. Thanks,
Anybody getting an ocassional dropout on cplay/cmp? Seems to have increased in frequency the last few weeks. If not what could it be, dirty connections? This is not a cpu interrupt hiccup its something else I believe. Sometimes it manifests itself as a stutter. Anybody else experience this?
Edits: 09/17/09
I don't know if it's the same thing as what you are talking about, and I've been waiting to mention it to "cics" until I tried to better diagnose it myself, but here's what's happening to me.
I don't know if it's a new issue, because I've only recently (in the past month) been resampling with SRC. Previously I used SOX or no resampling. But I think my issue is related to SRC. What happens is that when a song transitions to it's next segment or sometimes to the next track, then it stutters. (What I mean by "segment" is how the file is broken up into segments and delineated by the number in parentheses). I thought it was related to latency/buffer by my sound device so I haven't said anything before, because I was playing around with it. But one thing I did confirm, was that the stutter always happened at the same time. So I played a track that had (2) segments and it always stuttered at the same time, and then I put a 512MB RAM chip in so now the same track was not segmented and it played without a hitch. Changing the latency on the sound device did not seem to matter, and I was not experiencing any other dropouts at the lowest latency setting. FWIW I am using a USB device at the moment, so I thought USB was the culprit.
Any ideas?
This only applies to upsampling with SRC@145db. This stutter is eliminated when running cPlay with cMP. Also within cMP, Optimise should be set to Critical or Realtime.
Here's a test that I always run for gapless:
![]()
Transitions are totally seamless within cMP (except when Optimise is set to Player). Without cMP, its difficult to eliminate.
Yes, of course, I am running cPlay within cMP and Optimise is set to Critical. (And I'm booting straight into cMP). I've also made sure Cryptographic service is not running.
Do you have any other suggestions? I'll keep poking around to see if I come up with anything.
Well, I figured it probably wasn't the same thing as Theo and I think I confirmed it's just my POS USB device. I tried a few other things and then went back to the latency settings and cranked up the latency to 512 and now it plays seamless.
I'm broke right now, otherwise I would have bought a PCI card by now. I've been using the USB device because I need to listen with headphones on. (I only have time to listen when my wife and son are asleep - IOW I'm nocturnal).
Well, let me know if you have any other ideas to improve my current situation.
I don't know if my stutters/dropouts occur at segment transitions but I will try to observe. Lately I have not been using upsampling and I get the problem just the same. This is a fairly new problem...did not get it before. It is unpredictable...does not always happen.
Edits: 09/18/09
Edward I was popping fairly often this am so I checked Control Panel and that stubborn "Cryptographic Service' had re-enabled itself. So I stopped it and my stutters stopped (at least for a 60 minute file playing). But sometimes this phenomena just doesn't occur so I cant be sure it is root cause or not but check it out.
Cryptographic Service should be set to "Manual" but Windows will start it whenever a new device type is plugged into USB or doing a detailed query on a device. It's always good idea to check that only the 2 required services are running (RPC & Plug n Play). Some soundcards may require a 3rd service (Windows Audio).
With "Cryptographic Service" running, there's a notable deterioration in sound quality!
Wow, thanks for another great tip. After disabling 'Cryptographic Service' (and a few others like 'Spooling Service') through Administrative Tools, there is a definite SQ improvement such that some previous CD's that I thought were inferior (to the LP sound) transfer are now OK. Some of the missing emotional elements are back. Once again, thanks.
I found that the services do come back if you just set them to 'manual'. What I have to do is to set them to 'disable' in 2 places:
1. In the 'Services' window in 'Administrative Tools', and
2. Right click on the specific service, go into Properties, hit the 'Log On' tag, high light the item in the 'hardware profile', and then hit the 'Disable' button.
After power down and reboot, the services remind disable.
Refer to ReadMe.txt for important details & specifications.
Change Log (2.0b30 release):
- DSP refinements
- Optimisation for Intel's Nehalem platform (Core ix CPUs)
- ASIO Settings activated on invalid latency or output
- Other low level refinements (SRC & kernel)
cPlay documentation can be found at the new cMP² website .
Please REMOVE previous versions before installing cPlay 2.0b30. Normal (SSE2), SSE3 B9, SSSE3 B9 and SSE4 B9 versions are available. Running cPlay on a CPU not supporting the required instruction set will cause cPlay to exit immediately. For best results, uncheck "Timer" setting and test different "Buffer" sizes.
...I also got a Buffalo 32s running the digital power supply off of 3 LiFePo4's in parallel and yes I confounded several changes and I revoiced my system after all this new stuff was broken in. Only thing I haven't done is I2s with the Buffalo directly off of juli@. Several things right off the bat:
...yes Robert you are right the Buffalo is heads and shoulders better than the Benchie.
...the unprecendented highs,dynamics and tonality I am getting is making me (yes its a cliche) listen to everything all over again. This is what being an audiophile who tweaks is always waiting for: something that breaksthrough that sonic limit barrier. I love it!
... the number of Lipo cells one uses for either juli@ or Buf32s has a striking impact on sonics. The amount and kind of bypass caps one uses on the Lipo's also has an impact. I use 2 (3.3 volt 2000 milliamp hour cells)on Juli@ and 3 on Buff32s. Don't know whether this is an absolute optimum but it suits my system best. Also I use 6-10K uf electrolytic caps, 30 uf of polypropylene and .047 uf polystyrene caps in parallel with the LiFePo4's.
Now here is the real biggy for me I'm listening to cplay @ 44khz and prefering it to 192. You guys know I have been an unabashed supporter of 192 with -145 src but not now. Listening to the Benchmark from 44 was painful, it needed 192 to sound decent. But the Buf32s is so smooth I prefer it at 44. The airiness is less but still abundant and the precise sense of placing images on a classical music piece is a lot better and the bass and mid incisiveness is undoubtedly better. I had my wife listen to both and she said she preferred 192 because it sounded more mysterious than 44. To me 192 now sounds too ethereal (too impressionistic) or imprecise but pleasant. I could still go either way since I am not declaring dogma here, I'm just saying this is where I am right now. When I go I2s everything may change. Needless to say my metallics are gone @ 44khz. I still get (but less frequently) dropouts like Edward but I have traced it to my method of transmitting battery power to the digital supplies. So I think I can eliminate it.
After all of the above cplay 30 is wonderful: less edgy, less bright, more tonally and dynamically correct than its predecessors. No I did not go back and listen to my new setup with cplay 29 and won't. I love cplay 30.
One further thought. Since my system has many degrees of adjusting tonal balance (triamped, electronic x-over on bass) when I make changes I always prefer to go the way of better dynamics then adjust the tonal balance accordingly. Cplay b30 has best dynamics yet. Way to go Cics.
Because I had been away for 2 weeks so the system took a few days to warm up before I can give a real listen to the new version. What I find:
-- smoother high
-- much better high mid frequencies. Such that a lot of the recordings with weak middle, e.g. Paul Simon's 'Graceland', are much more listenable now.
-- tighter bass, I think,
-- the sound space is much more expansive. So now for majority of the recordings the listener is IN the sound space, instead of sitting in front of it, with sounds coming to him in all directions; when previously it only happened to a few rare recordings with lots of space information.
Thanks again cics, for bringing back the fun of listening to recorded music at home. Nowadays, with cPlay, I find myself too busy listening to miss the sound of my long dismantled LP system.
Hi Cics,
Thank you (again) for your work and this new version. There are many subtle differences with the last version which all together make this version a very nice one. I’ll try to describe them.
There is a whole new level of details. Percussion, cymbals, brushes, ect have a very accurate ‘attack’ and are very very detailed. But without being scharp, hars or edgy. The combined titanium/fabric Seas dome tweeter in my setup easily tend to sound somewhat sharp and harsh. But no trace of that with this version !
The new level of detail also lets me hear ‘deeper in to’ the music (i.e. the mix / recording). Individual instruments are better separated and there is more soundstage and image.
This B30 is a very nice and enjoyable version ! Trance, dance, jazz, classical, al types of music sound very good !!
LynxL22 dig i/o XLR AES/EBU -> Lavry Black DA10 -> XLR Mogami Gold -> Klein & Hummel O300
Could not agree more. Again I am surprised at how you continue to raise the bar. A level of realism has emerged that is quite surprising. Particularly in female vocalists. I have listened to some of my favorite Jazz and now opera, and in both cases its not only a relaxed realism, but a larger measure of emotion is evident.
Each time you have released a new version, its like someone drops off a new higher priced transport at my door with a note saying here, try this!
Great work, and again thank you.
In terms of SQ, the latest cPlay release is the best playback software that I'm aware of (compared with pro audio software such as Wavelab, Sonar 8, Samplitude, etc). And thanks to the cMP^2 "package", my computer transport now puts all my absurdly (relatively speaking) expensive CD transports to shame.
Thank you for making "hi-end" audio more affordable and enjoyable.
Wonderful Ver 30. Etheric sound... non tiring... Best timing
Thank you Cics
Question for Cics! A conundrum my main pc had a MB flatline, I had located an e7200 and an S2l but know I question the purchase. I was going to swap out my PC audio asus and 6300 as my output is native digital 16/44.1, lynx l22 to a benchmark dac 1, but I was restricted from using sse4. My question is with the new platform should I wait and what cpu would you recommend with the new Nehalem platform. Everything was easier yesterday. Lastly my impression of b30, most noticeable was the sound stage was both deeper and higher, more of the recording stage was present(Paolo Fresu and Uri Caine/Think) T.
No worries!
...but I was restricted from using sse4.
My question is with the new platform should I wait and what cpu would you recommend with the new Nehalem platform.
Hello, cics, and everybody. I am glad, that with this version resolution comes to a new level with much of the upper mids' dip finally flattened. (at least in my humble setup). The upper bass also got a minor facelift and overall sound is back to the "snappy" level of b28, brass and vocals are very good, indeed. Placement of instruments more pronounced.
Every time new version appears, I wonder, how much is still hidden from us by conventional cd-players. I would compare this version with Wavelab demo as far as the resolution goes, only less grainy and without unnecessary sparkle. IMHO, this is the way to go.
Thank You very much. Also, very interested about forthcoming processor and corresponding mobo.
Serge.
I agree Serge, I too hear more resolution and better separation (especially noticeable to me with vocals).
Thanks again "cics".
Does this imply that the "iX" will become the new ideal platform for cMP^2?
New 32nm Core iX is certainly worth testing. The change made in cPlay could also benefit certain AMD CPUs.
Intel has made significant improvements in Nehalem and with the right BIOS we could see power consumption drop to just 10 watts doing full 192k with SRC@145db SNR. We should see Core i3 32nm CPUs in the next few weeks which can use DDR3 RAM (which is better than DDR2 using just 1.7V).
Could this be a better platform? Not sure as the current platform is magnificent.
Read the test results in the link. I think that the cMP^2 platform has a (soon to be) new champion. Imagine the underclocking/undervolting potential of this beast!
"We should no more let numbers define audio quality than we would let chemical analysis be the arbiter of fine wines." N.P.
Hi cics--still evaluating version 30 which is difficult because I just upgraded to a Buffalo32s and it is still breaking in. Anyway question on the Nehalem platform (Core ix CPUs) will they fit in the current recommended gigabyte mobo's or will they require new mobo's?
Core iX CPU's require a new mobo.
There are a few 'if this or that' on how they will be offered/implemented.
After a week of Buffalo 32s and cplay 30 I have to say I agree with the sonic attribute comments made by others. While I am not entirely free of metallics I can say they are much more infrequent than with the Benchmark. The depth of soundstage is extrraordinary and the high frequency timbre is a bit laid back (I think I will try 44 sampling rate again as Alan suggests). I find I am turning my electrostatic panels amp up to compensate and the thing is very very hot. I have been listening to 192 src -145. I don't think I have the optimum combination yet. But while I wait for my battery clips to run LiFePo4's for the digital power supply I will not make a final judgement. Still coming will be an I2s connection to replace my spdif.
nt.
Maybe my two cents can be of interest here too since I'm coming the same path that Theo has chosen (I have replaced my Benchmark DAC1 USB with, well, two Buffalo32S, since I more or less destroyed the first one...):For me the sonic difference, especially in transparency, depth and width of soundstage, but also tonal correctness and just being so "right" is HUGE compared to the Benchmarke, it just floored it in basically any aspect IMHO!
I must add that I run a fully optimized cMP2 system that runs entirely on battery with caps, and the Juli@ connects to the Buffalo32S via i2s (not SPDIF as in Theo's case, never tried this). BUF32S currently is also running on battery power, but for the time being not optimized at all since I am waiting for the TWEAKER version of it. I too feel that this is not the optimum yet, but it is exceptionally and so very clearly the best setup I had so far. Although BUF32S is a (very good) compromise to implement the ESS9018 chip, I can highly recommend it over many DACs and definitely over the Benchmark. (BTW I have not had any metallics in this setup, unlike with the Benchmark, although I meanwhile run the system at 140MHz and 0.76875V)
And BTW: It DOES make a HUGE difference in my setup when I feed BUF32S with 24/192 data upsampled with SRC instead of 16/44.1, it is so much better in again almost any aspect.
Regards,
Robert
Edits: 09/15/09
Hi Robert I think I finally have the Buffalo set up right and I am torn between 176 and 44 khz (src @ -145, hcc @ 145, cpu volts at .8 something). Running it at the higher sample rates (either 176 or 192) gives me an ethereal sound that is quite good in imaging but 44 gives me more incisesiveness while still being smooth in the highs, bass and lower mids are definitely better @ 44. But I don't have my battery clips yet to run my digital ps off LiFePo4's and I'm not yet I2s. The Buffalo is better in the mids and bass than the Benchmark and has more depth but if I didn't have trouble with cplay/Benchmark metallics I could live with the Benchie. I suspect my impressions will get way more favorable for the Buffalo once I further optimize. Metallics happen at about a 50% less frequency with the Buffalo (@ 176 that is, @ 44khz no metallics as cpu usage is 2%).
Speaking of further optimizing the Buffalo have you considered or done anything to tweak it?
Hi TheoInteresting how different our findings (or is it taste already?) are - I would not want to go back to the Benchmark now that I have got to know the Buffalo, and I wouldn't want it to do anything other than 24/192 :-)
No, as said I haven't done any optimizations or tweaks to BUF32S yet (apart from running it from battery which is not yet optimized in any way) since I'm waiting for the TWEAKER version to become available. This I think will be much more suitable for tweaks and optimizations since it (hopefully) will have no regulators, no output stage etc. thus being a much leaner and more precisely to handle component compared to the current BUF32S which is quite an integrated device (of course a very good one).
Well, that said, it may well be that I still start playing around with it and doing some tweaks and mods if the wait for the TWEAKER gets too long and/or I get too impatient, but I don't really have plans to do so yet ;-)
Regards,
Robert
Edits: 09/15/09
Still assessing. But so far equal to or better.
nt
thanks
This 2.0b29 release is the best for me by far!!!.
Until now, I have always found some problems in each of the different versions. Not about the resolution but on the purity of tonal timbres of the instruments, especially in the treble that I always seem too prominent and wheezing. I couldn't understand why.
Not with b29!. I believe that this is a wonderfully well balanced version between resolution and tonal purity.
Cics, I am nobody to tell you what you have to do. But I think, that now, as some have said in previous messages, it is time to develop a library interface to allowing us to develop diferents browsers.
And, also, I think that would be urgent that cPlay could read the embedded cue in the FLAC files. This would free the need to use cue files with flac, greatly simplifying the management of audio files. I think that this does not require much work: rather than read the content in an external CUE file, you can read the relevant section into the FLAC file. They have the same format.
Regards.
Ignacio.
I respectively disagree. I believe cics should do what he does best--seek out the next level of transparency.
How do you know when you have reached the final level.
OR is it a case of developing it until the sound no longer bears any resemblance to the initial recording.
To further complicate matters Vista is at the end of its cycle,7 is waiting in the wings,yet the software is still being developed for XP.
In its present state I fail to see much of a future,more so when there are other players that offer far more with comparable sound that will work on the latest platforms.
If the sole intention of cplay is to be a minority niche player then it has achieved it,but will not achieve much more.
Hi Flipper,
In its present state I fail to see much of a future,more so when there are other players that offer far more with comparable sound that will work on the latest platforms.
What players are you referring to that offer comparable sound?
You don't. Your comments reminds me of a quote I believe made at the beginning of the 20th century.....'all things to be discovered or invented...have been discovered or invented already...' or something like that. Well that did not pan out did it.
I will repeat my bottom line.
If the sole intention of cplay is to be a minority niche player then it has achieved it,but will not achieve much more.
With every new version it gets bigger,give it another year and it will be the same size as the software you now run down.
I use Foobar playing via Asio into an RME Digi96/8 PST then into an aria headphone amp feeding modified Sony CD3000 phones.
The detail/speed/resolution of this rig far exceeds some of your speaker setups.
For anyone to pick any variation in sound between the 2 players,they would indeed have to have Golden Ears.
It seems that those that consider themselves to be audiophiles are not happy to use a player that is mainstream,but will instead extol the virtues of a player where you can not even vary its size.
I am afraid that in its present configeration you audiophiles are welcome to it.
The reason others are not coming into the fold is because it has nothing to offer above others.
Complexity is not even worth a mention,if you consider it to be complex then you have never had much to do with PC's.
There are thousands of people like me that have been using PC's as there music source for years.If cplay had something to offer above the others it would be welcomed with open arms.
It has Great potential as a player,do not burden it down with a lot of settings that have no impact on the sound.
If these settings are having an impact on the sound,then they are altering the sound,you may as well use a SW equaliser.
Most of you would find more of a variation in sound quality by changing your cards around.
No one wants a player with bells and whistles,there are plenty of those as it is,but they do want a player that they can customise.
The Interface is everything as far as SW is concerned if you want it to be adopted by more than the few on this forum that use it.
Flipper,
If the sole intention of cplay is to be a minority niche player then it has achieved it,but will not achieve much more.
I am certain cics doesnt care. I dont think he wants it to be the next foobar or winamp. hell, he doesnt even seem to want any recognition for his behemoth of an effort.
With every new version it gets bigger,give it another year and it will be the same size as the software you now run down.
I use Foobar playing via Asio into an RME Digi96/8 PST then into an aria headphone amp feeding modified Sony CD3000 phones.
The detail/speed/resolution of this rig far exceeds some of your speaker setups.
For anyone to pick any variation in sound between the 2 players,they would indeed have to have Golden Ears."
Personally I doubt that cplay will ever get that big. Take a look at its growth:
![]()
Compare that to foobar:
![]()
Neither are that big but it does seem that foobar has grown more, and cplay comes with things included like asio upsampling, etc.
Maybe it is your headphones but I hear a difference primarily in the sound stage of Foobar, winamp and cplay, and that was a long time ago. Foobar and winamp were close and produced 2 d images and cplay was more 3d with more space between the images. I dont know, but it was pretty clear and I certainly back then was of the cover art gui crowd, but my ears told me that cplay sounded better.
It seems that those that consider themselves to be audiophiles are not happy to use a player that is mainstream,but will instead extol the virtues of a player where you can not even vary its size.
I am afraid that in its present configeration you audiophiles are welcome to it.
The reason others are not coming into the fold is because it has nothing to offer above others.
Complexity is not even worth a mention,if you consider it to be complex then you have never had much to do with PC's.
There are thousands of people like me that have been using PC's as there music source for years.If cplay had something to offer above the others it would be welcomed with open arms.
It has Great potential as a player,do not burden it down with a lot of settings that have no impact on the sound.
If these settings are having an impact on the sound,then they are altering the sound,you may as well use a SW equaliser.
Most of you would find more of a variation in sound quality by changing your cards around.
No one wants a player with bells and whistles,there are plenty of those as it is,but they do want a player that they can customise.
The Interface is everything as far as SW is concerned if you want it to be adopted by more than the few on this forum that use it.
All that you write is 100% true. But what you clearly are missing is that cplay was not designed to be a stand alone player!!! If you have only installed cplay and that is your only experience you will conclude as you did. However, cplay was designed to be used in conjunction with cmp and the principals in cics The Art of Building COmputer Transports.
In that context and with cmp, the interface is dramatically different. The file browsing window goes away and cmp is used to pick artists albums playlist, etc. It is a completely different experience.
If you open up cplay and move the mouse scroll wheel, it doesnt move from song to song, but with cmp it does. This is just one of many examples.
I thought the cplay interface sucked and sizing was just one of the many things wrong with it, but with cmp and with the principles of the whole build, the interface is great. There is no need to change it IMHO. Try using foobar or winamp or album player SOLELY with a remote. I certainly couldnt get it to work, but cmp2 totally works with the mouse as a remote and no need to move it around like a mouse....just the scroll wheel and the left click button.
Build a cmp2 and you will see for yourself, and it will be easier to hear differences. Heck just getting the extra psu for the hdd and usb is huge for sq.
And whats up with the audiophile hatred?
I agree with theob; cics should do what he does best with his product.
I use .WAV and FLAC users could write their own interfaces if they desired.
The pursuit of SQ should take precedence over GUI and other user refinements. Someone could step-up to the plate and assist cics in his endeavor.
RayBan
Hey Ray,
I agree with you and theob and I was once an eye candy guy who (LOL) wanted pretty pictures.
However I would love it if cics took a break and fixed the documentation or updated the cmp2 website.
I am a huge fanboy but the documentation is so out of date now that it really is hard to recommend cmp2 to anyone but the savviest computer person, and even then you have to keep checking for the latest settings.
Most probably dont care because they have their cmp2 rig running, but man there should be alot more people embracing this approach and the documentation stops them...
I struggled with that very fact.
I started with no computer skills, few tools but a love of music, so I ventured where most would not go.
A few kind souls, (Ryelands, seger, GStew, AlanJ and RickMc) provided great assistance and guidance. Daily (hourly) checking at AA was also of great help as the search feature allowed me to go back in time to see how others had solved their individual problems.
I agree we all owe cics a huge debt. I wish I had skills that could be useful to him but sadly I don't
I can say I started from scratch, crashed and burned as others have, got up and continued onward. I am certain dedicated others will join the challenge and do the same!!
RayBan
I also agree with theob that SQ is of prime importance and everything else is secondary. Unlike some users here I have no particular loyalty to cMP/cPlay but I do want the best and I sincerely appreciate all of cics hard work and dedication.
I thought that new documentation at the end of the year (December 2009) would be timely since Windows 7 is officially out in October. Actually cics should be asking for contributors instead of shouldering the entire burden himself.
With some volunteer help, it's looking much better and hopefully would be done in a month or so.
Awesome.
If you want I plan to build my rig this week and wouldnt mind some proofreading if you need it. I was going to do that anyway for guide I will be using.
Of course, SQ is the first. But, what is SQ?. Transparency?. I don't think so. I believe more in naturality and fidelity, then less distortion and tonal balance.
Resolution and tonal balance is the key. You prefer resolution but I think that the key is the balance between both. And this b29 is the most perfectly balanced of all versions.
My main source of music is live music: classical, jazz and opera. When I listen a live concert, sound is not so "microscopically" transparent as we expect from our equipments. This level of transparency is both unreal and unnatural.
RayBan say: "I use .WAV and FLAC users could write their own interfaces if they desired.".
Are you sure?. How?. It's impossible write an interface if we don't have a interface library.
In Spain, we have a proverb, it could be translate as: "The best is the enemy of the good". That say that an obsessive search for the best never allow us to distinguish the good things that have already achieved.
Of course, the title of my last message was a friendly nod. But I really believe that the SQ goal has been reached with this b29 version. Then, what I proposed to Cics was a reflection moment. I believe that is time to think about everything that has been done so far, the goals achieved and what is really needed to get that this great project might be useful to everyone.
Indeed, as mentioned above, a new documentation would be good. Also, this thread and the cMP thread are excessively long. A new thread with a compendium of the essential from this threads is also necessary. And, of course, the interface library to develop news users interfaces.
I am not criticizing anything, quite the contrary. I am only asking for think the way forward from now.
Regards.
Ignacio.
"naturality and fidelity, then less distortion and tonal balance": nice recipe for SQ.
"The best is the enemy of the good" - what is the original Spanish, if you please, it's a real gem. As you explained it, it is a perfect antidote to the obsessiveness of audiophilia!
"what is really needed to get that this great project might be useful to everyone" - and I think the essence of this would be more flexibility and choice for the user, in terms of both functionality and interface. You have the programming skills that could really help, so I hope you get a chance to put them to use on cMP and cPlay.
.
You are right!.
The original sentence is: "Le mieux est l'ennemi du bien". In spanish we usually say "Lo mejor es enemigo de lo bueno". This is a sentence so much used, and has become a proverb. I know that is also very used in Argentina.
Regards.
Ignacio.
Thanks for the Spanish - sounds best to me. Voltaire may have been the first to use it in print, in Italian 8 years before he used it in French ("Il meglio è l'inimico del bene"), but I wouldn't be surprised if it was already a popular expression that he picked up.
I agree the thread is quite long and a new one should be considered.
If one were to follow cics instructions attached with each new release, the documentation necessary is mostly there.
I used the search feature to locate info I needed (as well as kind assistance from Ryelands, GStew and others) to solve problems I had as a newbie.
RayBan
Does anyone know if the file browser in cPlay supposed to be functional when the computer is fully optimized? Mine goes into an hourglass and stays stuck in the File-Open window when I click on the file folder in the cPlay main screen.
Daniel Gauthier
Montreal Canada
Gigabyte GA-G31M-S2L, E7200, 2 x 1 gig HyperX LL Memory, Antec SLP-450WR PSU, WinXP sp3 on Compact Flash Drive, Jili@ Analog Out direct to Electrocompaniet Ampliwire II Amplifier, Apogée Convergence Speakers
Untick the "cMP" option in cPlay, untick both "Suspend" options in cMP (svchost and lsass) and disable cicsRemote. Then you can use the file browser in cPlay from within cMP.
There is a slight reduction in SQ because of the unused "Suspend" options. cicsRemote doesn't affect SQ, but it ruins most of the hotkey functionality of Windows and media players (apart from cPlay of course).
I'm not willing to compromise SQ so I guess that option is a no-go.
Daniel Gauthier
Montreal Canada
Gigabyte GA-G31M-S2L, E7200, 2 x 1 gig HyperX LL Memory, Antec SLP-450WR PSU, WinXP sp3 on Compact Flash Drive, Jili@ Analog Out direct to Electrocompaniet Ampliwire II Amplifier, Apogée Convergence Speakers
Yes if you are not in cMP mode. In cMP mode, it hangs.
There is an option in cPlay for 'cMP'. If you select that, file browse is disabled.
Were you in cMP mode or regular Windows when you tried to browse?
Greg in Mississippi
I was in cMP mode.
Daniel Gauthier
Montreal Canada
Gigabyte GA-G31M-S2L, E7200, 2 x 1 gig HyperX LL Memory, Antec SLP-450WR PSU, WinXP sp3 on Compact Flash Drive, Jili@ Analog Out direct to Electrocompaniet Ampliwire II Amplifier, Apogée Convergence Speakers
I have a 24/96 flac of about 10 minutes, for some reason as part
of the album cue sheet cplay can't load it, though as an individual file
it is loaded fine.
I can solve this problem by adding an index to the cue sheet to "cut"
the song, but, wouldn't it be better if this will be done automatically
by cplay (as it seems to be doing to the individual file)?
Oh, almost forgot, b29 is great, thanks for all this hard work.
Settings:
192k, SRC 145db, AWE, no timer
I have seen a few folks listing the PHASE they have SET their cPLAY "to"
and must make a comment.
PHASE is not a setting like the others, well, maybe some folks do change ALL for each recording and for those with such an intrepid nature you have my full respect. BUT, PHASE is different. What other setting can be selected from the remote mouse? PHASE is absolutely dependent on the recording you have selected.
Sometimes the PHASE is vague and either setting sounds the same as the other but there are many recordings, the truly good recordings will reveal a best setting. One does need to learn to listen for this but one will learn and hear the difference. It does make a difference.
This is not, ideally, a set and forget setting. For the best sound it must be USED.
I feel much better now.
Bye,
Rick McInnis
Yes its different per recording. I would think it also helps when a system component (DAC, amp or x-over) that inadvertently inverts phase in which case a phase "bias" tends to be preferred. I hope this makes sense.
it really does not matter how your components are handling phase or what is the resulting phase at the end of the chain, the recording determines the final best position for phase.
Most loudspeakers have phase confusion within themselves which some would say eliminates the need for phase switching. In my experience, there is always a best setting for most anything but Phil Spector type productions.
If you have not sensitized yourself to this it would not matter where you had the phase switch since it is going to change back and forth with individual recordings regardless of the selection of components for playback.
It is a worthwhile skill to learn. Don't think this is one of those A/B things. One must, at first, listen for awhile in one position and then try another, going back and forth. You will eventually hear the difference. By this, I do not mean you are going to "get it" the first listening session. I usually find my preferred position early into a recording but many times I decide I had it wrong and change back. And sometimes I do not find a best position at all. The vast majority of recordings do have a best position.
You will only experience anything approaching absolute phase by learning to hear it. Listening critically to voice is the best place to start. We hear voices in phase all day long.
Bye,
Rick McInnis
Hello, and thanks for clarification. I'll try today to find my best phase for the Jethro Tull wavs that I use to evaluate new versions of cplay. I know the songs' every tiny "bit", so, I hope, it would be relatively easy. Before I read Your post I thought it was some strange exotic add-on.
Serge.
That the cMP2 setup had a non-intrusive phase adjustment was one of the things that attracted me to it. Back when I played a lot of vinyl, I setup dual inputs on my phono stage, in-phase & out-of-phase. Each album got marked with a preferred setting (right alongside the preferred SRA setting... but that's another contraversy).
To learn how to 'hear' phase, I recommend starting with some recordings with single singer or instrument. I find voice and drums to be very easy to hear correct phasing on. Listen, flip the phase, continue to listen. You'll find a preferred setting. Then try more complex recordings... but be prepared to be frustrated. Many pop recordings are not phase-consistent... for example, the lead vocals will be in phase along with the drums, but the piano, guitars, bass, and backup vocals will be opposite. They seem to be better nowadays, but many of my 70's & 80's recordings are like this. And this is why learning to hear phasing works better with simple recordings.
This is a key parameter to getting it right in my book.
Greg in Mississippi
Hi, I just discovered this site by the friendly advice of another HiEnd guy from my city, and reading through the endless postings about cMP/cPlay I have become very curious, I have decided to try to build one myself in order to bring my mostly digital playback closer to true audio fidelity.
Here is my planned configuration:
1. Motherbord: Gigabyte GA-G31M-ES2L (bought it today new)
2. RAM: 256 MB DDR2 533 Infineon HYS64T32000HDL-37-A (is this one possible?)
3. HD: WD Scorpio Blue WD5000BEVT (SATA300), 500GB, 2,5"
4. cpu: Intel E7200, maybe boxed first in order to try the OEM fan first.
5. PSU: Antec Earth Watts 430W
6. ESI Juli@ or even my older LynxOne via S/PDIF to my external DAC.
Not yet sure about the following:
7. Case: Don't want to invest much before not being sure that the whole thing sounds really better than my old setup. So would it be possible and save to leave all "naked" first, I mean PSU + MoBo with CPU and RAM + HD outside any case? Or does it need to be "grounded" or however it is called (I do not know much about this stuff)? Otherwise before investing into a Zalman 160XT or such, I'd rather take a second hand HTPC case from Ebay for 50 or 70 bucks.
8. DVD drive: I haven't yet understood how to transfer all those Gigabytes of music to the cMP: Better with a DVD drive or better by USB2 or FireWire? I have read that some people disable both - but then how to connect to the cMP to the normal PC?
Of course, the most convenient would be to use a USB- Stick with 8GB or so. But I have understood that USB is better disabled...
All the other tweakings like more than one PSU, battery powered soundcard etc may follow once I hear that the effort starts to pay off in terms of SQ.
OK, if you see something that in your experience will not work well (hardware related), I'd be glad if you'd reply.
Thanks in advance,
Bernie, from Berlin
1-6 is fine. Just remember that initial install of Windows requires more RAM (1GB should be good). Only switch to 256MB RAM once Windows optimisations are done. Also, install Windows on a small partition (5GB will do) and the rest of your drive (495GB) becomes another partition where you store your music library.
7. No need for an expensive case, some use a wooden box that offers sufficient venting.
8. It's best to rip using EAC in secure mode on your home PC and transfer files to your cMP² via USB drive, memory stick or temporarily remove & connect drive to sata port.
Hello all,
It has been a long time since I’ve contributed here, but I’ve followed along and enjoyed most of your suggestions and the improvements made to cplay over time. Cics may remember me as suggesting process explorer, autoruns, and some random windows settings a couple years back.
I just installed minlogon for the first time (finally), and although the transition seems to have worked and the benefits to cplay are obvious, Zoom Player will not play my media files. All I can do is launch the program and select a file to open; and that is when the program hangs until I reboot. Does anyone else have zoom player functioning with minlogon? I use my CMP for watching movies just as much as listening to music and would like to be able to use only one Windows partition with minlogon for all media types. If I must switch to another application, what else are you using?
Thanks,
Mike
Hello Mike,
It's good to hear from you and congratulations on Minlogon.
With all my old parts I built a separate dedicated cMP HTPC with minlogon - zoomplayer works intermittently. I settled with PowerDVD. Works for my basic needs.
I have two power supplies I bought directly from granite digital a year ago, but I’m wondering if they are not the same one everyone else is using and whether I received the correct product. It only powers up one type of drive I own. As soon as I plug the sata end of the power connector to my Western Digital 3.5” 400GB hard drive, the drive starts spinning. I then turn the power on to the computer and everything is fine. I assume this is normal operation. My other drives buzz very faintly, but do not spin up. So I’ve been using my computers power supply to run a drive for all this time (a carnal sin by now)
Can someone answer why none of the following drives power will up:
Laptop drives:
Fujitsu MV2120BH 120GB Rating 5V - 0.6A
Fujitsu MHV2200BT 200GB 5V 0.55A
3.5” drives
Western Digital WD10EACS 1TB 5VDC – 0.7A, 12 V – 0.55A
Western Digital WD5000YS 500GB 5VDC – 0.7A, 12 V – 0.75A
The power supply I bought from Granite Digital says FINENESS POWER power supply and has some German writing on it as well along with these specs
INPUT 12V 1 AMP
AUSGANG 12V - 2000mA (Ausgang means exit)
Output 5V – 2000mA, and I thought this should be sufficient for any of these drives.
I still have the invoice and it reads:
Serial # 7130 Universal Drive Mech Power Supply $19.95 each
So, shouldn't these drives power up given their power requirements being lower than the power supply limit?
More importantly, did I receive the wrong product?
Thanks,
Mike
Thanks,
Minlogon wasn't really as hard as I feared, but I wasn't about to try it unless I had some backup software and a few extra drives to play with. Getting married and having a kid plus changing jobs just got in the way. Still can't update the hardware though.
I think I'm actually getting zoom player to work after all. I think the problems I am intermittently having are more to do with my backup program and zoom player (which I have to completely reconfigure..along with the windows settings). Every time I reboot I see some different things.
I going to post another question about the power supplies I bought direct from Granite digital (even though they say FINENESS POWER power supply on them)
They don't power up my laptop hard drives or my WD 3.5" drives except for one 400 GB drive.
I thought I read almost all of the cMP/cPlay posts but either I missed or forgot what I read.1. What is the difference between cMP and cMP squared (cMP2)?
2. I have cMP/cPlay running on two different computer that have extra memory in case my or your program needs change. I run the sse4b9 version on a Intel Duo Core 2 E7400 with 4 GB ram. I run the sse2 version on an AMD Athlon 64 x 2 Dual Core 4400+ with 2 GB ram. I don't know why I chose the sse2 version for the AMD and I'm not sure that the sse3b9, ssse3b9 or sse4b9 versions would not run. The Readme installation text mentions a CPU-Z program to determine which version will run but I don't recall seeing that program nor a clear explanation of what to sonics expect with each version.
3. Likewise for many of the other settings, it seems hard to choose settings without established or suggested recommended settings and what each setting is likely to bring to the table. Thus in your experience do you prefer the highest upsampling or no upsampling, a small or large buffer, AWE on or off.
I sometimes like to explore as much as the next guy but sometimes I like an established benchmark upon which to base all mods and tweaks. So can you give us your best and perhaps update the December 2007 document by the end of the year.
Edits: 08/08/09
I have a more or less comprehensive "check list" for setting up cMP2 systems (cMP2 = cMP squared, i.e. cMP and cPlay working together).
Some others on the list say they found it useful. If you want a copy, e-mail me and I'll send it.
Regards
Dave
Refer to ReadMe.txt for important details & specifications.
Change Log (2.0b29 release):
- DSP refinements
- Gapless processing enhancement
- Volume processing efficiency improvement
- DSP buffer optimisation (when using AWE)
- SRC efficiency improvement
- Other minor changes
- Minor ASIO refinement
- UI timer processing improvement
- Fix for Windows Kernel error (in memory management)
- Increased tracks/splits per cue sheet to 160 (from 120)
For best results, uncheck "Timer" setting and test different "Buffer" sizes. Details on SoX settings and output measurements can be found under Release 2.0b26 Notes .
Please REMOVE previous versions before installing cPlay 2.0b29. Normal (SSE2), SSE3 B9, SSSE3 B9 and SSE4 B9 versions are available. Running cPlay on a CPU not supporting the required instruction set will cause cPlay to exit immediately.
For cMP² users, review these BIOS changes (underclocks FSB, CPU & disable additional USB items). New settings are in effect allowing for lower power consumption (below 20W). This release allows for further CPU underclocking under maximum load (192k SRC 145db SNR output) where CPU Host frequency can be lowered to 140 giving CPU frequency of 840MHz! This setting is only recommended for WAV files (for FLAC, use a minimum of 150). Recommended Gigabyte mobo GA-G31M-S2L is no longer available, use either GA-G31M-S2C, GA-G31M-ES2L or GA-EG45M-UD2H ( advanced settings discussion ).
I have a question about the choice of mother board since I am going to build a dedicated music server per cics's recommendation. Do the boards GA-G31M-S2C, GA-G31M-ES2L have the same flexibility in BIOS configuration as GA-G31M-S2L? And if yes, do we need a special version of BIOS?
The GA-G31M-ES2L does not have the option to turn off Spread Spectrum,otherwise all the other settings are available. My testing suggests that option would be useful, but not critical... you can setup a good cMP2 with the ES2L, it is just a little better with the S2L and Spread Spectrum turned off.Greg in Mississippi
P.S. Don't know about the S2C.
Edits: 08/16/09
This is very helpful
Hi again,
I'm having some issues with b29 that I never had with previous versions (I was using b27).
The problem is strange and I don't know how it happes, but after a while of playing music and loading files I get this strange metallic sound. It's not just tics or pops, it's all over the place. I tried quitting cPlay and reloading the CD but it was still there. When I played with the settings (I switched from SRC145 @ 192 to SRC145 @ 44.1) the metallic was still there and really louder. Same problem switching to SoX.
The good news is a reboot solved the problem!
The first time I got the problem was the day I installed b29 (ssse4), the metallic appeared after 2 or 3 CDs. At the time I thought I had to reboot due to the new install so it didin't bother me. But then, a week or so passed, I played over 30 CDs, and this morning, it was back?!?
It's not a huge problem since rebooting seems to do the job but I wonder what might be causing this.
Anybody else had this issue?
Thanks
Etienne
Yes I guess you can say I've had it quite a bit with my Benchmark usb dac (which I connect via bnc cables to juli@). If you listen to 96 you should be ok. I've tried for 8 months to get rid of it and while I have ocassionaly gone a few weeks w/o it, it always comes back at rates of 176 or 192. I have just ordered a Buffalo 32 which I hope (when fed with a 96 khz stream or lower) will fix this.
Cics and others believe it may be an interaction between juli@ drivers and certain dacs or that it is just a juli@ problem. While it may very rarely go metallic with juli@ analogue outs it is easily fixed with a restart of the track (no reboot is necessary).
Hi Theob
I'm glad that we're two with the problem. Maybe together we'll be able to fix it!!
I really doubt it's the DAC because a reboot of the PC fixes the problem... it might have something to do with the RAM. Next time the metallics appear, i'll try to clear the RAM without rebooting the PC just to see if that helps!
By the way, this issue never occured with b27 which I had running since its release! Maybe Cisc can point out what were the major changes in the code between those two versions.
I too am using the BNC connection of the DAC1 (via a RCA to BNC cable (using the breakout cable with the Juli@)). Have you tried the Optical connection? What drivers revision are you using.
I'd really like to know if anybody else is using the Juli@ paired with a Benchmark DAC1 and experiencing the same issues. If others are using this combination without any problem then we'll be able to cross out the DAC, or even the Juli@ and we 'll be able to compare our configurations. I'm also curious to see if anybody else is using the digital out of the Juli@ at 192 paired with any DAC whatsoever.
Thanks
Etienne
Yes, I am using the Gstew modified Juli@ on a BNC to a Benchmark DAC 1 Pre at 192. Just installed 29 and ran it all night before my trip.
I'll run it again when I get home late tomorrow.
Got stuck with some counterfeit mem (256) which turned out to be only 155 so cPlay wouldn't play. Had to insert my Kingston 1GB XLL back
RayBan
Ok thanks for the info.
Could you please play as much music as possible, without rebooting the computer, and give us feedback next week?
Just let us know if you get this weird metallic sound.
Anybody else with this setup?
Thanks
Etienne
Have you tried the Optical connection?
Yes it doesn't fix it.
What drivers revision are you using.
1.05
I have tried everything: New hd (cpu, ps, juli@, hd's, mobo), reloading windows--all of which did not solve the issue.
I do know that I had less metallics with cplay 27 and no metallics with cplay 18. But since 96 src upsampling never gets metallics and since cplay 29 @ 96 sounds better than cplay 18 @ 192 I leave it at cplay 29 @ 96.
Hi Theob,
Today, the metallic sound came back again!
I really wanted to figure out why it happened and how to fix.
I tried clearing the RAM ... the metallic sound didn't go away.
I tried to exit cMP+cPlay and reload everyting ... the metallic sound didn't go away.
I tried to load the file via cPlay only ... the metallic sound didn't go away.
I went in DEVICE MANAGER, disabled the Juli@ Controler and enabled it ... the metallic DID go away!
I'm pretty sure it has absolutely nothing to do with the Benchmark DAC1, since disabling and enabling the Juli@ did the trick...
Maybe it's a driver problem with the Juli@...
Anyways, I'll do more research the next time the metallic occurs.
Thanks
Etienne
Very good analyses! I have just had it though and have already ordered the Buffalo. Good luck.
So did I but it's still not here yet!!!
RayBan
Hello,
Thanks a lot for this amazing release!
I really see an upgrade over the last version I used (b27)!
Plus, this release completely solved my "Not Enough Memory" problem. I had an issue with SP3 and 256MB of RAM. My GA-EG45M-UD2H has 34 MB of "Shared Memory" and there wasn't enough available memory to play a file via cMP/cPlay, my workaround was to kill ciscRemote everytime I loaded an album. But now, there's plenty of available memory and everything's working like a charm!
My settings are SRC 145db @ 192k.
Timer disabled
AWE enabled
Buffer size small
I'm still wondering what level I should set the volume control! Does it affect anything? I didin't hear any difference but I didin't do an exaustive testing. Should I set the level to 0 and control the volume via my DAC?
One last thing... What's the impact of each version on CPU load?? Should I use sse4 just because my CPU supports it?
Great Work!
Etienne
Thanks again!
cPlay's digital VC offers 53 bit resolution (64 bit double precision) throughout. I use this as my main VC over dCS' Scarlatti DAC (digital VC). In your case, I would test which VC to go with, but definitely avoid using both (i.e. one of them should be at full volume). If your DAC does digital VC, use cPlay instead.
CPU load definitely changes a little although same amount of MACs (Multiply Add Computations) are done. Best option is SSE4 when your CPU offers it as this employs a key CPU instruction in the most critical area of cPlay.
Shared memory (34MB) on mobo can be lowered - check your BIOS under video setting ("On-chip Frame Buffer Size") under "Advanced BIOS Features" (I think) and set to 8MB (note: do NOT use the lowest 1MB option).
Thanks for the info.
And i've looked everywhere, there's no "On-chip Frame Buffer Size" for the GA-EG45M-UD2H motherboard. At least not in the latest bios rev. F3.
Anyways, it doesn't really matter because I don't have the "Not Enough Memory" error anymore, even with 256MB of RAM.
Etienne
Curious... why not use the lowest 1MB option for the On-chip Frame Buffer Size?
TIA!
Greg in Mississippi
One of the best Version of cPlay. It feels as if I upgraded my amplifier.
Cannot decribe in words but music sounds as it should sound.
THank you Cics for this Jewel of a Audio Player
Firstly, thank you for the feedback - its excellent. B29 is a breakthrough (again)! In particular, violins and piano are rendered beautifully. There's great balance and emotional energy. I was hesitant to do this release given the many changes done/required. Anyway, this after b25 where I thought more improvements would be unlikely/slow...
I need to know what is your preferred settings: output rate, buffer size, resampler (for SoX other details as well), phase (0/180), awe & timer setting.
Well I've been playing with a "loaner" RME Fireface UC for the past month, so these are my (finally) definitive results based on that playback device.
-Output Rate: 192KHz
-Buffer Size: Small
-Resampler: SRC @ 145db
-Phase: 0
-AWE: Yes
-Timer: No
A couple notes about the RME Fireface UC. It is similar to a Fireface 400 only with a USB input (rather than a firewire input). The USB protocol is BULK mode (as is their firewire version) which is supposed to have similar benefits as asynchronous mode. And contrary to many beliefs, that USB is limited to 24/96, this does 24/192 no problem. With one caveat. I think the USB implementation is too CPU intensive, because I was not able to keep the FSB @ 140MHz with the CPU speed at 840MHz when playing upsampled 192 (with SRC @ 145). I could keep the CPU at 840 when playing upsampled 24/96 (with SRC), and could play 24/192 upsampled with SOX at this speed as well. However when playing 24/192 upsampled with SRC @ 145 I had to increase the FSB to 150 and the CPU at 1.2GHz. But playing 24/192 (with SRC @ 145) was so much better than anything else that it was worth it.
And "cics" did not ask this, but I have a couple additional "tweak" preferences to add: I prefer to set the SATA mode (in Integrated Peripherals section of BIOS) to "Enhanced" rather than "Non-Combined". This requires that you install Windows with the AHCI driver. And finally, I formatted my one HDD with two partitions. The OS partition is default, but the music partition is formatted with a cluster size of 64KB (rather than the default 4KB - and I also tested 16 and 32). Both these HDD "tweaks" result in a deeper soundstage with better separation and three-dimensionality. IME.
Cheers.
B30 is in the works and should see some CPU load reduction but I doubt it would help your USB setup.
I've tested 64kb block size but not with AHCI driver under "Enhanced" SATA mode - this is definitely worth trying!
Another very fine release. thank you!
My preferred settings are:
44.1
buffer large
SoX VHQ Phase-minimum, bandwith-97 alias off
Timer on
AWE off
Phase 0 volume 0.0
running Toshiba laptop with CMP, winxp with all the recommended mods
asio4all out to a peter daniel NOS DAC==> lightspeed attenuator ==> Peter Daniel Chip amp; no interconnects.. hardwired with silver and gold
I enjoyed version b28 so much and I thought that it is not going to be advanced any further in the near future. I was so wrong. Thank you so much Mr. cics for enhancing what is already a very fine player. b29 brings more detail and refinement comparing to b28 such that it is like seeing a better focused picture. Left to right balance is better and a wider sound stage. Better dynamic too. A significant version indeed.
My setting is :
Output -- 48K (the max acceptable by Devilsound DAC)
Buffer -- Small
Unsampler -- Sox, VHQ, Intermediate, 97, alias
No Timer nor AWS
btw, I have never been able to set AWS because I could not get lock privilege following the procedure in your document. I cannot find "User Rights Assignment" after hitting the 'add' button. Any indication why? I am using sp3 instead of sp2.
This setting can only be done under the "Professional" version or in Vista's case, "Ultimate" version.
Navigate step 7's tree hierarchy until "User Rights Assignment" is available in left pane of MMC editor.
Now I know why. I think its because I have the Home Edition of Windows XP. So the AWE option would have to wait until I replace my temporary server (running on an Aspire netbook) by something more substantial in a couple of months time.
Thanks.
It works like a charm.
Daniel Gauthier
Montreal Canada
Gigabyte GA-G31M-S2L, E7200, 2 x 1 gig HyperX LL Memory, Antec SLP-450WR PSU, WinXP sp3 on Compact Flash Drive, Jili@ Analog Out direct to Electrocompaniet Ampliwire II Amplifier, Apogée Convergence Speakers
- Output rate 44,1, SRC 145 db
- Buffer small
- Fase 0
- AWE On
- Timer OFF, also I often minimize the window during playback and switch off the monitor.
When I start CMP I have task manager running and as I don't use my Zalman now and have a monitor, I kill sicsremote and set cmp priority to realtime.
This version is very clean and the sound is getting lighter with the last 2 versions.
Thank You very much. Serge.
I need to know what is your preferred settings:
- Output rate 44,1 (Sabre ES9018 DAC seems to be insensitive for upsampling)
- Buffer large (more emotional)
- Resampler none
- Fase 0
- AWE OFF (More listening is needed for prefering 'on' with Cplay B29)
- Timer OFF
need to know what is your preferred settings:
output rate:
prefer 176 or 192 although I am very happy with 96. I tried going back to cplay 18 which never goes metallic (@ 192) but cplay 29 @ 96 is way better
buffer size:
prefer small altho could go with any of 3
resampler:
prefer src @ 146
phase (0/180):
prefer 0 but 180 works on some files
awe & timer setting:
prefer both on
I meant to comment yesterday, but...This version is just superb. As someone else said it is more airy and detailed and yet the music is more intense and affecting with imaging that is quite real. Congratulations, cics, and thanks.
Hi Cics,
Personally I like this version over v28 alot because (1) I feel less bass booming, and therefore more airy and details. (2) Emotionally when I listen to my favorite track and I can feel the mood of the singer much easier, that I had goose bumps when listening to the song. Congratulations.
Hi cics, thanks for this other very fine release.
Just one concern though: my "available ram" has decreased dramatically from previous version (b28).
I use one 512M stick and the maximum track size I can load dropped from about 9 minutes to 5 minutes for FLAC @ 44.1. and about 2 minutes for 24/96.
Is it due to a new way of treating data, and why you increased the track/split to 160 ?
Should I get a 1GB stick or is it just fine this way ?
Sonically speaking, b29, so far, is a new breakthrough.
Thanks
Martin
No need to go with 1GB RAM. Track split switching is done after 5 minutes playback. I would recommend going smaller to 256MB (split size is ~2:15) when media is largely 44.1k. This switching doesn't interfere with playback however flac does demand heavy cpu load and it may be a good idea to increase cpu speed (from low of 840MHz). Its better to increase from 6x to 7x instead of increasing cpu host frequency (leave this to lowest stable value ~140).
I use 256MB RAM.
A new memory model is used in b29 (to avoid some low level Windows kernel issues).
Thanks for the fast answer.
I'll try to get my hands on a 256 ram module, hard to find now, I fear.
My favorite settings for now: CMP2, 96k output(max. for my DAC), I hesitate between small and medium buffers, SRC@145dB, AWE enabled, Phase 0, no timer.
Phenomenal sonic balance, tight bass and smooth highs. And incredibly large soundstage, out of this world !
Is CPU speed X6 enough for 96K output, if I get no glitches ?
Superb work, once again, Mr. cics
Thanks again.
Martin
Is CPU speed X6 enough for 96K output, if I get no glitches ?
Yes, x7 would be needed for 176k or 192k output.
This is the first version of Cplay where i experience a benefit with AWE turned on. In previous versions AWE was not my favourite setting. AWE gives me now more precize sound, without getting edges.
Thanks Cics for your continously refining the code
I was able to get metallic free sound (@ 192/145/small buffer) from cplay 27 but I am not so fortunate with 28. Tried everything from slowing clock and lowering upsampling to 176. Even tried listening uniquely to music files off of flash drive. For whatever reason while listening to anything higher than 96 I get 40-50% cpu usage which always goes metallic. At 96 khz I get 25% cpu usage which is stable. 176 begets 40% cpu usage which varies + or - 4% (this is the most stable greater than 96 cpu usage I can get). Anything that can improve efficiency on future cplay releases would be appreciated.
For me as a new user of cPlay I´m interested in the related hardware. Which Intel CPU do you use? E7200, E7300 or E7400? Which soundcard?
Ruediger
All by the book...Intel 7400, juli@ soundcard plus all other recommended items. Only things I don't use are wireless mouse and touch screen hardware.
![]()
I´m new with cPlay and still a foobar user with a large music archive. The music has been ripped by EAC in flac format for single title without cue sheet. Is cPlay able to play flac files without cue sheet?
Ruediger
Ruediger asked:
Is cPlay able to play flac files without cue sheet?
Not individual files as shown in your screenshot but Al Jordan's Recursive Cuesheet Creator makes cPlay- and Foobar-compatible cuesheets for the largest of libraries in (literally) moments. See link.
Dave
Ingenious it works. The Recursive Cuesheet Creator is a fine tool. :)
Hi,
In my CPlayList Editor application, I have written some functionality to control CPlay across the network. Within my application, I create a run-time instance of CPlay.exe with the temporary playlist as an argument. This works fine.
As part of the handle to CPlay, I could send information to Cplay if it supported a standard input stream. It doesn't appear that CPlay supports an input stream. Instead, I've been able to control CPlay by creating an instance of Java's Robot class, which basically allows a person to send key events and mouse events to Windows. I've been able to add support for next track, previous track, fast forward, rewind, volume up and down, pause / play and phase by sending the appropriate key events to the CPlay application.
This only works when I am controlling CPlay over the network, because the Robot class only sends key strokes to the windows application that has focus. When I send a playlist to a remote machine, my application creates an instance of Cplay on the remote machine, and CPlay naturally has windows focus at that point. I can send whatever keystrokes I want to the application through the Robot class. Obviously, it doesn't work when running locally because my CPlayList Editor would have focus instead of CPlay. It doesn't really need to happen on a non-networked setup because CPlay has the same controls anyway, so it is a non-issue.
However, even across the network, I don't the Robot class is a very graceful manner of controlling CPlay. The reason being that if some other system event caused another window to appear and get focus, my application would then send the keystrokes to that window instead of CPlay. This hasn't happened yet, but I imagine it could. There is no manner in Java to get a list of running non-java windows application and have an application grab focus.
So, is there any other manner of getting information to a running instance of CPlay? It would be perfect if CPlay had standard input stream capabilities, as I could then send the same keys to the input stream instead of the inherent ugliness of using the Robot class. In this manner, I would know if my handle to CPlay still exists, and wouldn't have to worry about window focus or sending keystrokes to the wrong application.
Thanks,
Alan
In your case, I imagine it would be combination of FindWindow and SendMessage with WM_SETFOCUS.
I'll need to give this some thought on how best it can be achieved. cPlay uses direct keyboard input. Standard input stream may be the best way to do this.
In cMP mode with full optimisations and Minlogon, there's no chance for other dialogs grabbing focus (cMP hides until cPlay exits). Only exception is when an error occurs.
cics,
it would be good that all user interface's services (displaying and keyboard functions) were completely separated from the main engine of the application and to communicate them via a messaging API -as ussual in Unix aplications-.
This will facilitate us the development of new interfaces, without having to worry about the main engine.
Regards.
Ignacio.
Regards.
Ignacio.
Good idea. Skin is a separate component and a simpler version can be implemented as you described, i.e. a simple DSP message processor. This could be cPlay as a DLL version. Your thoughts?
I believe that cPlay must work as a service, getting his usual configuration paramaters from a configuration file and, while running, wait for messages from the user interfaces. A DLL must be a static component, not an active component.
The first step would be to fix the functions of cPlay and the interfaces and to design the messages protocol between both.
There are many ways for this messages exchange: you can use memory buffers, exchange registers, FIFO queues, etc, if you want to maintain the integrity of the design. But -and I know that you prefer not initiate the communications services-, I believe that best would be use a tradicional IP port. This standardizes the design and makes it independent of the architecture. In this case, for example, I could get a very simple PC exclusively dedicated and optimized to cMP/cPlay and the client interface in any other system: windows, linux, mac, etc, without consuming resources of cMP/cPlay. This, also, would facilitate the use of external storage and release cPlay the need to control it.
Even better: what if we released cPlay of the tasks of the DSP and it is limited to handle the audio you are serving, decompressed and upsampled, from an external computer?. I am absolutely sure that the resources and power consumption of communications services are much lower than the DSP are. You yourself has admitted that cPlay sounds worse when must to decompress the FLAC files. Let alone deal cPlay control audio and outsource all other tasks.
Now, I have got some time. I could help in the design development. Send me an email.
Regards.
Ignacio.
Regards.
Ignacio.
Unfortunately some win32 apis must be used for best performance in cPlay (although this is limited and can be ported without too much effort). My preference is to limit interaction to just the kernel. I'll meditate on a possible solution, perhaps just a simple .lib that you can use...
Refer to ReadMe.txt for important details & specifications.
Change Log (2.0b28 release):
- DSP refinements (reduce low level overheads & enhance threading)
- ASIO refinement (support for rare soundcard types removed)
For best results, uncheck "Timer" setting and test different "Buffer" sizes. Details on SoX settings and output measurements can be found under Release 2.0b26 Notes .
Please REMOVE previous versions before installing cPlay 2.0b28. Normal (SSE2), SSE3 B9, SSSE3 B9 and SSE4 B9 versions are available. Running cPlay on a CPU not supporting the required instruction set will cause cPlay to exit immediately.
For cMP² users, review these BIOS changes (underclocks FSB, CPU & disable additional USB items). New settings are in effect allowing for lower power consumption (below 20W). This release allows for further CPU underclocking under maximum load (192k SRC 145db SNR output) where CPU Host frequency can be lowered to 140 giving CPU frequency of 840MHz! This setting is only recommended for WAV files (for FLAC, use a minimum of 150). Recommended Gigabyte mobo GA-G31M-S2L is no longer available, use either GA-G31M-S2C, GA-G31M-ES2L or GA-EG45M-UD2H ( advanced settings discussion ).
PROBLEM FIXED: The WiFi showed up in the USB device list. I disabled the WiFi Radio and the problem went away. Sorry for the false alarm.
Cics,I am trying out 2.0b28_sse4b9 (same thing happens on other 3 versions) on a Dell M6400 laptop running 32-bit XP PRO SP3 with a new EMU 0404 via SPDIF.
Every time I select content to play in cplay via the SPDIF out, I get a error that the driver requested a reset and to restart the player. This does not happen if I select the analog outs.
If I put it into diagnostic mode, it will hang indefinitely on "Loading...". I can not kill the task and have to reboot to recover from it.
I have uninstalled and reinstalled both EMU and cplay.
Any ideas on what is causing the driver reset when using the SPDIF out ?
Does cPlay work on 32-bit XP Pro SP3 ?
I will get 1 or 2 seconds of garbled audio before it stops. Both the EMU window and the DEQX window show the rate changes are detected if I try different rates (e.g. 44.1, 48, 96).
I made the changes to allow locking of memory and have the buffer size set to auto.
M6400 Laptop -> USB -> EMU 0404 -> SPDIF -> DEQX HDP 3.0 -> amps -> speakers.
2.53GHz QuadCore Intel Core 2 Extreme QX9300
Dual Boot: 4GB - 32-bit XP Pro; 12GB - 64-bit XP Pro
RAID0 (2 x 320GB drives)
MMX, SSE, SSE2, SSE3, SSSE3, SSE4.1, EM64T
Thanks,
Tim
Edits: 07/25/09 07/25/09 07/26/09 07/26/09 07/26/09
There are cPlay users on XP SP3 and have not had similar problems.
The fact that it works on EMU's analogue outs is very strange and seems to suggest a problem with EMU's driver. Do you have the latest drivers installed?
Cics,
"The fact that it works on EMU's analogue outs is very strange and seems to suggest a problem with EMU's driver. Do you have the latest drivers installed?"
I was only able to get it to playback without throwing the error when it was set to analog outs and may have had other configurations incorrect. This observation may have been a red-herring because I did not have the TRS/XLR analog cables at that time so I could not hear anything from the stereo rack. I just saw that player count up with no external audio. I would dismiss this observation.
I believe I have the latest drivers. I am using the drivers that came with the EMU install discs. I did not find any newer drivers, but I may not have looked in the proper places.
I spent a day trying different software and driver permutations. I had NetJack2 installed and then de-installed that. I started blowing away processes and still no go. I tried ASIO4All and that didn't work. I tried other players including Foobar2000 and J. River Media Center. J. River would randomly output clear content, garbled content, and silence. The next day I went out and bought a pair of TRS to Male XLR interconnects from Sam Ash Music Supply to try the analog outs. Still no go with the analog outs going into my pre-amp.
After all of this, I started looking at the USB entries in the device manager (versus the audio drivers). When I noticed WiFi was listed as a USB entry, I disabled the WiFi radio from the tool-tray and everything started to work (CPlay, Foobar2000 and J. River Media Center).
BTW, I am comparing a Mark Levinson #390S CDP XLR digital out against the EMU 0404 SPDIF digital outs going straight into a DEQX HPD 3.0. So far the ML #390S still has the edge, but I haven't tried doing any PC mods, EMU power supply mods or fancy USB cable mods yet. It works so well straight off the bat (after disabling the WiFi radio) that I will be building a dedicated PC around it. My goal is to replace the ML #390s with a better sounding network player.
Is it possible to launch CPlay in realtime such that it will play 1 song and then quit? When I use the .bat file, it does not exit after playing. I would like to make a light weight network command interface to feed it a song at a time and possibly support other controls so a tablet or laptop could remotely control it without the graphics overhead of VNC or other remote desktop software. This would also offload the database searches, artwork, artist, etc. displays to the laptop or tablet PC. The only thing that would go to the playback machine would be the cplay command with the filename (and possibly other exposed commands).
Thanks,
Tim
Is it possible to launch CPlay in realtime such that it will play 1 song and then quit? When I use the .bat file, it does not exit after playing. I would like to make a light weight network command interface to feed it a song at a time and possibly support other controls so a tablet or laptop could remotely control it without the graphics overhead of VNC or other remote desktop software.
"...BTW - I had the ML390S directly into my ML33H's..."
Kindred ML Spirits.
Has your cMP / CPlay implementation bested you ML390S yet?
My amps are a ML336 for bass and Jeffrey Rowland Model 10 for (mid+highs). Formerly with the DEQX, Magnepan 3.6's and Martin Logan Descent, I had 11Hz to 20KHz +/- 5dB in room response. I have since swapped out the 3.6's for 20.1's and don't have the new numbers dialed in yet.
Tim
Hi cics
I am just starting to try out your cPlayer/cMP, it looks a great concept, thank you.
I am trying to get cPlayer working first. I have AMD 4850e (SSE3, confirmed by CPU-Z). However, neither of the following versions will launch;
cPlay_2_0b28_sse3b9_setup.exe
cPlay_2_0b28_ssse3b9_setup.exe
The following do run fine;
cPlay_2_0b28_sse2_setup.exe
cPlay_1_4_Final_SSSE3_setup.exe
Do you have any tips. Otherwise, which of the last two would you suggest I use?
Thanks
Oli
System;
Windows XP Home (SP3)
Gigabyte GA-MA78GM-S2H rev1 (latest BIOS & chipset drivers installed)
AMD Athlon™ X2 Dual-Core 4850e
Trust 5250 (CMIDriver-1.2.5-bin-x86)
ASIO4ALL
I assume you're able to install the "SSE3B9" version but cPlay doesn't start.
This may be an issue with AMD 4850e (very unlikely though) as cPlay SSE3 is fully tested on a Pentium 4 SSE3 capable processor. What I suspect is the "B9" feature that's affecting your system. Any chance of using XP SP2 or SP3 Pro?
cics
I tried rolling back to SP2 - same problem. Yes, I can install SSE3B9, but it will not run.
Thanks
Oli
I have found that only SSE2 versions of cPlay will run on three AMD AM2 processors. I have tried SSE3 without success on a 4000+, 5000+ black, and 5400+ processors.
BTW - I find that the latest versions of cPlay with SoX VHQ Intermediate Phase seem to sound best in my two all digital systems using Juli@ soundcards and s/pdif output.
Jack
Hi Cics,
Whow this version 2.0b28 really, really swings !
Very enjoyable…..
Every previous version added (more or less) better highs, mids, lows, 3D, ect.
But this version…whow….
This version ads the last missing ingredient: PRAT !
Pace, rhythm and timing.
Although i’m not playing any musical instruments myself, i’m blessed too be surrounded by many friends who do/can and some music-producers. They all have one simple rule to judge sound quality: “do you want to turn it louder or do you want to turn it down ?”
Well….. with this version I want too turn the volume up !
The more volume, the more it swings.
And it doesn’t get aggressive or harsh… or boomy.. or …..
The seas tweeters in my Klein + Hummels really like the highs of this version
The mids are beautifull two.. Saxofoon ect sound wunderfull !
Cics, I must thank your daughter for messing up your CD-collection.
Otherwise you wouldn’t have produced such a nice player. .
Thank you for this version.
LynxL22 dig i/o XLR AES/EBU -> Lavry Black DA10 -> XLR Mogami Gold -> Klein & Hummel O300
I have been spending the entire weekend working on the cMP2 + cPlay system and as a result have locked up the G31M-S2L quite a few times with just about every combination of settings.I know have a 2 inch thick binder with all the documentation and all the relevant posts littered with post-its!
Here are the settings that seem to work best to keep a stable system:
CPU Clock Ratio: x7.0
Fine CPU Clock Ratio: +0.0
CPU Host Frequency: 175 mhz
SPD: 2.0
FSB DeOverVoltage: -0.15
CPU GTLREF Voltage Ratio: 0.566
CPU Voltage Control 0.875CPU-Z reports the following:
Core Speed: 1225 mhz
Multiplier: x7.0
Bus Speed: 175 mhz
Rated FSB: 700 mhz
Core Voltage: 0.864v
Memory Timings: 3-3-3-8 (Set to Auto in BIOS)
CPU Temperature: 48 deg. C.I have done almost all the Windows optimizations (including AWE) except for Minlogon (For which I am waiting to have finished with Windows changes).
AWE configuration instructions are not included with the 2.X versions of cPlay postings. Is it still used? cPlay diagnostics shows it is successful.
For some strange reason I can't set the CPU Clock Ratio to 6, the board refuses to go lower than 7, it locks up and resets the CPU Host Frequency to Auto.
I'm currently using 2 sticks of 512 meg Kingston Value RAM and the memory seems to be the most touchy part of the system. Setting the Fine CPU Clock Ratio to +0.15 and the CPU Host Frequency to 170 mhz resulted in a BSOD related to RAM. Next thing to try is the Hyper-X of which I have 2 sticks of 1 gigs each. I need at least 1 gig of RAM as I listen to a lot of classical music with tracks of more than 30 minutes. I wonder which is the best, 1 stick of 1 gig or 2 sticks of 512 mb (to enable the dual channel)?
I also tried to lower the PCI-Express Bus Frequency to 90 mhz and that caused instabilities and a BSOD in Windows. Is this setting still recommended?
As far as the sound quality is concerned it's amazing. The Vangelis Mythodea album sounds superb, very open, and I hear many details I'd never heard before. The voices sound so natural and the bass response is very strong but not muddled. If anything this version highlights my slightly too bright tweeters, but the sound is never aggressive.
Daniel Gauthier
Montreal Canada
Gigabyte GA-G31M-S2L, E7200, 2 x 1 gig HyperX LL Memory, Antec SLP-450WR PSU, WinXP sp3 on Compact Flash Drive, Jili@ Analog Out direct to Electrocompaniet Ampliwire II Amplifier, Apogée Convergence Speakers
Edits: 07/12/09 07/12/09 07/12/09
Your experiences suggest EIST is enabled in BIOS. It's best to use just one RAM module (either 512 or 256). Are you using the F6x BIOS?
PCI Express should be left to its default value.
After checking and changing for above, leave VID at 0.875V and reduce CPU Clock Ratio to 6x. If stable then look to reducing CPU Host Frequency followed by CPU voltage VID. All else looks OK.
I've only been able to get back to the player a few days ago. I replaced the RAM with the Kingston Hyper-X 1 gig LL module and today had the time to try different settings. I've finally got to where I should be! The RAM made all the difference in being able to lower the clock speed and the lowest voltage I can get to is 0.775 volts. Any lower and the board refuses to start. I am finally satisfied.
CPU-Z now reports:
Core Speed: 900 mhz
Multiplier: x6.0
Bus Speed: 150 mhz
Rated FSB: 600 mhz
Core Voltage: 0.752v
Memory Timings: 3-3-3-7 (Set to Auto in BIOS)
CPU Temperature: 38 deg. C. (10 deg cooler!)
Now all that's left is the min-logon which I will try after Ghosting the compact flash.
I've also installed an external RAID box that sits in the next room connected by an eSATA cable. The box is accessed from my other PC via USB and has 1.5 terrabytes online. I now have access to all my music on the player and it sounds absolutely wonderful.
Daniel Gauthier
Montreal Canada
Gigabyte GA-G31M-S2L, E7200, 2 x 1 gig HyperX LL Memory, Antec SLP-450WR PSU, WinXP sp3 on Compact Flash Drive, Jili@ Analog Out direct to Electrocompaniet Ampliwire II Amplifier, Apogée Convergence Speakers
Reduction in temperature is a very good indicator of lower power consumption. You're very close to the optimal setup.
Use Seger's method (for both SP2 or SP3) to implement Minlogon. You should also reduce RAM size to 512MB.
Hello cics.
EIST is disabled and I am using the F6x BIOS. I think the problem preventing me from lowering the CPU clock is the RAM. I will replace it later today and see what happens.
Daniel Gauthier
Montreal Canada
Gigabyte GA-G31M-S2L, E7200, 2 x 1 gig HyperX LL Memory, Antec SLP-450WR PSU, WinXP sp3 on Compact Flash Drive, Jili@ Analog Out direct to Electrocompaniet Ampliwire II Amplifier, Apogée Convergence Speakers
Has there really been a recommendation to do this?
From cics or from a pasticpant?
Took me by surprise.
I would think it futile trying to slow your machine down until all optimizations are done, most assuredly including MINLOGON, and even then I have found one needs to "ease" into the lowest possible settings, especially CPU voltage.
Bye,
Rick McInnis
I actually don't remember where I saw the PCI Express thing. It was written on a post-it stuck in my printouts of the important threads.
All optimizations except Minlogon are done, I'm going to do a full backup before trying that to ensure I don't trash my Windows install.
Daniel Gauthier
Montreal Canada
Gigabyte GA-G31M-S2L, E7200, 2 x 1 gig HyperX LL Memory, Antec SLP-450WR PSU, WinXP sp3 on Compact Flash Drive, Jili@ Analog Out direct to Electrocompaniet Ampliwire II Amplifier, Apogée Convergence Speakers
CICS.. Once again thank you!! First I did not expect another release, and second I thought it would be minor fine tuning.
This release is a significant enhancement. On my system, SOX pulled out way ahead of SRC. I have yet to experiment with all the buffer sizes. I am running without any up sampling feeding a good quality NOS DAC from a notebook computer driving CMP. The clarity of sound improved while musicality preserved. The sound stage on SOX took a step improved in 3d. I am hearing details in music never before heard. In my opinion on my system SRC sounds like it is processing the sound as you hear with up sampling where SOX is giving full transparency. In any case it is excellent and a surprise gift late last night when I realized it was available and plugged it in.
The bass less muddy and the little haze is gone. More space in the sound stage. May be that's what HP called transparency.
Hello, Cics and everybody! I wonder at your silence! This is a wonderfull new version, indeed. The lower region got so clear and dimensional, the sense of pace and rhythm is so pronounced! I think it is a great achievement overall, and it sounds so clean and HEALTHY. I stopped using SOX.
Serge.
More detail with more musicality! Micro dynamics and vocals are wonderful.
Are you using SRC@145db?
Its a long time ago that I posted something. I am very busy in my job but try to keep up with the development of what is happening here. Also I have changed my hifi system with the help of alfred (sonics) completely. Only the speakers survived. This morning I installed the new player and was imeadetly hooked by new details and natural sq. For me especially voices and piano, but also drums and bass have a new quality I didnt expect to be possible. Thank you cics. Like several times before you made my day. Walter
Sonics (i.e., Alfred) is one of those very knowledgable yet sharing kind of gems on Audio Asylum that makes this such a great place. I too am a student of his.
cics wrote:
"More detail with more musicality! Micro dynamics and vocals are wonderful."
Well, here's an "unsubstantiated assertion" to get 'em going.
You're right.
Running v 28 into an as yet only partly-configured "headless" Fit-PC2 and a NOS DAC, I don't think I've ever heard music reproduction like it, at least not in my home.
As you suggest, unaccompanied choral music, which often on disc can be somewhat flat and lifeless, is sounding uncannily real.
I have no idea how my setup compares to high-end this or mega-buck that (and don't much care) - what I'm hearing is very good and that'll do me.
I haven't commented on a cPlay upgrade for months but this time at least you are definitely to be congratulated.
Dave
Yes! I tried different sox settings, but somehow they all seem to be compromising one aspect in order to push out another, that leads initially to a WOW, but soon to a "don't believe" condition. SRC is more natural to my ears.
Yes, You are right about microdynamics and musicality, every minute detail is here and has a right feeling and shape. Even my incorrigible Robert Plant - Mighty Rearranger started sounding more smooth. Interesting that Your player is Not smooth, but it brings so many details that such terms as sharp, harsh, smooth are not relevant any more. I am not trying to flatter You, I have a vast experience with very smooth and liquid players - accuphase 65, 67 and 500. Well, They are smooth and tame and very "nice" sounding but at the expense of the hf details. Your player sounds smooth when musicians want it, and, I hope, it says all. Thank You for this constant pleasure.
Serge.
It is going to be a fine weekend!
Edits: 07/11/09
Thanks yet again. I'll say the same thing about this release that I've felt about every release. It is yet again another increment in improved clarity, diminished smearing, and improved 3-dimensionality.
Initially I love 28 sonically--sounds more transparent top to bottom esprcially top. However the last couple of days I have been back into metallics. When I run src/146/192 I get 50% + 7% cpu usage and ultimately the annoying ringing. I get it even with sox. Anyway I notice that in Task Manager performance I get a red plot showing up. When it surfaces (running 96 src) the total % cpu usage shoots up 7% to 33% (or way over 50% for 192) . I looked up Performance Help description in windows and it indicates the red can reflect amount of cpu consumed in kernel operations. So I assume this is what I'm seeing. However when I ran for 2 weeks perfectly I got 25% cpu usage with src/146/192 now I get 50% + 7% which eventually goes metallic (I don't remember getting that much % usage variation). So question what do others get for cpu usage with cplay full out? I see this as a major indicator of how stable my pc runs. If I can get back to a steady 25% usage I feel confident I will run good. Any ideas?
Very nice except for the garish interface,the colour's and buttons on the previous version look a lot cleaner than the new ones.
I run it on vista[32" monitor and it looks crap].
Hey Flip,
I agree with you about the looks. But none of the pretty players sound anywhere near as good as cplay on my rig.
having stated that I thought one could get a pretty interface and great sound, I have been sorely disappointed.
Cplay is now what I use and at aleast on my 17" monitor cplay looks fine.
Also if you do the video optimizations ONLY cplay will look good on your screen....
Have switched back to b25,that great big ugly red button is to much for me to handle.For a player to be great it has to look good as well as play good.
When I want a bit of eye candy I switch to Foobar.
Hey Flip,
Keep in mind that Cplay arose due to inadequacies in other players like Foobar. It was designed to fit into the memory player concept where OS bloat and many other things are treated.
If you do the necessary video optimizations to maximize sound (like 8bit color) no player is going to look good. I had winamp with pretty cover art and well it looks like crap in 8bit color.
I have to hand it to cics for staying with his audiophile sound first principles even at the expense of looks....
Hey, my dear beloved wife chose those beauties!
I dare not change them for a while unless I'm willing to loose some liberties (like implementing more refinements... b29).
Please do not get me wrong,I love the program.
There are many who use it on there main pc's without all the os tweaks.
If you want to see what it looks like run it on Vista on a large screen.
I have just had another look and I will stick to b25,maybe some time in the future there will be an option to resize those buttons
No offense taken whatsoever.
Each cPlay release is tested on a Vista machine - the 'spartan' look is much better than in XP. These machines have normal 32 bit color resolution. What res are you using?
Running 1920x1080 on 32",like I said b25 looks great,but would look a lot better if I could change the size of the task bar/text.
I do realise that some people operate it from a distance so they need large buttons/text etc.
I have it open alongside Foobar to compare and the text and buttons are about 3 times larger.
Well, because of the cheap USB device I'm playing with at the moment is limited to 24/96, that's what I'm running at, so my information may not be entirely helpful to compare to your 24/192. But maybe it is at least some indicator that might help you.
I just listened to an entire album and the CPU graph was pretty consistent throughout. It only spiked an additional 5-10% (on CPU 1) when loading the next track, but this seemed to be dependent on the size of the track. In the Task Manager, under View, I enabled "Show Kernel Times" that way I could see the red graph as well as the green.
So running at 24/96 with SOX set to VHQ, Minimum Phase, 99 Bandwidth and Alias on (and CPU at 840MHz):
The overall CPU usage bounced between 10-15%
But looking at the graph for each CPU: CPU 1 Green was 14-22% and Red was 1-6%
CPU 2 Green was 3-8% and Red was 2-7%
I hope this information is useful to you.
thanks it tells me my cpu is laboring maybe unneccesarily
just as mysteriously as it started 2 days it has now just ended---I'm back @ 25% for a full out cplay.
How about a link to the source code? I am learning to code ASIO using C++. This source coce would make a great tutorial.
I second that. This project made me very interested in ASIO coding.
.
cPlay 1.0 Final has been packaged (includes instructions, precompiled libraries and DLL) and is easier for learning ASIO.
cPlay 2.0 will be packaged later.
Could I solicit your thoughts and experience in connection with passive cooling for the E7300 CPU on a Gigabyte Gigabyte GA-G31M-S2L mother board. The Thermalright SI-128 originally suggested by Cics is discontinued. I'd welcome some recommendations for other passive coolers which have been found to work okay.
Thermalright AXP-140 Low Profile 6-Heatpipe Aluminum & Copper CPU Heatsink
RayBan
Thanks for this suggestion.
I just realised that I stupidly put my original post in the cPlay string when I ought to have put it in cMP. I've posted it again in the correct string and I apologise to all for this dual post.
Anybody using cPlay with an Empirical Audio Off Ramp 3?
I loaded the latest version of cPlay and it sees the ESI Julia and ASIO4ALL in my music server but not the Off Ramp on USB. I'm certain the Off Ramp is recognized in the system (Foobar playing). I tried changing USB ports to see if reassigning it would do something and it didn't help.
A solution would be greatly appreciated. I'm very much looking forward to getting cPlay working.
Thanks.
You guys are using asio4all with the offramp? I thought Steve had his own drivers and didn't require 3rd party. My Offramp turbo 2 uses the mtransit drivers, no need for asio4all (which I don't think sounds so good).
The issue is getting the Offramp to work with cPlay. It must go via an asio driver to work with cPlay, hence the use of asio4all. Steve supplies directsound2 for use with Foobar 0.8.3 and Offramp, but this will not work with cPlay. Is the mtransit driver that works an asio driver?
The Maudio Transit driver only works with the older Off-Ramp units. The new units use Centrance USB interface which is plug-and-play with Windows drivers. Same as the Benchmark DAC-1, Belcanto 24/96 and PS Audio.
If this does not work with Off-Ramp, It will not work with these others either. It think the problem is ASIO.
steve N.
Yes, no asio4all on my setup. Cplay sees the unit an an m audio transit. FWIW, I think cplay sounds superior to the Nugent recommended foobar 0.8.3/directsound 2.
Edits: 07/10/09
I used earlier versions of cPlay (those up to early spring) with offramp 2 via asio4all. cPlay sees asio4all(not offramp) and asio4all sees the offramp. If you get this and it still doesn't play, try doing a clean reinstall of asio4all.
Got it working. I uninstalled ASIO4ALL and reinstalled it with the offline dialog, found the Off Ramp there, then cPlay/ASIO4ALL showed the options for the Off Ramp.
My first install of ASIO4ALL excluded the offline dialog. cPlay was seeing only the association to Julia.
Thanks!
For anyone considering a linear PSU + picoPSU
In the process of optimizing the power for the P24 connection, I did some current measurements on the 12 Volt DC powerline that feed the 12V Micro Power Supply from Mini-Box. Model: PW-200-M (200 Watt).
In some posts current draws on the 5 V lines of the P24 connector off over 4 Amp’s are reported (On a cMP optimized computer transport). So just to be sure that enough power can be delivered, I bought an 200 Watt version (PW-200-M) and also an linear bench/lab PSU capable of delivering up to 6 Amps at 12 V.
After measuring the current into the 12v Micro PSU the whole seams a little over dimensioned.
During boot-up the current to the 12V Micro Power Supply reads between 1.6 to 1.95 Amps with one short peak on 2.1 Amp.
During playback (or idle) it's 1,28 Amp. At a constant value.
This value is only a little influenced by bus speed, monitor resolution / color settings, type of sound-card. Ect. The LynxL22 draws 0,1 amp more than the Juli@ (without the analog part)
So in my cMP-setup the P24 pin uses/consumes:
- At start-up: 1.6 – 1.9 Amps. 1 peak at 2.1 Amps. x 12 Volt is about: 20 - 25 Watt’s
- when playing / idle: 1.28 - 1,35 Amps. x 12 Volt is about 15 – 16 Watt’s
If anyone considers buying a picoPSU for the P24 connector:
the smallest picoPSU version of 60 Watt, should be enough (for a cMP optimized PC).
Also a small linear bench / lap PSU (with low ripple specs) capable of delivering 2,5 Amp’s should be enough.
PSU-wise my cMP setup now looks like this:
- P4 pin: 1 linear bench / lab PSU output: 12V DC / 2,5 Amp max
- P24 pin: 1 linear bench / lab PSU output: 12V DC/ 6 Amp max -> Mini-Box PW-200-M. (with smoothing caps on all 3 voltages (3,3/5/12)
- USB, HDD, DVD, etc: ANTEC AW 430
Pin 16 (PS_ON) on the ANTEC P24 connector is grounded to pin 17 to start the Antec. The P24 pin isn’t used to feed the MOBO P24 connector. (The picoPSU now does that). Only the ANTEC’s molex connecters (5V + 12V) are used to feed USB, HDD and DVD as explained in Cic’s AOB manual)
Sound Quality <-> costs
ANTECH AW 430 PSU costs about 47 euro (in The Netherlands)
A picoPSU costs 37 euro + 30 euro for linear bench / lab PSU = 67 euro (in The Netherlands)
IMHO the extra 20 euro’s are definitely worth the extra costs.
I hope the placebo-effect is not fooling me but I think the picoPSU + Linear PSU sounds better than the ANTEC AW 430. Not a gigantic difference, just a little better. More space, better imaging, better treble , etc. Everything a little better.
And…… as with the power optimization for the P4 connector: its very, very easy to implement. Anyone can do it in 15 - 30 minutes.
Are there any other inmates that can comment on SQ-improvements when using a picoPSU (versus a 'normal' swithing ATX PSU) ?
Mark
N.B. How the currents measurements were done:
1 linear bench / lab PSU -> multi-meter -> Mini-Box PW-200-M.
cMP2 optimized as per Cics recommendations: all BIOS optimizations done (no EIST function) + all software + min logon. GigaB G31M-S2L, E7300, Core Speed: 840 MHz (140x6), FSB 560 MHz, 1Gb Kingston HyperX, at 3-3-3-5, FSB not undervolted, CPU GTLREF ratio at 0.65, CPU Voltage at 1.0000V (just enough for allowing passive cooling)
LynxL22 dig i/o XLR AES/EBU -> Lavry Black DA10 -> XLR Mogami Gold -> Klein & Hummel O300
Yes, I'm using the Pico with a Daniels inspired Linear PS built into the case along with a Hammond 193L inside also.
I added AR power cables and a GStew modded Juli@ digital at the same time so the sudden increase in SQ was difficult to attribute to any single change.
No fans, no switching PS, runs cool and sounds great.
Ryelands inspired me to order Element twisted pair speaker cables and interconnects as my arthritis prevents me from making my own as he has so capably done.
RayBan
Refer to ReadMe.txt for important details & specifications.
Change Log (2.0b27 release):
- Minor ASIO refinements
- DSP refinements (new advanced design implemented) giving efficiency improvements
- Disable keyboard processing when file browsing
- Faster 24 bit WAV processing (improving RAM load times)
- Compiler optimisation done (for FLAC)
Details on SoX settings and output measurements can be found under Release 2.0b26 Notes .
Please REMOVE previous versions before installing cPlay 2.0b27. Normal (SSE2), SSE3 B9, SSSE3 B9 and SSE4 B9 versions are available. Running cPlay on a CPU not supporting the required instruction set will cause cPlay to exit immediately.
For cMP² users, review these BIOS changes (underclocks FSB, CPU & disable additional USB items). New settings are in effect allowing for lower power consumption (below 20W). This release allows for further CPU underclocking under maximum load (192k SRC 145db SNR output) where CPU Host frequency can be lowered to 140 giving CPU frequency of 840MHz! This setting is only recommended for WAV files (for FLAC, use a minimum of 150). Recommended Gigabyte mobo GA-G31M-S2L is no longer available, use either GA-G31M-S2C, GA-G31M-ES2L or GA-EG45M-UD2H ( advanced settings discussion ).
Hi,
I have a little suggestion related to audio file loading using cuesheet.
Would it be possible that cplay be modified in the following way:
- If cue file mention FILE "foo.wav" WAVE
-- try to load foo.wav from current directory and if it fails fallback to the loading of foo.flac and if it fail display an error.
This could be usefull for people using flac as modification of cue files generated from EAC or other software would be necessary anymore.
If you don't have time for this, I would be interested in sending you a patch for that feature if you can send me the source code to pbmtp AT free DOT fr
Thanks for this software
What will be most useful is a playlist tool. I'd like to remove the "File Browse" button and replace it with a more general playlist tool button. Alan has built a standalone cue sheet creator / playlist tool in Java.
Idea is when you hit this button, an elegant playlist tool takes the place of the old file browser. One can either select a file / track or build a new playlist and send it to cPlay for playback.
This will need to be built in c/c++ using either WTL 8 or native WIN32 api (preferred). Also, avoid a resource editor.
I ripped the whole cd into one large image FLAC file + a CUE file using EAC. With the cplay "File Browser", I just load the CUE file and play the music. That works very well with the existing cplay file browser.
If we move to "playlist tool" instead of file browser, is it possible to create play list selected tracks from these CD images base on the information of image CUE file? That would be ideal. I don't want to rip all music again into "one file per track".
s
YES,
You CAN create a cue sheet playlist. Alan Jordan who posts on the computer asylum created such a program. I am pressed for time, but doa search for somthing like "cue sheet playlist creator" or "recursive" and look for posts from his moniker "aljordan".
It is a fantastic program.
using cuesheets? Playing the WAV file, playing the FLAC file or displaying an error.
Rip one of your cd with EAC in flac mode, ask EAC to create a .cue file.
Once it is done load the cue file with cplay and you will get an error as inside cue files have .wav extension whereas on disk they have .flac extension.
To fix you need to open the cue file with a text editor and replace all occurence to .wav with .flac.
My idea of trying to load .wav file and if it fails trying to load .flac file would fix such issue without the need of editing all of your .cue file
.
I had upgraded cPlay from b15 to b27 (I know I am a bit lazy but I was away from my main system until a week ago) and the different is huge. The sound I am getting now is very close to the analogue (LP with Infinity RS1) sound twenty years ago. Thanks to cics for getting me close to what I have been missing for so long, the emotion of the music.
I have been trying out the FLAC option in cPlay last night and am totally puzzled! The sound is quite different from WAV from which the FLAC file was derived. I expected that the difference would be minor and that WAV would sound better. On the contrary, the difference is quite noticeable on the get go and what is most surprising, FLAC sounded better, more analog! (There is one case-Joni Mitchell's big band version of 'Both Sides Now' which I haven't decided yet on which is the better version. But the difference is there.)
Does anyone have the same experience or am I just delusional? And can someone explain the difference in sound? I thought the FLAC file was first decoded when load and that would not be any difference at all.
I have given the system another listen this morning and discovered another recording that I prefer WAV over FLAC: Van Morrison's Astral Weeks (Remastered)--Japan import only. My guess is that:
-- there is more high frequency content in WAV playback (easily detectable but why?)
-- less high frequency in FLAC renders some recordings more analogue sounding
But why is there difference between WAV and FLAC playback?
Jumping from b15 to b27 is massive! Analogue together with recreating that natural "un-electrified" sound remains the only reference - doing 192k SRC output of 44.1k material to an external high-end DAC is very special and pure...
Depending on mobo / CPU / RAM setup and power supply, FLAC will have an impact. cPlay decodes all FLAC material into RAM before playback (thus ensuring exact same audio data). FLAC decoding is however intensive ito CPU load. This intensive decoding process should be seen as a severe electrical storm being unleashed. We get a large PS load being generated. How your PSU and mobo setup handles this will affect SQ thereafter. All my music is stored in WAV thus avoiding this issue. A way to reduce this PS spike (which can last several seconds) and its "after effects" is to cap your CPU, i.e. set maximum CPU frequency to lowest stable level ~900MHz.
By turning off alias in SOX, I find the 'electronic' sound of WAV files playback deminishes
I have settled on using juli@ latency of 64 and it in combination with cplay 2.0b27 has been favorable. I can now lower my host clock to 144 and voltage to .85 and I get excellent sonics with sox @ 192 and very few metallics. I'm still not sure about buffer size in cplay. I like small the best for the pristine highs but large is not bad either for the good mids, bass. But I'm confused on what is theoretically the best for cpu usage and sonics. There was a lot of discussion in AA pc audio on this but I was lost in all the debate. Can anybody clarify this?
Relative to 27's sonic impact vs 26 I believe Serge has nailed it for me as well.
Also try SRC and Medium buffer size as host frequency of 144 should not be a problem.
ok will do
I have been running src and 192 sample rate(64 latency in Juli@) for a day and a half ---digital out into my Benchmark---no metallics and the rabbit sounds wonderful!!! I don't know what happened but I am at 140 host clock control, .85 volts in bios and it is rock stable. I always thought that power supply had something to do with it. Now I have cpu on a linear supply, all hdd's on gd's, no fans and I guess the only thing running off the Antec is digital power and analogue power. So with digital out that is a very low load on the antec. Maybe this is the reason, maybe not but any way real nice!!!!
Major thanks to cics for requesting I try it.
Have you removed Juli@'s analogue PCB?
latency now at 48, analogue board off. Please remember I was getting some metallics from juli@ analogue outs with sox. Now nothing @ 64 latency.
Edits: 06/22/09 06/22/09
Got to shut down , going out for awhile. No metallics aftwer 1 hour.
m
.
it goes an hour or so then metallicca. sorry I changed parameters midstream but I have been planning/building the battery supply for some time.
Alfred these are LiFePO4 from a usa company in california.
...they are everything you said. Soundstage becomes large, hardness is eliminated, timbre is spot on.Thank you for this maybe best tweak ever on cmp^2.
One interesting phenomena on the metallics... I am running 50% cpu usage with 192/-146 src but I think I remember usage in the 30% range yesterday before the batteries with src. I get mid teens with sox. But I thought cplay 2.0b27 was an improvement on efficiencies. Can someone who is running 27 with 192/-146 src get a read from task mg on cpu usage? Certainly 50% usage was consistent in the past with metallics for me so that is not a surprise.
ok now I'm getting 25-27% cpu usage with src @192, what the heck would cause it to jump to 50% before??
I have gone 1 hour src / 192 no metallics @ 25% cpu usage. But but the sonics of the battery are overwhelmingly great---several discs I heard as if the first time. One is a classical piece wherein the depth gets 30-40% deeper. Another one with 2 female vocalists singing 'Walk away Renee' it sounds as if they both stepped forward several feet into my room. The air around them is uncanny (you can hear around them). Alfred thank you again!!
Maybe after I modded my juli@ for batteries it takes several hours to settle. I believe now I can use cpu usage as an indicator of whether metallics will come. If its 50% they will come if 25% they will not.
2 + hours no metallics
6 hours on src 145 / 192 sample rate / 48 latency on juli@ and small buffer on cplay---no metallics. The metallics I suffered from earlier was probably due (I'm surmising) to the soldering on juli@ not settled in yet and the power supply being off while I modded the sound card.
It was really unusual that I was getting 50% cpu usage (when metallics ocurred) and now I'm getting 25-28%.
My first impressions after listening for a number of hours, very very impressive!
This is superb. I have a CD which was recorded while I was in and around the orchestra videotaping, and I remember the sound very clearly. This version is the first time where I really feel I am back there in that same hall. Every instrument is clear and distinct and I can tell exactly where it is on the stage, exactly where it was in reality.
CICS, you have performed quite a feat here, thank-you for all the hard work. This has to be one of the few ultimate ways to listen to digital audio.
Daniel Gauthier
Montreal Canada
Gigabyte GA-G31M-S2L, E7200, 2 x 1 gig HyperX LL Memory, Antec SLP-450WR PSU, WinXP sp3 on Compact Flash Drive, Jili@ Analog Out direct to Electrocompaniet Ampliwire II Amplifier, Apogée Convergence Speakers
- Have you tried using just 1 RAM module (reducing memory to 1GB)?
- How is music accessed if using CF drive?
- Any mods on Juli@? By ESI's own admission, Juli@'s 114db SNR DAC is not the best (they instead went for better ADC). The big advantage is using 192k cPlay output which bypasses AKM's internal 192k interpolater. This way, SQ is exceptionally good (including digital VC) from this "average" AKM DAC chip. Very interesting.
1. Have you tried using just 1 RAM module (reducing memory to 1GB)?Actually I was mistaken, I'm using 2 x 512megs of Kingston ValueRam, The 1gig modules are in another PC at the moment, but as soon as I have a chance I want to swap and test different configurations of memory to see what sounds the best. I think I wouldn't go below 1gig as I listen to a lot of classical music and I like the possibility of loading an entire movement into memory.
2. How is music accessed if using CF drive?
Only Windows and software (C drive) and a spare backup drive of a total of 4 gigs are on Compact Flash, the music is on a 2.5" 120 gig hard drive at the moment, but it's much too small. I'm thinking of getting an external SATA box to put 4 drives in RAID 0. The box would be mounted on the other side of the wall in the next room.
3. Any mods on Juli@? By ESI's own admission, Juli@'s 114db SNR DAC is not the best (they instead went for better ADC). The big advantage is using 192k cPlay output which bypasses AKM's internal 192k interpolater. This way, SQ is exceptionally good (including digital VC) from this "average" AKM DAC chip. Very interesting.
No mods yet, I plan to power the card through a dedicated 5 and 3.3 volt linear supply and cut the power connections to the PCI bus. I would love to get an external DAC, say like the Apogee, or the Benchmark since people seem to get very good results with them, but it's not in the budget for a while yet.
I am also using only an Antec PSU to power the computer itself, no mods there yet, but I'm planning to add caps and later to use dedicated linear supplies for everything.
I've got these monster power supplies made for the Meridian-1 phone system that gives all the necessary voltages, I want to try that first, if it works you could probably get them cheap on eBay, but they are heavy, about 25 pounds!
Daniel Gauthier
Montreal Canada
Gigabyte GA-G31M-S2L, E7200, 2 x 1 gig HyperX LL Memory, Antec SLP-450WR PSU, WinXP sp3 on Compact Flash Drive, Jili@ Analog Out direct to Electrocompaniet Ampliwire II Amplifier, Apogée Convergence Speakers
Edits: 06/22/09
Thank you Cics! I agree with other fellows :) I feel the focus of sound has improved, resulting (1) less booming (2) deeper and wider sound stage (3) I can turn the volume louder without losing the sound stage or feeling fatiqueHowever for some reason I miss the "booming effect" in v26... Because I am using bookshelf speaker haha
And I have deleted all my up-sampled files (300MB+ each) and now only rely on SoX.
Edits: 06/18/09
Sox is available to use as a standalone program
http://sourceforge.net/project/showfiles.php?group_id=10706
May be you can compare offline upsampling as well
Hi everyone, I have been on another planet for the past 6 months but am now catching up on all the posts, as well as updating my system to the latest versions.About the composer name issue in cue sheets, there is a tag specifically for this in the specs called SONGWRITER. Me and my friend have been working on this issue and are transferring huge collections of music to hard disc, I've got now over 1000 CDs with the same number of cue sheets to edit and this is a big issue, not to mention managing the GENRE tags as well.
It would be nice if there was a way to support this eventually in cMP and cPlay to make it a little easier to navigate a big collection of music.
I've included a part of one of our cue sheets as an example below as well as the cue sheet standard description for that tag.
Frodan
REM GENRE Classical-Symphony
REM DATE 1977
CATALOG 4988005503473
REM ORCHESTRA BPO
REM FIRM DG SHM-CD
SONGWRITER "Brahms"
PERFORMER "Karajan"
TITLE "Symphony No. 1 in C minor Op. 68"
FILE "Brahms - Karajan - Symphony No. 1 in C minor Op. 68 - 1977.wav" WAVETRACK 01 AUDIO
TITLE "I. Un poco sostenuto - Allegro"
SONGWRITER "Brahms"
ISRC DEF058702021
INDEX 01 00:00:00TRACK 02 AUDIO
TITLE "II. Andante sostenuto"
SONGWRITER "Brahms"
ISRC DEF058702022
INDEX 00 13:22:02
INDEX 01 13:26:00
SONGWRITER
Description:This command is used to specify the name of a songwriter for a CD-TEXT enhanced disc.
Syntax:SONGWRITER [songwriter-string]
Parameters:songwriter-string - Name of songwriter. If the string contains any spaces, then it must be enclosed in quotation marks. Strings should be limited to 80 character or less.
Example:SONGWRITER "Paul McCartney"
Rules:If the SONGWRITER command appears before any TRACK commands, then the string will be encoded as the songwriter of the entire disc. If the command appears after a TRACK command, then the string will be encoded as the songwriter of the current track. Note: If your recorder does not support CD-TEXT, then this command will be ignored.
Daniel Gauthier
Montreal Canada
Gigabyte GA-G31M-S2L, E7200, 2 x 1 gig HyperX LL Memory, Antec SLP-450WR PSU, WinXP sp3 on Compact Flash Drive, Jili@ Analog Out direct to Electrocompaniet Ampliwire II Amplifier, Apogée Convergence Speakers
Edits: 06/17/09
I don't know if you are aware of the interim solution I have been using which is to list composers from the flac files' Composer tag as genre in the case of classical files. Al Jordan has included an option for this in his Recursive Cuesheet Creator. This accomplishes much though not all of what you are after. Under genre display, if one selects a given composer then his works are listed by performer-title.
If you want to use Songwriter in the cuesheets, you have to have a way of entering composer data from the music files into this and then a way of displaying Songwriter as you have described or some other way.
Next cMP release (much later) will look into this.
I have a few suggestions on how to implement it in cMP2.1. Have an ini setting to turn it on or off, Songwriter=True.
2. If it's on, you could have 4 buttons along the bottom for searching:
IF Songwriter=.F.
- Show buttons as they are now.ELSE Songwriter=.T.
- IF button = GENRE --> After selecting a GENRE offers a list by
- - SONGWRITER, after selecting a SONGWRITER, offers a list by TITLE,
- - then offers a list by PERFORMER.- ELSE IF button = TITLE --> After selecting a TITLE, offers a list by
- - SONGWRITER, after selecting a SONGWRITER, then offers a list by
- - PERFORMER.- ELSE IF button = SONGWRITER --> After selecting a SONGWRITER, offers a
- - list by TITLE, after selecting a TITLE, offers a list by PERFORMER.- ELSE IF button = PERFORMER --> After selecting a PERFORMER, offers a
- - list by SONGWRITER, after selecting a SONGWRITER, offers a list by
- - TITLE.
- ENDIF
ENDIFBTW I love the fact that with a keyboard one can type the first few letters of a search and cMP goes directly to the first entry that starts with those letters. It makes things a little easier when navigating though a collection of 1000 titles!
Daniel Gauthier
Montreal Canada
Gigabyte GA-G31M-S2L, E7200, 2 x 1 gig HyperX LL Memory, Antec SLP-450WR PSU, WinXP sp3 on Compact Flash Drive, Jili@ Analog Out direct to Electrocompaniet Ampliwire II Amplifier, Apogée Convergence Speakers
Edits: 06/22/09 06/22/09 06/22/09 06/22/09 06/24/09 06/24/09
Hello, Cics, hello everybody! We are happy to get one more wonderfull version. This one is less "intensive" as 26, but much cleaner. Soundstage is very detailed, and the instruments and their respective lines are naturally separated, which was not so evident before. Also the highs are on a new level of intelligibility, being sharp and "furry" but smooth. Like sable fur. With this version SRC is gaining up, especially for me, with Sox my best setting was 97 minimal no alias. Overall a shining new development of my only player. I also have to report that, even after minlogon, I had been having short dropouts-hiccups appr. every 12-17 minutes with 26, while with 27 I experienced only 1 and much less pronounced. So, if any changes were done in this area, they must have been beneficial as well. I really enjoy 27.
Serge.
My question to Cics - can You tell me how does volume control affect the resolution? I am afraid to touch it, it is always 0.
First, I'd like to thank CICS for his wonderful contribution to audio. I agree with Serg in his observations on 27. It is wonderfully musical while remaining precise. On my system much better bass. Wider presentation of sound stage. I am running on a dedicated lap top with most of the cmp mods applied. Perhaps with less availble cpu power the code optimizations have an even bigger impact.
Last section of this post explains modern digital VC. cPlay offers 53 bit numerical precision, i.e. no resolution loss. I'm currently using this digital VC [-24db .. -30db] into DAC.
Analogue VC only has advantage when preamp's SNR betters DAC's. Juli@'s DAC offers a SNR of 114db - not sure if you can get an affordable preamp that better this!
This is an excellent software volume control.
I am using the lightspeed attenuator posted on diyaudio and find it a wonderful companion to cmp/cplay. Very affordable and to my ear allows the full potential of cplay to come through. No question that with cplay at 0.0 and the lightspeed handling -40db of attenuation the sonics are much more complete.
Dear cics,wow, 840 MHz and my system is still running absolutely stable and without any dropouts or so - maybe I should even try to get lower... ;-)
I had settled on CPU Host Control voltage of 0.76875V, couldn't go any lower - tried again today and it seems as if i can go lower now - quite a surprise, I had to do a CMOS reset before when going any lower - is the system "burning in" or somehow getting adjusted to lower voltages over time...?
Anyhow - look forward very much to see how 2.0b27 will perform in a few days - I really believed it when you said "no more new releases in mid-erm", and here we are with (i) a release that has implemented a new and very promising upsampler (I'm not using it currently due to Buffalo32S though) and (ii) a new release with major refinements re footprint and performance - thank you so much cics, life is good :-)
Best,
Robert
Edits: 06/15/09
Lower CPU frequency allows for lowering CPU VID. In my setup, I have VID @ 0.73750V (one notch lower than before). At this level, CPU temperature (at max load) as measured in BIOS is 35°C and through RealTemp its 46°C/45°C (core 0/core 1).
No, sorry, I have to correct my yesterday's post - a cold boot (system was off overnight) is still not possible with VID values below 0.76875, even with CPU Host frequency now at 140 MHz.
When doing a cold boot before, the system as kind of "standard procedure" always had failed to come up at 900MHz so BIOS always reverted CPU Host Clock Control to "Disabled", System Memory Multiplier to "Auto" and DRAM Timing Selectable to "Auto", which I then always reset to "Enabled", "2.00" and "Manual". A warm reboot was always possible without that procedure if the system hadn't been idle too long (~5-10mins).
So it really seems I can't get any lower than 0,76875V. I would have loved to do so though, since every notch or two down brought what I considered still another (although small) improvement in relaxed audio reproduction with a tad more transparency and less distortion.
Or did I miss something?
Hi cics,
not very important, but your new buttons are not looking very well on (my) different screens. Want the old ones back :-))
Regards
Thomas
I currently have one 2.5" laptop drive but am going to need another soon. I was thinking of a SSD hd. I know cics recommended keeping all hd's of same type. But what is the risk of mixing lap tops with ssd?
As far as I know none. My cmp2 & cPlay system boots off of a 4 gig Compact Flash card and the music is on a 2.5" laptop drive.
Daniel Gauthier
Montreal Canada
Gigabyte GA-G31M-S2L, E7200, 2 x 1 gig HyperX LL Memory, Antec SLP-450WR PSU, WinXP sp3 on Compact Flash Drive, Jili@ Analog Out direct to Electrocompaniet Ampliwire II Amplifier, Apogée Convergence Speakers
Fro,
I know I can search for that info on the net, but in the interest of time and clarity, can you say more about how you have your 4Gb CF Flash Card setup as the boot drive and how it's connected into the computer? Also, does it connect via IDE or SATA?
Finally, what type of sonic differences do you hear with that as your boot drive versus a regular drive?
TIA!
Greg in Mississippi
Hi Greg
i use the same device . You can easyly change cards with different configurations and windows is separated from my music storages.And Cheap!!!!.they are connected with an CF to SATA Adaptor.It's not better sonically than an external SATA Drive with external PSU , but very practical.
As sonics said it's using an SATA to Compact Flash adapter, and the only advantage I see is less vibration from a hard disk - there's enough of that with the one with the music, and it makes the drive that has the music dedicated to it - ie put Windows on a different drive, so that random Windows disk access won't interfere with cPlay's access to the music drive. I'm planning to have the drives in a separate box later but in the meantime there's only one 2.5" disk spinning.
Daniel Gauthier
Montreal Canada
Gigabyte GA-G31M-S2L, E7200, 2 x 1 gig HyperX LL Memory, Antec SLP-450WR PSU, WinXP sp3 on Compact Flash Drive, Jili@ Analog Out direct to Electrocompaniet Ampliwire II Amplifier, Apogée Convergence Speakers
Edits: 06/19/09
Can you guys tell us which brand and model CF card you are using? Some of my research on this subject is saying a fixed one is recommended.
Brad
I use a Kingston 4 gig elite series compact flash card. I have had no problems so far and have rebooted the pc at least a hundred times.
Daniel Gauthier
Montreal Canada
Gigabyte GA-G31M-S2L, E7200, 2 x 1 gig HyperX LL Memory, Antec SLP-450WR PSU, WinXP sp3 on Compact Flash Drive, Jili@ Analog Out direct to Electrocompaniet Ampliwire II Amplifier, Apogée Convergence Speakers
For your Kingston Elite what file system did you format with, FAT32 or NTFS?
Brad
NTFS
Daniel Gauthier
Montreal Canada
Gigabyte GA-G31M-S2L, E7200, 2 x 1 gig HyperX LL Memory, Antec SLP-450WR PSU, WinXP sp3 on Compact Flash Drive, Jili@ Analog Out direct to Electrocompaniet Ampliwire II Amplifier, Apogée Convergence Speakers
Thanks Daniel for your help.
I had sucess with this CF adapter
http://www.syba.com/index.php?controller=Product&action=Info&Id=838
and your recommended Kingston Elite 4gb CF card. Also I think the drives shows up as fixed since its location is under hard disk drives in MY Computer.
What I like about this route is if a new revision comes out you could leave this CF card alone and install the new revision on a new Kingston Elite which are very cheap!!!
Oh yeah, boot up time is right under 30 seconds without any optimizations or bios changes yet.
Thanks again, Brad
I'm glad it worked well! I have an older version of that CF card so it takes longer to boot.
Daniel Gauthier
Montreal Canada
Gigabyte GA-G31M-S2L, E7200, 2 x 1 gig HyperX LL Memory, Antec SLP-450WR PSU, WinXP sp3 on Compact Flash Drive, Jili@ Analog Out direct to Electrocompaniet Ampliwire II Amplifier, Apogée Convergence Speakers
I bought mine from this guy and paid extra for priotity shipping, it arrived in 3 days.
Brad
ebay store
http://stores.shop.ebay.com/Yogi-Computers-Inc__W0QQ_armrsZ1
Thanks, that's allot cheaper than the ones I was looking at.
Brad
The places I have looked for the kingston elite showed they were out of stock so I explored some other options.
I found a new open box CF to sataII that included a Seagate ST1 5gb CF for 20 bucks shipped on ebay.
Link to adapter
http://www.syba.com/index.php?controller=Product&action=Info&Id=838
This adapter has the option to use 3.3v or 5v with a jumper with the source power from a floppy molex (wish it had a sata power inlet).
Link to Seagate CF
http://www.seagate.com/staticfiles/support/disc/manuals/ce/ST1%20Series/ST1%20Series/100325578h.pdf
and a review
http://www.xbitlabs.com/articles/storage/print/seagate-st1-5gb.html
Didn't know this till I looked it up that the seagate is an actual HDD at 3600rpm with 2mb cache.
I went ahead and purchased a 8gb ST1.2 for 17 bucks to try out too. At these price points I am hoping this works out. I have yet to build the cMP yet but have purchased all of the parts and am just waiting for arrival. Will let yall know how this works out.
Brad
Just a follow up of the seagate drives. After installing the OS on these drives I kept getting an error before Xp loaded saying "Your System has no paging file, or the paging file is too small". From what I could find the fix for this would require the purchase of an MFT editor so I ditched the effort. The other negative with this drive was it showed up as removable storage. Time to sell these drives on eBay.
Brad
Another solution is to get the Sata harddisks external with external psu.
I use the 'Sharkoon Sata quickport DUO'. It can has own psu and can handle 2 hdd's. Very easy.
Sorry if this has been answered before, but I was wondering what was the number besides certain track number when I load a CD into cPlay?
example :
01 26:50 Feierlich. Misterioso (04)
Whats the (04) ?
It's not in my cuesheet...
One other thing, will it cause me any trouble if I sometimes have ISRC codes in my cuesheets?
Thanks
It's the amount of pieces the track was split into, in order to fit into ram. In this case, the 26:50 track was split into 4 chunks.
I can't help you about ISRC.
Have a nice day
Hi,
I made some updates to CPlayList Editor: See the attached link for information.
Alan
aljourdan,
I've worked with your recursive cue creator in the past and it works wonderfully, placing the cue files in the same directory as my artist/albums.
Using the cPlaylist editor in the same fashion, and placing the files in the same directory, I cannot seem to get them to play with cPlay either by loading them directly from the directory or by selecting the optoion to play the list right from the cPlaylist application.
Any ideas on what the problem might be?
Thanks,
Jack
Thank you Alan - great job :-)
Regards
Thomas
I guess I should have looked before I leaped.
So I guess one if it does do a good job of collecting the information you would still need EAC to make a cue sheet. Not the worst thing to have to do BUT one wonders why this thing won't do it for you.
Oh, well ...
Of course, it I am incorrect I would greatly appreciate being corrected.
Rick wrote:
". . . you would still need EAC to make a cue sheet."
Al Jordan's freeware Recursive Cuesheet Creator (RCC) makes cuesheets that work with cPlay and (at least for me) Foobar.
It's far quicker and more reliable than writing cuesheets with either EAC or Foobar. Once I was familiar with it, I deleted all my cuefiles (search for *.cue) and re-wrote the lot. It took less than five minutes.
And while you're at it, take a look at the photographs on Alan's homepage.
Best
Dave
and the photographs are certainly worth viewing.
What an interesting fellow. I have bookmarked both.
I will report if I can hear any difference between the EAC and db... files.
Thanks again for the help,
Rick McInnis
Nope, you are correct. Sorry.
They have been "just about to come out" with cue sheet support for years now.
Eac is still the one to use IMHO for a cmp box.
Dear fellow inmates,I would be quite interested if anyone of you is playing SACD recordings through cMP2 - somehow... I know that SACD is typically 1/2822.4 DSD with a totally diffrent physical layer etc., but I've seen so many creative and brilliant solutions here that I would not be surprised, and thought I'd just ask, maybe I can learn once more... ;-)
Many thanks
Robert
Edits: 06/03/09
it has been said the main reason SONY developed it was to eliminate the ease with which we can play with REDBOOK.
I start this comparison with the caveot... your mileage may vary. Watching the oversampling vs no-oversampling discussions currently taking place just reminds me that how a particular piece of gear or tweak will sound to a person will vary based on their listening bias and their system. So this is MY comparison... you may hear other differences and come to different conclusions.My setup is a highly-tweaked cMP... 7 separate linear supplies, one each for P4, ATX-24 (provided by a modified PicoPSU), HDD, touchscreen/USB, Juli@ digital, Juli@ DAC analog supply, and Juli@ analog output stages (see my threads "PSU Follies" and "Revenge of the Power Supply Follies"). I've applied all the customizations and blazed the way a bit for the GA-EG45M-UD2H... I still need to write a report on setting up that m-board. I'm using the analog outputs of the Juli@, which are performing very well with some mods (see my thread "Juli@ follies"). I used a 256Mb stick of ValueRAM at 3-3-3-5 on both motherboards, preferring it to both the 512Mb Mushkin and 512Mb HyperX RAM. It's cased in a damped Zalman with everything taken out or off that is not needed to run the system... for example, I've even removed the door that hides the front ports.
The rest of the system is a pair of modified Eminent Technology LFT-IVs planar magnetic speakers driven by Electronic Visionary System's 500M's (modified B&O ICEPower 1000ASP modules), a home-built shunt attenuator using a Vishay series resistor and a silver-contact Shallcross switch, and various cables, most of them hard-wired at one end. AC filtering is provided by a couple of Hammond chokes, one for the amps and another for the cMP. I'd also had a few days off for the US Memorial Day holiday and I spent some of the time catching up on my system tweak backlog... rebuilt the speaker crossovers (now all film-caps), braiding a Rylands-inspired extension cord for my dirty supplies, rebuilding my m-board and dirty supplies with larger filter caps (15kuf-> 47kuf), plus a few tweaks to the m-board supply regulators. These all helped make the system a bit more natural and revealing for the comparison.
On the GA-EG45M-UD2H... I did a new WinXP install for it, disabling the un-needed devices in the BIOS before the install to minimize un-needed devices that needed to be disabled later. I've also gone a bit farther on some of the other customizations, going after all networking drivers I could find in Autoruns and disabling all USB devices except the one providing the connection for the wireless mouse and touchscreen. My latest BIOS settings for this board are on the cPlay 2.0b25 thread and as I said above, I'll publish a lessons learned post in the near future. Also in preparations for this comparison, I updated its BIOS to the latest F3 version from the Gigabyte website, which didn't add any new settings or make any sonic difference as far as I could tell.
To setup the comparison, I configured two harddrives (same models ordered and received at the same time) with a Windows setup for each board (this is made easy by having a separate system partition on the drives AND using imaging software to make and restore system partition backups. Then when I put in the G31, I spend a few hours first updating (to F6), then re-configuring the BIOS (this was not the board I used for my cMP before, but the one out of my ripping machine) and performing some of the additional system configuration tweaks I'd learned setting up the EG45M. But ultimately, I need to do a fresh install of WinXP for the G31M... it was my first successful cMP setup and and I now know some better ways to do the install and configuration that will make it a slightly cleaner install (that may sound a little better too, I hope)... but see below, it was pretty good without doing that.
And I stuck with cPlay 2.0B25 for the comparison to limit the variables, as B26 had just become available and it looks to need some tweaking to perform optimally with SOX (my next project!).
So enough of the preamble... how'd they compare?
On the side of the GA-EG45M-UD2H, as I'd mentioned in my preview, it has a very robust sound... bass and midrange sounds are strong and full. This was noticable from the start and was enhanced the BIOS settings were moved towards optimum. But also as I mentioned in that preview, the highs and level of details lagged behind the G31M.
On the side of the GA-G31M-S2L, first, even though the EG45M has a very 'robust' sound to its bass and mids, the G31M does not sound deficient in those ranges... it's just that the EG45M has a stronger sound there. And the G31M did have more delicate and detailed trebles which provided an increased sense of details and a "see-through' quality to its sound... as I have mentioned in the past, the cMP with this board gives even average recordings a sense of being a direct-to-disk recording, something not evident with the EG45M. But I was surprised by what I had not remembered about its sound... I get a much more engaging and emotive sound out of the G31M. PRAT and drive are better... I find myself being draw more into the music and wanting to listen to more and more recordings... something I didn't get with the EG45M. Bass was very detailed and tuneful and the overall sound is very dense with details... yet all are there and separate from each other. During the setup and breakin of the EG45M, I'd gotten used to its sound and forgotten about that aspect of the G31M's sound. I should say that the EG45M is not bad, but in my setup, with my listening bias, it does not provide the magic that the G31M does.
While doing the comparison, one thing I noticed was the difference between processor chips... I kept the one I've been using in my cMP in the EG45M and used the one from my ripping maching in the G31M. That one would not allow as low of a VCore setting as the other... .77500v or so was about as low as it'd go. So I swapped processors between the boards... no significant sonic differences that I could tell (one is a 7200, the other a 7300) except that the 7200 goes down to 74375v and may go a bit lower and this provides some subtle, but important benefits in the darkness of the background.
I did try turning back on 'Spread Spectrum' on the G31M board to see how much of a difference it makes in the sound... it made a noticable dent in the purity and clarity of the highs and details, but it didn't make the G31M sound like the EG45M in the treble. If I get time over the next week or so, I'll try duplicating the EG45M's processor speed settings on the G31M to see if that is the cause of the differences. The G31M will give a stable processor speed of 900Mhz while the EG45M bottoms out at about 1200Mhz... this along with the absence of a setting to turn off Spread Spectrum are the two most significant (AFAIK) omissions in the EG45M BIOS settings. cics provided a few welcome suggestions to the EG45M BIOS settings which are noted in the cPlay 2.0b25 thread that eliminated some of the other omissions I could see.
Bottom line... the GA-G31M-S2L is staying in my cMP, I find it more musically satisfying and engaging. I put my $ where my heart is... there's a vendor with a few of these boards on Ebay and I purchased another as a spare. I really, really hope that we can get an update to the BIOS of the GA-EG45M-UD2H board as the delicacy and PRAT of the G31M combined with its robustness in the bass and mids would be killer... but for me, the G31 has the magic, the EG45M does not.
Greg in Mississippi
Edits: 05/31/09 05/31/09 06/01/09
Carcass,Thanks for filling me in on this. DANGED, another setting we need on this board!
Have you tried using 3rd-party software like Double-Dawg (http://www.mark-knutson.com/dawg/) to make these settings changes?
cics,Thanks for the suggestion. I will give that a try on the GA-EG45M-UD2H (and also try the Mushkin and HyperX RAM again), likely this coming weekend. And of course, I'll report my findings.
Greg in Mississippi
Edits: 06/01/09
Ok... put DoubleDawg on the cMP and with the G31M on my fully-configured cMP, there are no other PCI connections shown except for the Juli@ card.
That's a good thing, right?
Greg in Mississippi
You need "PCI Latency Tool" - link below.
This latest version 3.1 is not the best, in the way it 1) doesn't always show the soundcard itself, and b) installs a device under Hidden devices. But it's still useful, since it allows you to see all onboard devices - I guarantee you'll see at least 5-6 of them, more if USB is enabled etc. Not all of them are programmable, those that have default setting of 0 are most likely not.
I actually use version 2.7, which should be still available somewhere on the 'Net.
Ok... I installed the PCI Latency Tool 2.7 on my cMP tonight and it showed several (6, I think) more PCI devices, but all already had a latency of 0. Only the sound card was at 128.
A couple of them were devices I'd disabled while doing the cMP optimizations.
Am I good now?
Greg in Mississipp
So, is it true for both G31 and G45?
I only tested the G31 last night. I'll publish results with the G45 when I get it back in the system... now looking like a week or two out.
Thanks for confirming that I've got the PCI Latency setup correctly now and also thanks for educating me on how to do that and what tools to use.
My only disappointment is that I didn't have a sound quality improvement to gain!
Greg in Mississippi
Thanks for the info.
I've located what is supposed to be v2.7... will try it tonight and report results.
Greg in Mississippi
I'm suggesting this as this mobo forces a higher RAM speed and your current 3-3-3-5 setting may be too aggressive.
At the BIOS level, that is correct, both m-boards provide a single setting for PCI latency.
Of course, if you're using one of these boards for a suggested cMP configuration, that is not an issue as you have just one PCI card being used.
Is this an issue for how you're using it?
Greg in Mississippi
... also setting latencies for all other (on-board) PCI devices. In my system, based on "old-fashioned" Biostar mobo, you can easily hear an improvement from setting latency for couple of devices to 0 (directing them to release bus immediately), in addition to setting sound card to 128.
As I understand from some other posts, neither of these two boards allows setting PCI latency individually by device using PCI Latency config tool - although they allow setting it for all devices in BIOS (cics recommended 128).
Is that correct?
Refer to ReadMe.txt for important details & specifications.
Change Log (2.0b26 release):
- Increase support for large WAV files to 4GB (from 2GB)
- SoX 14.2.0 resampler implemented. Settings allow for VHQ (Very High Quality) or HQ (High Quality) converters. Important resampler effects includes phase, bandwidth and (above bandwidth) aliasing.
- DSP and ASIO refinements
- Added SSE3 B9 version for CPUs supporting SSE3 but not SSSE3. SoX resampler is not compatible with a compiler optimisation. This impacts SSE2 and can be avoided by using SSE3 or higher versions.
- UI allows for most settings ( excluding ASIO, Output Rate, Buffer & AWE ) to have immediate effect. "Apply" button in Settings applies new settings during playback enabling easy comparisons.
- Improve UI (new icons and enhanced Settings window)
- Enhance gapless processing and background loading
Please REMOVE previous versions before installing cPlay 2.0b26. Normal (SSE2), SSE3 B9, SSSE3 B9 and SSE4 B9 versions are available. Running cPlay on a CPU not supporting the required instruction set will cause cPlay to exit immediately.
For cMP² users, review these BIOS changes (underclocks FSB, CPU & disable additional USB items). New settings are in effect allowing for lower power consumption (below 20W). Recommended Gigabyte mobo GA-G31M-S2L is no longer available, use either GA-G31M-S2C, GA-G31M-ES2L or GA-EG45M-UD2H ( advanced settings discussion ).
SoX 14.2.0 Settings
SoX implementation in cPlay enjoys full optimisation as that for SRC (Secret Rabbit Code). This includes 128bit DSP processing and other optimisations. Only VHQ (Very High Quality, 175db rejection aka Stop Band) and HQ (High Quality, 125db rejection) converters are supported. SRC at 145db SNR (154db rejection) and 121db SNR (120db rejection) remains as before. SoX VHQ offers better than 170db SNR performance (see measurements below). Other SoX resampler options available in cPlay are (as per SoX Manual):
- Phase (0-100)
All resamplers use filters that can sometimes create `echo' (a.k.a. `ringing') artefacts with transient signals such as those that occur with `finger snaps' or other highly percussive sounds. Such artefacts are much more noticable to the human ear if they occur before the transient (`pre-echo') than if they occur after it (`post-echo'). Note that frequency of any such artefacts is related to the smaller of the original and new sampling rates but that if this is at least 44.1kHz, then the artefacts will lie outside the range of human hearing.
A phase response setting may be used to control the distribution of any transient echo between `pre' and `post': with minimum phase (0), there is no pre-echo but the longest post-echo; with linear phase (50), pre and post echo are in equal amounts (in signal terms, but not audibility terms); the intermediate phase (25) setting attempts to find the best compromise by selecting a small length (and level) of pre-echo and a medium lengthed post-echo. Note that phase responses between `linear' (50) and `maximum' (50..100) are rarely useful.
- Bandwidth (90-99.7%)
Band-width is the percentage of the audio frequency band that is preserved. A resampler's band-width setting determines how much of the frequency content of the original signal (w.r.t. the orignal sample rate when up-sampling, or the new sample rate when down-sampling) is preserved during conversion. The term `pass-band' is used to refer to all frequencies up to the band-width point (e.g. for 44.1kHz sampling rate, and a resampling band-width of 95%, the pass-band represents frequencies from 0Hz (D.C.) to circa 21kHz). Increasing the resampler's band-width results in a slower conversion and can increase transient echo artefacts (and vice versa).
The -s `steep filter' option (99.0% bandwidth) changes resampling band-width from the default 95% (based on the 3dB point), to 99%. Band-width values greater than 99% are not recommended for normal use as they can cause excessive transient echo.
- (Above Bandwidth) Aliasing
Aliasing above the pass-band is allowed. For example, with 44.1kHz sampling rate, and a resampling band-width of 95%, this means that frequency content above 21kHz can be distorted; however, since this is above the pass-band (i.e. above the highest frequency of interest/audibility), this may not be a problem. The benefits of allowing aliasing are reduced processing time, and reduced (by almost half) transient echo artefacts.
Resampler Measurements
All output files are generated from cPlay. Both 16/44.1 and 32/44.1 1 kHz test tones are used:
![]()
Upsampling 16/44.1 > 192 (SoX -vLb99.7 on left and SRC 145db right):
![]()
A closer look at the bandwidth option using 16 bit test tone (noise floor ~130db):
![]()
SoX at 99.7% offers all frequencies up to 22kHz (max is 22.05kHz). SRC offers ~96% and cannot be changed.
Upsampling 32/44.1 > 192:
![]()
Downsampling 32/192 > 44.1:
![]()
SoX SNR is better than 170db whilst SRC is better than 150db (notice small artefact ~150db above 20kHz).
Phase measurements done using a 2kHz (2 cycles per ms) 24/44.1 test tone. Transient is a sudden volume increase (50db) from -90dbfs to -40dbfs for 5ms:
![]()
SRC output at 192:
![]()
Both pre- & post-ringing is equally distributed.
SoX output at 192 (minimum phase):
![]()
No pre-ringing but greater post-ringing. Overshoot and ringing level is greater than SRC (due to higher rejection db, higher SoX bandwidth and no aliasing). Note that:
- SoX non-linear (non-50) phase setting affects HF phase response of original input signal (i.e. a trade-off is made wherein increasingly higher frequencies undergo greater phase shift). Intermediate setting (anything between 0..50) offers better balance regarding this trade-off.
- Observed ringing is beyond human hearing when "min(input, output)" is 44.1k or larger.
- Ringing effects can be further reduced when "alias" setting is enabled. This results in a trade-off wherein alias artefacts above Nyquist cut-off is added, e.g. for 44.1k input and 192k output, distortion above 22.05kHz occurs. These levels are low (below 100dbRMS) and not in the audible range.
![]()
Hello, everybody! I know that many people like resampling to 96, or even 192. As I work in graphic design sphere, I often come across images, resampled by other people, who do it to conform to printshop standards. As our wave files can easily be represented, and they are (in professional programs like samplitude), as a black&white image, it is very easy to understand with the above picture, what resampling does to original image (wave). On the picture You see a detail of my cat. To the left - non resampled original (16 bit 44,1 kHz), to the right the same image (resampled 400 % - read to 96 kHz) The arrows show artefacts of resampling. Take a close look. Resampling was done in Photoshop, the king of professional image processing, so it can be compared to Sox or SRC.
If you watch carefully you will certainly find more. To better see the difference, save the image and zoom in on it in picture viewer program.
This is by no means to discourage You from resampling, I thought it might be interesting to represent the fact this way
Serge.
IMO the pleasures are not dubious. Similarly, if something looks or sounds worse then IMO the shortfalls are not dubious. Only if one cannot discern whether something looks or sounds better is anything dubious.
The good thing about cPlay, as I understand it, is that upsampling is an option that can be bypassed. You may find that your DAC has better playback at 24/96 than 16/44.1 or 24/192. You may also find that software upsampling is superior to hardware upsampling and perhaps you prefer software upsampling with a NOS DAC or no resampling with a NOS DAC. On the other hand do you really want to resample in both software and hardware and does it sound better?
It always seems to come back to implementation is everything.
Take a low-res pic which would be our 44.1k. Then take a hi-res pic (@ ~4x) to be our 192k version. Yes the upsampled results are as per the hi-res picture if strict bandlimited interpolation is done (SRC does this; I'm not able to verify SoX but their site suggests so).
At issue with the upsampled hi-res pic would be added artefacts ("ringing" or more correctly "Gibbs phenomenon"). New information is added wherever we go from low to high intensity (i.e. discontinuities). This means at these boundaries the low intensity area now contains information that's NOT visible to us. This would be in the UltraViolet (UV) spectrum. Aside: Radio, TV, Microwave & Infrared are lower down the spectrum to visible light. Beyond UV are harmful X-rays and Gamma-rays.
So does this invisible artefact affect the visual? Experimentation with SoX confirms this to be true. SRC offers no settings but has a redeeming feature in that the pre-ringing (same for post-ringing) is just ~0.3ms in duration. SoX offers settings that reduce the pre-ringing but at the expense of changing some aspects of the visual (phase shift). Although the pre-ringing is short, the levels are relatively high (some 36db lower than transient peak) bringing into question how well your speaker's tweeter can handle this ultra-sonic load. Should this be at or near the break-up or resonance frequency, your tweeter is most likely to loose the plot...
Hey serge,
Forgive my ignorance, but I thought all this depends on the dac. Some handle different rates better dont they?
Hello! Yes, I think that generally resampling takes place within dac but not in case of cMp-cPlay, I believe, where it is done by src or sox and then fed to the dac to be used for conversion unaltered. The process for resampling can be different depending on the math. In graphics, which, I repeat, is very similiar to wave, the math can be bicubic, bilinear and some others, also different dithering (even the terms are the same) processes are involved like bayer, fatting, floyd-steinberg math, diffusion, dispersion... Next, if I remember correctly, some DAC manufacturers employ tracing - the process that converts our wave into scaleable peak which is closest to analogue (according to their decision and or understanding). IMHO one of the best but hardest ways to achieve the best sound is to try to read and convert to analog every bit to the last one, but I am not an engineer nor am I a program developer, so I may be mistaken, but resampling does NOTHING to bring us closer to truth. I will prepare more pictures with dithering and vectorization (bezier) examples, used in graphics that closely resemble what is happening to the wav within the dac or resampling program (if it is interesting). Also, once again I want to say that I am not trying to discourage anybody from resampling, nor am I ungratefull or trying to say that Sics is doing something wrong.
Serge.
what the heck does that have to do with anything. You are listening to a live event through speakers , in a different room, with different amplification, at different volume, at a different distance, with different cabling, in digital, with different DAC"s, and different computer chips, different versions of cPlay (or other players), oversampling DAC's....etc.
It's a friggin representation of an event.
You should know that with your photographic endeavors. Fidelity is a goal but it is more than pure specifications.
If every system sounds different what is TRUTH?
Now remember that I am sensitive to this being a professional photographer. I hear comments like "is that the way it really looked?" all the time. I try to explain that the slide film I use responds differently than the eye to a scene and the Ilfochrome has it's own character which is dependent on the illumination etc etc. Even the most accurate photograph is a representation and fine art is an "Artist's" representation.
So everything is but a representation and higher bit depth and sampling rates are the way of the futuure. Right now the playback hardware is ahead of the media. Non oversampling DAC's will probably never be made again in light of HD audio. So we live with it and find better ways.
Bob
www.PlateauLight.com
D,
I try to explain that the slide film I use responds differently than the eye to a scene and the Ilfochrome has it's own character which is dependent on the illumination etc etc.
I think that is what I was getting at. I thought some dacs like getting info at say 24/96 while others might be better off at 24/192.
I mean the initial wav file. Nothing more.
Serge
I am a graphics pro also.I agree with your assessment however I combat the fuzzy uprezzing by grossly over rezzing and then downsampling.
It is the equivalent of sampling to 352.8 then do a DC offset and downsample to 192.
Of course this is in Photoshop and I have no way of doing it for audio.
Just and interesting idea
Bob
www.PlateauLight.com
Yep, I agree, by upsampling to 400% and then viewing the result in 25% image size, Serge has in fact do two resampling processes. I can usually accept images downsampled using Photoshop internal function, but for upsampling I prefer it done with external plugin such as Genuine Fractal Print Pro. I prefer editing wave file with Wavelab software, but for upsampling, I prefer SRC.
Sopian
yes I agree upsampling de-focuses the sound just like it does in the picture. but I recall what j gordon holt said when somebody told him that solid state devices sound more accurate than tubes to which he said then I like tubes better than accuracy. I go through periods where I like 16/44 better than 24/192. Its all an illusion.
My schoolboy's limited understanding of the vastness of Hindu culture, I do wish I could take the time to truly understand, but the Hindu concept of ILLUSION is one of the most important concepts in coming to "understand" the world.
All is illusion and one makes of it what they will. This hobby is about illusions within an illusion, no one knows what is really TRUTH, it is illusive and its discovery will do no one any good, or harm for that matter, for it is meaningless. Especially in this context.
We try to find something that allows our consciousness to experience an event, an event we most likely did not attend in a hall we most likely have never entered, through microphones and recording equipment we will never own, blah, blah, blah.
I think steppe had a clever insight, but not clever enough. I know there is no "bad" intention on his part, it was interesting, just not very useful. DHT for ME's analogy was far more accurate, to my mind.
This is an interesting discussion but it leads nowhere. With such bare bones as RED BOOK we have to find a way to fill in the blanks and a best version of the "blank filler" will come. Just as with the LP, no one heard what we are able to hear now with LPs at the beginning of digital era, some amazing refinement occurred after what was the end of the LP's commercial life. Who knows what will happen with the CD, but all of us have lots of them and we would like to hear what is on them as best we can, as it was for those with large LP collections twenty years ago.
My fear is that the future, except for a few specialist exceptions, is moving towards even less information, not more. In ten years we might be pining for the CD to return. I would bet someone will find a way to upsample/extrapolate this information into something much better than we would have ever thought possible.
Bye,
Rick McInnis
Hello, Theo! You have found exactly the right word - de-focuses, it softens, washes away the edge. Unfortunately I cannot post a bigger image - or tiff image, where everything is much more explicite - it won't fit on the screen here. Also an old CRT display (which I cherish) is much better at such subtle differences. If we zoom to highier percentages the "96 kHz" right side picture looks awfully washed-out.
I personally prefer no resampling .
Serge.
I just compare SoX VHQ Linear 99 vs Izotope, 44.1 -> 192, and I feel that SOX online upsample is better than Izotope offline upsample. SoX has lower distortion and background is quieter. So I get a better imaging and resolution is higher. I use Lynx aes16e-> Mytek 8x192-> Beyer 990. Result is the same when I switch to Chord+MUBE.
My perception has always been that online upsample cannot beat offline... so I will try a more days. If SOX is confirmed successful then I can delete all the 192 up-sampled wave files (300MB+ each). SoX is not a CPU hog. My E7200 runs SoX without pops and clicks at 800Mhz (Asus P5KPL-CM) and heat sink is cold as room temperature.
Otherwise b26 sounds just as musical as previous versions. I guess this is the result of good ASIO integration.
Gee I thought B25 could not be improved!
Just received an email from a cMP² user on a similar experience as yours, i.e. Online betters Offline.
Which filter do you use with Isotope?
Sorry ackcheng I need to tell you later... My main PC just doesn't boot anymore after I insert the E7400 :( I need to fix it...
With SOX, how the sample rate conversion is performed?
Is it done on the fly?
Or is it first complete the conversion and then play?
I think it may sound better if it is first converted to the target sampling rate before it is loaded to the memory and play.
Thank you.
Given that there's less RAM processing with realtime upsampling, it would be better. It would be interesting to compare batch converted SoX file vs realtime upsampling.
How good is SOX? How SOX compare to leading sample rate conversion software such as SARACON by Weiss if I use it to upsample from 44.1kHz to 96kHz?
The SNR is one of the best. SARACON uses linear phase filter which SOX allows. If you look at the comparison site, SARACON is not that good afterall
Being a lazy man who does not much want to find out for myself; I follow the leader.
I hope this is the one that will keep me from wearing out my cartridge!
Thanks,
Rick McInnis
For starters, I would remove the preamp altogether (maybe give that cartridge a well deserved rest).
I've settled with B26 SSE4 B9 with large buffer setting (E7xx processors have 3MB L2 cache - think of this as a free resource so exploit it). I use a single Kingston 256MB ValueRAM set at 3-3-3-5.
SoX is very good and certainly has the upper hand with transient performance. On some recordings / genres it may be preferred. SRC offers better tonal decay and does a marvelous job of preserving those delicate harmonics. Overall SRC (@145db SNR) is no slouch on transients but its harmonic purity makes it a perfect choice for long term listening. It's like tubes without the haze.
Nothing would make me happier than to not have to depend on the tweaky turntable and cartridge!!!
Nothing like a click of the mouse and then music is presented.
I am looking forward to hearing if that CD ripping software makes an audible difference.
I have yet to install #26 so I am confused about some of what you said. I guess one can go back and forth between the SoX and SRC and make a choice? All of this is attractive even though I will look forward to their being a unified, this is best, configuration. At the same time I will take delivery of an absolute best phono cartridge!!!
I admit, I am lazy, though once one finds a way to optimize you are compelled to use it. At one time I wanted to believe polarity control was ridiculous, until I learned to hear it.
Have you implemented some of the power supply options the faithful have offered? Still using the onboard DAC of the JULI@ or back to the AERO CAPITOLE?
If I was a cat I would be dead.
Thanks!
Rick McInnis
To clarify, I prefer SRC@145db - harmonics & tonal decay is superior to SoX. If I use SoX then I set it to intermediate:30, bw 96% & no alias. Other settings are also good like minimum, bw 97% and alias.
Setup still the same (Scarlatti DAC used with 2 cMP²: RME / Toslink / 96k & Lynx / Dual AES / 192k. AA Prestige SE & Scarlatti Clock collecting dust).
I use Juli@ for development & testing. It still offers best value for money but I wouldn't rely on it to better your TT (and new cartridge). Perhaps doing GStew's mods together with I2S feed to ESS Sabre's 9012 Stereo DAC will suffice. Clean and isolated power to Sabre's digital and analogue voltage inputs is a must.
interesting that you chose large buffer vs small (opposite of my choice). I'll try large. Is there any cpu load differences between large/small buffer? I have several resampled sox files I will listen to both at same parameters and post.
I got the loudest version of metallics ever today while listening to juli@ analogue outs. Who knows maybe I have a real ghost in my machine. I listened all day yesterday and only ocassionally got slight static-y sound when loading ram sometimes--not gross. It was not constant but when I rebooted the 'load static' went away for most of the day. Today the metallics completely obliterated the music. Had to reboot and it went away. I'm running cpu volts set @ .85 VOLtS (READS .800 VOLTS ON CPUZ), 157 host clock control. Maybe I'm at the limits of my machine. Really thought I was going to be ok with sox. The gross metallics started when I tried to go to a prior track during cplay playback. BTW when I compared pre prepared 192 sox files with realtime sox upsampling in cplay the sonic results were almost identical. could not specify a preference. Cpu usage never went above 28% while running sox 192. I can't understand why this monster defect keeps coming back.
The first test involves increasing Juli@'s latency to a maximum 128 samples (that's the maximum I would recommend for 192k output). I suspect you have tested this and got metallics.
Second test is to never jump directly to a track, instead hit Next till desired track.
I want to use Juli@'s I2S output to ESS Sabre's 9012 Stereo DAC (133db SNR) but this problem bugs me.
Did not like 128 latency but I have settled on 64. Sounds good plus I seem to be having fewer metallics as my linear ps on p4 settles in. Also I'm running sox at 192 with alias box checked. This means no filtering of aliases right? If right that means lower cpu load.
I have tried no mouse navigation within a file and for 2 days or so I've gotten 1-2 metallics a day--not bad. Now I will try 128 latency with the juli@ control panel (in past I've preferred 48 sonically) but I assume this reduces cpu load.btw sox upsampling within cplay is better (by a small margin sonically) and infinitely better for parameter setting and disk file usage. I've just saved 15 gigs of hd!
so thanks again for a great cplay 2.0vs 26 release!
Edits: 06/04/09
I ended up setting cpu voltage higher (.896 on cpuz) and host clock control to 153 and it seems to be better (soundwise and metallics wise). Thank you ray konkle for this suggestion.
I got metallics a couple of times while trying to change tracks with the mouse (sorry I just forgot). And although I had to reboot it seems not as bad while running sox at 192.
Also I implemented a Power One power supply (12 volts. 1.7 amps thank you sondale, mark, robert, ray and gstew) for the p4 and it sounds really good (I'm sure it has more breaking in to do) but it also seems more metallic resistant. I have to figure out where to put it now.
But all in all I may be getting a Lynx 2b. Hate to abandon the juli@ but it is really a distraction to reboot all the time. If I can get less than one a day or so metallics I could live with the juli@. We'll see.
I tried only the 'next' or 'previous' buttons or keyboard for navigating within a file all day yesterday and no got metallics (more below in response to Bertel). When I was actively working as an engineer doing problem solving we used the technique called 'Is/Is not' map. It simply was the listing of conditions that are associated with problem occurence or non problem occurence. With cplay version 2.0v18 I never got one experience of metallics ever-over maybe hundreds of hours of use. I went back to it several times just to listen to music during 'heavy metal' days and it never failed in this way. Yes cplay versions higher than 18 are way way better sonically but all were subject to metallics. Not trying to blame cplay or cmp this is just a statement of fact that may shed some light on what is happenning.In addition to only using keyboard for within file navigation I am currently @ 157 hcc (was 155 the day of the big metallics)--so far after 1.5 days ok.
Edits: 05/31/09
Cics,
I can already comment on the second test:
My version of the metallics occurs even when cplay just continues playing and moves to the next track - then (at random) metallics can occur or not. I restart the track, sometimes once more, sometimes again, rarely again, and sooner or later the track starts without metallics. I could discover no pattern whatsover yet to give me a hint on possible causes (other than the combination of latencies maybe).
However, since I am currently focussed on getting the system optimized for Buffalo32s first, it isn't an issue for me for the time being, i feed it with 16/44.1 since it does upsampling anyway. Once that is running fine I'll try 24/192 and will report.
Best,
Robert
you said '...My version of the metallics occurs even when cplay just continues playing and moves to the next track - then (at random) metallics can occur or not. I restart the track, sometimes once more, sometimes again, rarely again, and sooner or later the track starts without metallics. I could discover no pattern whatsover ...' --------- was very close to my experience before I went 2.5" hd, fanless cooling except ----when metallics were bad (real bad) only a reboot would make them go away. Now that I am into 2.5" hdd, fanless cpu cooling and using sox I get metallics much less frequently and I get the kind that are less than 'real bad' (except for other day). yesterday, with using only keyboard to navigate within a file I got none. Also with sox I have not yet gotten any at auto track changes or loads.
Robert I found out that cpu voltage and speed settings affect my metallics. I am right now at .85 volts (as set in bios) and 157 host clock control. I'm sure your equivalent but very low settings may impact your metallics.
what's very frustrating to me is I can hear more purity @ 150 hcc but I get the problem more frequently plus sometimes I get instabilities (system shuts down). So finding the sweet spot is perhaps different for all of us. One thing though I believe impacts metallics, is power supply purity. I am using the P4 Ryelands cap mod which is great sonically but I'm not sure it makes my sys more 'metallic' resistant. Others with linear external ps P4 feeds may want to comment how it may effect metallics. As I recall you use a linear ps for P4 so maybe it just depends on voltage/speed levels plus whether you use juli@ or not.
one last thing I have a benchmark too but I like juli@ analogue out better. When using Lynx did I understand correctly that you can only go digital out with an AES cable? Is there no i2s out with Lynx?
Theo,
I am at 0.76785 (not sure about the last three digits, can't look it up since I am not at home today) and host clock at 150MHz, that's stable but lower values don't work (have to restart from BIOS). At that values everything was fine with Lynx AES-16.
IMHO power supply purity does not help to cure the metallics - I am on pure battery power with well-sized caps (so I think) for P4 and P24 and got the metallics just the same as when I power everything with the Antec (I have tested that already). Maybe things are different when I power Juli@ externally as suggested by Alfred/sonics, will test in the next few days.
However, my task for the next few weeks is to go from Juli@ via i2s to Buffalo32S and optimize power supply (and possibly also caps and resistors?) in that chain. Since Buffalo32S does upsampling I output 16/44.1. Once that is running fine, I will switch to 24/192 and 24/178 (thanks Jean) and see if the metallics are back and how stable and effortless Juli@ can perform. Will report.
And yes, as far as I know you can only go out digital from Lynx with their breakout cable that provides you with an AES/XLR plug which I ran into the Benchmark DAC1, worked absolutely stable and reliable. I tried both to find any possible way to (i) go out i2s and (ii) power the AES16 externally, but failed with both (probably could find a way for (ii) but (i) is impossible AFAIK, Lynx techs have confirmed this). Otherwise I would happily stay with Lynx without any doubt.
Cheers
Robert
thanks for info.
Ok I am trying not to ever use my mouse again--using keyboard to navigate. I was planning on getting a buffalo32 also but I guess you're saying metallics might interupt that too. I will try juli@ latency at higher values than 48 and see what that does (I know 48 sounds best). does that help alleviate cpu load?
Theo,I'm afraid this doesn't really help to solve the issue, but I thought I'd just share:
I am a late Juli@ convert, have been running my highly optimized system (at least so I believe) until last week with Lynx AES16. I left everything as it was (e.g. 256MB ValueRam, E7200 with CPU voltage at 0.76875 I believe, all optimizations done etc.), just took out the Lynx card, uninstalled the driver, plugged in Juli@ and installed the unified driver 1.05 - and instantly and for the first time I had the pleasure to get to know your metallics - it's dreadfull...
Since I'm currently going Buffalo32S (which does resampling and reclocking, something I definitely don't like...) I went back to 16/44.1 and everything is fine. So this is just to say that a completely stable and (for me) close to perfect system tipped over just by exchanging Lynx with Juli@. I definitely have to have a Juli@, since Lynx can't do i2s and external power (and cics said its ASIO implementation is so fine), but I don't really have a doubt that it is either Juli@'s driver that screws up, or Juli@ hardware - now the latter was my theory before you said that you have the metallics even through Juli@'s analogue outs, I now expect to get them even over i2s, still to verify, if so it can't be the digital out part. Ah yes, and I had the idea of experimenting with PCI latency setting in BIOS, don't know if that could make a difference, haven't tried yet but could make sense.
Just sharing another experience...
Robert
Edits: 05/29/09 05/29/09 05/29/09 05/29/09 05/29/09 05/29/09 05/30/09
thanks Robert. Maybe the solution is the Lynx. I know its expensive but I bought power cords more costly. What is you theory about pci latency? Do you think something other than 128mb might impact it. Are you running a ssd hard drive? I am happy to hear you report it. If more people taslk about it the chances of a fix go up.
Theo,Lynx is definitely a solution - if you can live with XLR output to say the Benchmark DAC1 (as I had before, that was good and so stable and absolutely hassle free) and do not fancy i2s (like I do now to feed the Buffalo32S). On the other hand, once you go Buffalo32s you only need 16/44.1 since it does upsampling and reclocking anyway, so there will not be a problem. I am intending though to build my own solution based on ESS 9018 chip later that year (8ch dual mono configuration) so I've got to find a solution to this, but fortunately that's still far enough away ;-)
Well, my "blaming" PCI latency is in no way scientific or based on any facts yet, just a gut feeling - my metallics occur whenever I start a track (when cPlay set to SRC @ 145.68db and 192.0), it's either there or it isn't, or maybe it also starts with glitches or dropouts - I just have to restart the specific track, it either reoccurs or is gone, no pattern foreseeable. I'm just guessing, but looks to me as if processor's output and soundcard input are not in sync, and that as far as I understand it can only be driver or something "on the way", during transport, i.e. PCI and its latency I guess. As I said, just a gut feeling, haven't had the time (and need due to 16/44.1) yet to really check, one would need to simply try for some time with a setting other than 128. However, my metallic issue was definitely introduced with Juli@ and its driver, everything else has been there and unchanged before and has worked perfectly stable and without any glitches or dropouts or even any distortions noticable.
And yes, I have a SSD only.
Cheers
Robert
Edits: 06/03/09
Seems unlikely to me.
I know I have a DAC with the BURR BROWN upsampling chip and my set-up sounds MUCH better with the computer making the calculations.
It seems implausible to me that any chip could do a better job than a powerful microprocessor and these very clever upsampling software choices.
Have you tried it both ways?
I had the metallic problem when using the computer after 8 weeks of being turned off and installing #25. After it settled in it never happened again. To be honest this was only over a two days period.
After Theo has exchanged so many things I am seriously thinking it could be the mouse. Are you using the cics recommended LOGITECH? I had my temporary problem when making some quick mouse clicks (with the recommended device). Could a somewhat recalcitrant mouse send confusing commands to the computer creating this condition? I know I am making a stretch, but this has gone on so long there must be something less than obvious at work.
Bye,
Rick McInnis
I would have never thought a mouse could be a source of problems but I use a p/s2 type mechanical roller ball mouse with a 6 foot extension (to reach my listening chair). Is this a problem? If I could navigate the music library with my keyboard I'd get rid of the mouse.
when I thought about how the problem was triggered in my set-up it isn't completely ridiculous. A baffling solution to a baffling problem?
I would think you would find this easier to use:
http://www.newegg.com/Product/Product.aspx?Item=N82E16826104080
This is what cics uses (or did use when he recommended it to me). I have always tried to use what HE was using to assure I was getting the same end result. One would think a hardwired mouse would be less likely to misbehave, BUT at the same time you have this antenna in your room and no telling what could be happening.
Even if it doesn't settle the problem it is a pleasure to use.
I know you might be starting to think I work for NEWEGG as much stuff as I have told you to buy! My email address is not a ruse.
I sincerely think this is something worth trying. After all you have tried EVERYTHING else. (If this works I think Stephen Jobs should give me a job at APPLE even though I hate their machines)
Bye,
Rick McInnis
always wanted to try it (but didn't you get this problem with the logitech?). Rick I cannot thank you enough for always trying to help to solve my problems.
Edits: 05/30/09
and that was in the settling in stage after eight weeks of being turned off and then restart and install #25/disable EIST and trying to lower the processor voltage too quickly/
After being on 24 housrs it did not happen again.
And it was, each time, started by some fumbling with the mouse. Hitting multiple buttons quickly.
At least you could get rid of that wire!!!
Rick McInnis
Rick,yes, I've had this confirmed by the designers of the Buffalo32S, the oversampling FIR filter is also clearly mentioned in the data sheet (although they don't really disclose how the upsampling is done and to which frequency), the ESS9018 chip has what they call the Time Domain Jitter Eliminator which basically is upsampling and reclocking, as with your Burr Brown a technique that is commonly used. The ESS9018 is highly praised for exactly this feature, it is said to be very efficient and to do things very well, but I agree with you, I have yet to hear a solution where a hardware upsampler does a better job (or at least not worse) than a software upsampler like SRC or SoX.
And yes, the chip is designed with the ability of turning this feature off, it requires setting a few registers via i2c, but the firmware used by the Buffalo32S designers does not offer that functionality, you need a custom firmware for that. Surely the Buffalos are a very fine piece of audio design and manufacture, but I believe the chip can do even better. Since like you I do not want the DAC to do the upsampling, plus I do not think that our cMP2 environment produces too much jitter anymore that needs attenuating (the measurements cics had done in March last year were already impressive, and so much improvements have been made since then), I will build my own DAC later this year with two ESS9018 chips in dual mono configuration (recommended by ESS for best audiophile performance) and compare - at least that's the plan ;-)
Best,
Robert
Edits: 05/30/09 05/31/09
Robert,
I'm planning same thing.
I was reluctant to select Sabre unit over TI 1704 for the same exact reason you mentionned, ie internal upsampling / downsampling.
This technique allows the widest compatibility with the different sampling rates sound format availlable (everything from 32k up to 192k)
Although this is common practice in Silicon SRC and good to rmove jitter of common source, here we have a machine and a software that provide us with minimum jitter at the PCI Sound Card output which mean (if well implemented) at the DAC Input.
I'm thinking for my dac implementation to try to achieve as low jitter as possible all the way down to the analog which means the use of a Master Clock close to the DAC chip.
The problem here if one wants to avoid too much complexity implies to select a single as universal sampling rate (178K4 or 192K).
Should you select 178k4 it would be perfect for original files sampled at 178K4 and close to perfect with 88k2 and 44k1 providing that you select the right settings in Sox (that i suspect may be full system dependant).
So the question is what will be the standard sampling rates (if any) for Hi res files pick the right clock for that standard and accept imperfection for other sampling rates.
Now that you explained us that Sabre can be forced not to use its internal SRC, it comes back to interrest to me because of its ability to work with 32 bits resolution.
So, if you move on with this project and if we can share energy, you can count on me.
Best,
Jean
Jean,absolutely great to hear you're heading down the same path - or rather up :-)
Exactly, the first assumption I'm making is that we in our cMP2 environment have eliminated or reduced jitter to such an extent that the downside of on-chip jitter reduction techniques, e.g. hardware upsampling and reclocking, more or less clearly outweigh the benefits of doing the upsampling in software and closely coupling the DAC via i2s - in my expectation this is superior, but we'll have to test and see. IMHO the ESS9018 is currently the best DAC chip around, but need to properly re-evaluate.
Secondly, I'm also with you that it is key to use the best master clock available on the DAC. As far as I can see, that is NewClassD's Neutron Star (http://www.newclassd.com/index.php?page=36), Lars has done an outstanding job with this.
Thirdly, the question of selecting one "standard" sampling rate for everything and optimizing for this is easy for me I thought - since all of my playback is through cPlay, I intended to use 24/192 (or rather its 32bit incarnation through ASIO2) through SRC or SoX exclusively. Your comment on 176k4 though suggests that this probably might be the best choice (another assumption I make for me is that most important for me is data derived from red Book CDs, thus 16/44.1, that's what I focus on), probably wil go for that, need to get more info on this, and then choose clock and settings etc. accordingly.
And lastly but most important for me, I'll be absolutely glad to sync with you and join forces once I get started with this part of the project - look forward to this very much! Please PM me (if possible) so I have your coordinates.
Best,
Robert
Edits: 05/31/09 05/31/09
Hi Robert,
Glad to hear from you.
I've selected 176k4 as the best for CD material for the exact same reason mentionned earlier.
Up to now i could only test on my reference system (i'm hifi dealer) with 44k1 and 88k2, to be honest i didn't even try 96k.
But i suspect that upcoming hi res material will be 192k ...
Reagarding clock i was also interrested in Audiocom stuff. Also praised as the best one in town.
The next important thing is power supply and IV conversion.
I will be playng soon with different PSU for evaluation.
My problem with Sabre is how to get it, as i don't want to use Buffalo, buffalo one not any more availlable and at least no more support, buffalo two is too expensive if you take into account that i consider changing IV and PSU. Second big problem Board design.
That's why i am on the conservative approach thinking to use TI 1704.
Best regards,
Jean
Jean,you're probably right that over time there might be a shift to 192kHz material - for the time being for me this still is almost irrelevant since by far most of my music ("old" one I have as well as "new" one I buy) is 16/44.1 because it comes on CD. Let's see how SRC vs SoX as part of cPlay develop, I'd very much be interested in cics' opinion on what sample rate he sees best for Red Book material (176 vs 192) and for what reason (SNR? etc.)
Audiocom is certainly good - need to ask Lars why he's better :-) My knowledge in clock technology and architecture needs to be expanded and updated before I will decide on which to choose.
Well, power supply - you certainly have seen that I fancy battery plus proper capacitance - can only recommend that, need some extremist spirit though ;-)
Buffalo32S IMHO is not expensive at USD 469 at all, given the top-of-the-hill solution you get with it. I'm currently setting it up as my reference, then quite likely will go with 2 x ESS9018 evaluation boards as a next step (not a bargain at 2 x > EUR900), and then take the experience of those two to get to some "final end-all-search" DAC, who knows ;-)
Cheers
Robert
Edits: 05/31/09 05/31/09
I know we will all look forward to your implementation.
Bye,
Rick McInnis
thanks for your response. I'll try different pci latency values.
very nice, the ability to swap sox parameters on the fly is very usefull for finding a 'personalized sonic sweet spot'. I still prefer M for phase but am now experimenting with bandwidth. Seems like allowing aliasing is a no brainer (for me anyway). I don't have a handle yet on overall sonics (but very very close to 25) so far.
Excellent release! I like it!
There is an interesteing article on SRC and filters here. The author compared minimal phase, mixed phase and linear phase filters
http://tech.groups.yahoo.com/group/acourate/files/Uli/
I got it finally
can't sign on to link can you copy / paste here?
I am not sure how to post the white paper as it is a pdf and is quite large. If you send me your email, i can post it to you
I usually let one of my music servers play 24/7, which means that I prefer that it play my entire music library unattended.
Since cPlay requires cue sheets, I thought I could just combine the cue sheets into one large cue sheet to contain my entire library. But it seems that when I combine too many cue sheets (I don't recall the exact number - perhaps two dozen cue sheets representing 24 albums or 300 wav files), cPlay reports an error and fails to play. If I split the combined cue sheet into two separate cue sheets, cPlay plays each cue sheet perfectly.
Is there a way to have cPlay play a large music library unattended? Can cPlay play all the cue sheets in a particular subdirectory folder?
To play all your cue sheets, setup a batch file that calls cplay for each cue sheet (in cPlay settings, set playback to exit):
@echo off
"c:\program files\cics play\cicsplay.exe" "cue sheet 1"
...
You can list all your cue sheets into a file using "dir *.cue > allcues.bat" then edit .bat file as above.
Why are you doing this?
Thanks!
Answer to why are you doing this?
1. I enjoy listening to music
2. It’s similar to playing the radio all day and night, except it’s my collection of favorites
3. No further effort from me is required unless I want to hear something different
4. Too much music, too little time – unattended, unlimited playback means I hear more music
5. The music collection playback may be structured, but my attendance during any given 24 hours is random – thus I am pleasantly surprised or pleased with what I hear
6. I am used to continuous or random playback features in MediaMonkey and iTunes
7. From the Audio fanatic Arthur Salvatore “Optimizing Digital Components -IMPORTANT ADVICE: Never* turn off any digital equipment. It usually takes around 2 days of playing for it to "break in" and sound optimum. Once turned off, you have to go through the entire cycle all over again. Accordingly, do not judge digital equipment unless it has been on and operating for at least 48 hours.”
8. Because I can and it’s so easy
does the SNR setting have an effect if you are at 44.1?
If input media is 96k and output is 44.1 then downsampling is done using selected SNR. If they match match, resampling is totally bypassed.
Hi all,After reading posts of Gstew (linear PSU), Ryelands (extra caps) and others on power supply (PS) improvements I decided to try their recommended PS improvements on my cics memory player.
I can really recommend GStew (linear PSU) and Ryelands (extra caps) suggestions on PS improvements. All the suggested improvements on the PS made (sometimes big) improvements in sound quality (SQ) in my setup.
First my setup:
Zalhman case, GA-G31M-ES2L, E7300, 2Gb HyperX, Samsung 500Gb, passive nexus cooler, Lynx AES16 -> xlr AES/EBU -> Lavry Back DA10 -> xlr mogami gold -> Klein + Hummel O 300 active studio monitors.
DVD, HDD, USB separately powered as per cics recommendations.
But i didn’t like the thought of running 230 AC (!!) Volt wires inside my PC for powering the Granite PSU’s . So I choose an different way too power them. I use an external desktop PSU. (Search for number 998343 – 89 in the search box at www.conrad.nl. It’s not possible to directly link in to their site)Like Ryelands suggests, if one buys a few of these extension cables
http://www.maplin.co.uk/Module.aspx?ModuleNo=34109
http://www.maplin.co.uk/Module.aspx?ModuleNo=98854
It’s really easy to swap suggested PS improvements in (and out) place. Swapping in and out, also makes it very easy to hear the improvements in sound quality (SQ).* Adding a linear power supply the easy way ( very much recommended !! )
The SQ improvement I most recommend is a linear PSU on the P4 pin. Because GStew decided to build his own linear PSU, its looks very difficult too implement. But if one simply buys a 12 V (low ripple !) linear bench power supply / linear lab power supply, than its realy easy and simple too implement. Within 15 minutes the job is done !
And whow ! What an improvement difference in SQ.
Recommended ! Thank you GStew for the suggestion and inspiration.A simple, but good (low ripple ! ) 12V linear bench power supply / linear lab power supply will cost about 30 – 50 euro and a P4 extension cable 3 euro’s or so.
About the aesthetics. Almost all simple linear power supply’s look like this one: http://www.maplin.co.uk/Module.aspx?ModuleNo=231 .
So very ugly. But since the PC, the external desktop PSU and the linear PSU are stored in a board. I don’t mind.* Adding 4700 uF caps (to each of) the +12 V P4 PS lines on the ATX PSU.
(see also Ryelands post) If powering the P4 pin with a linear PSU is not an option, than I can realy recommend adding caps to the P4 power line of an ATX power supply. This also gives a nice improvement in SQ. And - I really hate to say this- I do hear a very slight (!) difference in SQ between ordinary (brandless) caps and audiophile (Audyn) caps.
I soldered the caps directly to a P4 extension cable. Again one might think it’s a hassle. But again it’s surprisingly easy to solder the caps directly to the P4 extension cable and swap the P4 extension cable in place (preparations, soldering, swapping in ect will take 40 – 50 minutes). Because it’s easy to swap the P4 extension cable in and out, one can easy hear the SQ differences.
Note: the SQ differences where much (!) bigger on the Zahlman PSU (with I used first) than adding the caps to the ANTEX Aerth watts 430 PSU P4 PS-line (which I use at the moment)
Note: adding extra 'smoothing' caps to the linear P4 PS-line, made no difference in SQ at all !
* Adding 1000 uF caps to the P24 PS lines.
Al though its more time consuming it’s also easy to solder caps (to the P24 power lines.
It took me 2 hours to complete the job. It’s a precise and a bit boring job. Soldering so many little caps to al those wire’s.
Luckily no caps are needed on the ‘power good line’ (grey), ‘power supply on’ (green) , ‘standy line’ (purple), and the white line (- 5 V).
Swapping in and out the capped P24 extension cable gave only some (not much) SQ improvements.Summary:
1. The biggest SQ improvement in my setup was: a linear PSU on the P4 pin.
(a must do !! Very much recommended)
2. When not using a linear PSU for the P4 pin but a common ATX PSU:
ad the caps to the P4 pin as Ryelands suggests.(a very nice SQ improvement)
3. If you have the time: you can also put caps on the P24 lines. But that only gave minor improvements
-> Final thoughts on power usage and power supply (sink <-> source)
As already suggested by some other inmates here (I can’t find there posts anymore because I forgot there names): lowering the power consumption might give better sound quality because of less demand on the PSU and the need to deliver less power to various parts.Hence: optimizing the power supply could make the need for lowering the power consumption less important or even irrelevant.
Personality I don’t hear any SQ differences in my setup, when the CPU is running at 0,8 volt, 0,9 volt of 1.0 volt or 1,15 Volt. In my setup, it al sounds the same to me. Now mind you: I do hear minute differences in sound quality in my setup:
- when I leave my keyboard connected tot the PS2 connection on the motherboard,
- between various cPlay releases
- after the kernel optimizations,
- when giving priority to background services, disableling services,
- low busspeed to get faster RAM timings
- when adding caps and or adding a linear PSU,
- when powering DVD, HDD and USB separately, etc, etc
I my setup I can hear all these minute differences in SQ.
But there's only one I don’t hear: that’s under-volting the CPU. In my setup (with 3 PSU’s) I don’t hear any change in SQ….Do other inmates really hear SQ differences between various levels of under-volting?
PC fully cMP2 optimized -> Lynx AES16 -> XLR aes/ebu -> Lavry Black DA10 -> XLR Mogami Gold -> Klein + Hummel O 300
Edits: 10/31/09
allow power from a linear supply to flow.
I had suggested this many months back though I did and do not have the motivation to give it a try. Nor the expertise to know which rails would need to remain attached to the P24.
It does seem to me this would make the process rather easy in comparison the approach mentioned.
Bye,
Rick McInnis
Rick,
I'm really sorry, I have been to quick on this... I re-read your posts on this and know understand that your suggestion wasn't to use relays to simply do the "ordinary" switching over to the clean sources once the system is completely powered up, but to use the ATX PSU to do the exact timing and switch on the respective lines, coming from clean sources, by relays at exactly these points in time where the ATX PSU powers up those lines itself, intended to give power to the mainboard but now actuating the respective relays that thus give clean power to the mainboard at this point in time, right?
Clearly a very elegant approach, my congrats - sorry that I didn't get it the first time I read your suggestion.
On the other hand, as far as I can see at the moment I'm afraid there isn't much gained:
- You still have an ATX PSU around - in your concept not used to power up but to actuate the relays, but still it needs to be there, both our concepts can't get rid of it.
- What's more, in my concept I can turn it off once I have switched over, but when actuating the relays it has to stay up, doesn't it?
- You have to provide the "special" rails in a clean version, i.e. linear or battery, too. 5V is already there, that's not the problem, and you can turn them off too once everything is up, but you also have to provide a -12V line (which I have powered by the Antec during boot-up and simply switch off then).
- As far as I know relays wouldn't provide the ramp rates (don't know how critical that is, could be irrelevant or covered by kind of the "rise time" the relay needs to close, would need to test)
I like the elegant concept of kind of "externally" do the timing and sequencing (that's why I was looking for a module or circuit that would do that as any ATX PSU has designed in), while my switching off/over is rather brutal. But I think I would really like to combine both ways and have the switching off and over done by the push of just one button quite elegantly with relays, I think that'll suit me best.
Again, many thanks for bringing that back into the discussion!
Best,
Robert
Robert,
I agree with all of your concerns and was aware that this would be the compromise - you cannot get rid of the thing entirely, my conjecture is that you do not need to. I cannot think of any uncompromised audio component - there is always something that cannot be ideal. The key is to find which approach allows the least interference from the weak link.
If all of the relays are identical (respond similarly) I think the ramp rate concern goes away since the delays would be consistent across the board, the timings would still occur in the correct sequence, yet, all would be delayed the same amount of time. A phase shift ...
I would worry about switching between the power supplies; this could eventually take its toll, like my experience with destroying a GB MB with turning the HDD on and off while the board was "on". I thought I was saving wear and tear on the HDD when I would leave the same selection to repeat over and over for burning in purposes. I did save wear and tear on the HDD!!!
Back to having to retain the switcher - let's face it: these things are not fatal to sound quality or none of us would be where we are. No need to fear the switcher, just minimize it.
I wish I was motivated to pursue the idea. At this point I am back to analogue. I do not think even a perfect power supply will get me where analogue takes me effortlessly (but noisily it is true). I mean the sound is effortless. I have spent many hours getting the system right.
I have a forty years collection of LP's, if I was younger and did not have all of these I would be far more motivated to get digital to sound really good.
Sorry to be so pessimistic but the leap in sound between my old Theta Jade and cMP is HUGE and important but the crevasse between analogue and digital is planetary in size and I do not think the trickest power supply in the world will get us there. I wish, I wish ...
I look forward to hearing what you come up with!
Rick McInnis
Rick,thank you again for your very much appreciated thoughts!
Regarding ramp rate: I agree to what you say, that if the relays are all more or less identical the timing and sequencing, although delayed a bit, should still work fine. However, the ramp rate is a different issue - it is described in the ATX12V Power Supply Design Guide as such: "There must be a smooth and continuous ramp of each DC output voltage from 10% to 90% of its final set-point within the regulation band, while loaded as specified in Section 3.2.3. The smooth turn-on requires that, during the 10% to 90% portion of the rise time, the slope of the turn-on waveform must be positive and have a value of between 0 V/ms and [Vout, nominal / 0.1] V/ms. Also, for any 5 ms segment of the 10% to 90% risetime waveform, a straight line drawn between the end points of the waveform segment must have a slope ≥ [Vout, nominal / 20] V/ms." I am not sure how strict and restrictive the Gigabyte MoBos are on that, but it sounds not trivial to me and might spoil the elegant relay solution. I for my part want to get the switcher out of the system, and I can the way I do things by just switching it off after the rails are switched over to their clean sources - but that surely is personal taste, or probably extremism and spleen are better words to describe it :-)
Your comment about putting the MoBo at risk by doing the switching over an extended period of time is something I indeed fear myself. It is also for that reason that I consider using relays, I think I remember that they don't always have to produce "spikes" like a switch does (I believe, haven't confirmed yet) when switching over from one source to another, but can allow for a smooth turn-on if designed accordingly. Definitely need to look into that.
Although I am not an analogue guy myself (I bought my last Thorens 23 years ago and "went digital" completely some 13 or 14 years ago, for the theory that digital by its binary nature is always more "right", just as I prefer solid state over tubes for quite the same reason), this is one of the things I definitely want to do when finally some day I consider my personal cMP2 incarnation good enough (the power supply is just one, although IMHO important, area of construction to get there): compare with one or two reference turntable solutions! I look forward to that probably quite revealing experience a lot! :-)
Best,
Robert
Edits: 05/29/09 05/29/09
I do remember reading all of that stuff and being suitably deterred.
Then there is the picoPSU ... when you look at that thing it is hard to believe that the ramp rate is being controlled by that! And when you look into the ANTEC, though far more complicated, STILL hard to believe there is that kind of circuitry contained.
I think all of this is controlled on the board. The supplies supply voltage and nothing else. I believe that the relays would have nothing to do with this happening properly.
The only thing I think one needs to retain the SWITCHER for is the absolute first moment of turn on when there is communication between the 5 volts rail and, I think, the -12 volts rail, after that I think all of the complicated actions are controlled from the MB.
Please excuse the fact that all of this is conjecture. I practice empirical computer architecture as an avocation.
Then there is the other side of the supply question; instead of getting rid of this mode of operation getting someone like the fellow at HYPEX to design an audio grade switcher.
I enjoy the conversation.
Bye,
Rick McInnis
Rick,
am I repeating myself? Ido this with pleasure - many thanks for your highly appreciated critical thoughts! :-)
Well, picoPSU and Antec doing the ramp thing: You don't need big and complicated for this, at least not at an outside glance - e.g. look at Actel's Fusion (a one-chip solution, see http://www.actel.com/documents/Fusion_PIB.pdf) or Linear Technology's LTC2928 (even smaller and "simpler", see http://cds.linear.com/docs/Datasheet/2928f.pdf), there's FPGA's and small stuff that I can very well imagine being on such a tiny piece as the picoPSu - the thing is we don't know and are guessing... :-) We could find out by simply trying.
However, I'm comfortable with my switching off/over solution for the moment and for the time being have put off the efforts to find a solution for the sequencing (and possibly ramping) issues.
I do want to get rid of switchers though - not so much for practical reasons but for purity and integrity of the concept - strange extremist position, I know, but "black and white" works best for me :-)
Best,
Robert
Rick,
yes, you are absolutely right, using relays would be a very elegant way to do the switchin off and over (as described above). I have thought of that too but have not used relays before, have no experience with them and (most important) do not know yet "what they do" to the signal. However that's definitely something I will look at next, if this switching is best done with high quality switches or if relays do a better job in signal integrity while not adding any harm to the digital environment (another DC user...).
Thank you for bringing this back to my mind and to our attention in this context!
Robert
how do you power up your lab 12 volt supply to p4? is the lab device turned on first then the p24 power supply or the other way around? this is a a very interesting way to do it. also which amp ratimg supply do you use?
Theo,hope you don't mind if I answer this: It doesn't matter at all which one (P4 or P24) you turn on first - the power-up process is initiated with the push of the power button on the case, so once you do this both P4 and P24 need to be powered on.
Or in other words: Turning on power for P4 or P24 DOES NOT power up the PC - this is done by pushing the button, that causes the PC to power up.
Clarification note: It is exactly that power up process that requires proper sequencing and ramp rates. I have not been able to achieve that "manually" (i.e. switching various lines in a specific sequence etc.), that needs to be done by a module that has the logic to get the timing and ramping quite exact. For this reason I power up with P24 powered either by Antec or by picoPSU, and when the system is completely up I switch over to the "clean" supplies. (P4 does not need any sequencing or ramping etc., you just power up the PSU of your choice and you're good.)
Hope that helps
Robert
Edits: 05/27/09
thank you very much Robert I appreciate both your and Mark's contribution to the power supply saga on cmp.
Mark,
Short response...
I can't comment on Robert's 'SERIOUS' capacitance add-on or separate manually-switched linear supplies... they sound very interesting, but I just haven't gone there yet.
But I can comment that a Pico-PSU with some capacitance mods powered from a good-quality linear supply... it is a greater leap over a 'good' computer PSU (Antec 430) with run-of-the-mill added capacitors than that supply with capacitors was over the stock Antec.
So from the perspective of a fairly simple-to-implement and fairly safe (for your motherboard) solution, a modified Pico-PSU with a good linear supply is a significant upgrade from the Antec-430, stock or casually-modified.
BTW, for mods to the PicoPSU, I'm using 2 Black Gate 1000uf/25v NX replacing the caps on the PicoPSU boards and have added a Black Gate 2200uf/6.3v NX across the 5v and 3.3v outputs of the Pico. I'm not tied to these particular caps (the Mundorf sound very interesting), but I do know these work and know that some others will not work as well.
Also implementing a 12V linear supply is a pretty easy thing if you can solder and follow a schematic. After gethering up all the parts, I did my first setup with dual supplies (12v P4 & 12v into PicoPSU) over a weekend. I just rebuilt them with different filter caps and a different regulator configuration in an afternoon.
Robert,
You've provided a lot to chew on, especially your comments on the Mundorf caps and your research into bettering the ATX-24 supplies. I'll spend some time reading your sources and also get some Mundorfs to try.
Greg in Mississippi
P.S. Mark, here's an interesting article on the value of 'audiophile' caps:
http://gboers.xs4all.nl/daisy/home/g3/139/measure/capacitor-comparison.html
Hi Mark,
many thanks for another valuable contribution to our "Special Interest Group power Supplies" :-)
Although I try to achieve similar results with battery instead of linear power supplies, I hope I can further help by commenting on a thing or two.
Surely the most important recommendation to follow is to add extra capacitance on both P4 and P24 to achieve a clean, steady and quick supply of power in order to support the digital circuits' and components' power demands best possible. Since there is either "power" or "no power" (think of a square wave signal), the ideal power supply has to follow this almost infinitesimally small rise time as good and as quickly as it can, and capacitance (especially the true "low ESR"-like types) support that very well.
My brand of choice to reach that goal is definitely Mundorf (http://www.mundorf.com/english%201.1/kondensatoren.htm, see link below). Their MLytic SI capacitors offer very low inner resistance of somewhere between as little as 11 to 55mOhm, depending on which size you choose, and do so at 100Hz which is where it is most needed in terms of (non-)interference with audio signals, plus a design that is targeted on audio use with lowest internal noise etc. I have received excellent results with them.
In terms of sizing the capacitance I have gone a slightly different route, depending on the current draw on the respective lines:
- On the "low current" lines, i.e. 12V on P4 and 12V and 3.3V on P24, I use 5 x 2.200μF of the 63VDC version of Mundorf's MLytic SI. Since it is crucial how quickly the caps can release the current stored, I chose that value, in my perception they are unmatched for that purpose. I topped them off with a 47μF Mundorf TubeCap (unbelievably low ESR of only 7mOhm!) plus a 1μF Mundorf Supreme as bypass cap. The latter keep the loss tangent as low as possible, shifting possible interference very high into the MHz range where the PC's digital circuits aren't bothered anymore and thus can operate in a clear and electrically quient environment.
- On the "high current" line, i.e. 5V P24 (4-5A is already quite a high current for a computer environment), i choose 5 x 10.000μF Mundorf MLytic SI 63VDC. They offer an ESR value of 20mOhm (with 5 in parallel going down to 4mOhm in total) plus an unrivalled ability to quickly release the current required even in such quantities as required on a 4A draw. These are also followed by the 47μF and 1μF caps for filtering purposes.
Although I prefer to also have the source as clean and AC-free as possible (hence battery for me), in my experience with this capacitance setup it is gradually becoming irrelevant which power source you have "behind" them filling the caps as the first-hand power reservoir, as long as the AC interference isn't becoming too big.
Regarding your remark on P24's Pins 11/12 and 23/24:
You're right, as chipsets, components and interface cards required more and more power, P20 was extended to P24. However, there's not a dedicated correlation of Pins to consuming component or groups of components (like "Pin 11 is used to power this and that chip on the MoBo") - in fact all Pins for all power rails are just multiple pin-outs for one single rail of every voltage. This has been done to reduce interference by the PSU's cables, plus to keep their operating conditions modest by using them kind of "in parallel" and thus reducing both their internal resistance and the current that has to come across them. (You can easily "proof" that for yourself by putting an ohm meter to various Pins of e.g. the 5V line, the resistance between them is 0)
For us that results in the fact that we do not have to care about which Pin powers which component and how we best cap that or such - get a decent capacitance setup (e.g. like the one described above) per rail and you have done everything you can. The fact that those single rails are then subdivided on the MoBo with legions of sub-regulators etc. is a completely different sad story that adds interference to a clean power signal provided, but I'm afraid there's not much we can do against that (other than starting to look for and unsolder regulators on the board... *turns in horrror*) ;-)
Best,
Robert
Hi Robert,Thank you for your response. It’s very helpful.
Optimizing the Pin 24 is my last and ‘final frontier’ in my hobby project of building cics cMP2 memory player.
I know battery power comes most close too ‘the ideal power supply’ but I do not find it practical. So I started looking for the next best way to power the P24 pin.
Since powering the P4 with a linear PSU gave a significant improvement over conditioning the power output line on the Antec ATX PSU to the P4 pin with (audiophile) caps, also started looking for a better way to power the P24 pin than with adding caps.
Thank you for explaining how cap parameters are important for conditioning dc power that will be delivered to digital circuits that process audio-information.
I read to many posts where approaches from the analog audio domain are simply copied to the digital audio domain.
Now I know ‘digital audio’ doesn’t exist and that all ‘digital audio’ simple obeys to ‘analog’ physics and electricity laws. But what is valid and relevant in the analog audio domain cannot simply be copied too the digital audio domain.
So I am always very skeptical when I read in posts, that one has to use audiophile caps (true in the analog audio domain) when conditioning the dc power output of a PC PSU for a PC which is processing digital audio information.
I have an engineering degree but not in electronics. I only have some basic knowledge about AC, DC, caps, coils, electronic circuits, etc, ect.
But that’s just enough to understand the concept that when a square wave signal is turned into a shark-vin signal, a jigsaw signal or a mountain/valley signal, that the time on which it’s considered to be a ‘zero’ or a ‘one’ will differ. Or that threshold levels which decide if something is a ‘one’ or a ‘zero’ should be steady, the importance of small rise and sink times, ect, ect. But I have no knowledge at all of, on how these concepts are brought in to practice in real-life electronic products.
But nevertheless these shortcomings in know-how and knowledge, my challenge is to improve the dc power delivery to the P24 pin beyond just adding caps.Since a lot of ‘google-ing’ didn’t shed any light on why there are five red +5 volt wires, eight (yes: 8 !) black ground wires, 2 orange +3,3 Volt wires and two yellow 12 volt wires. I did expect it too have some very good technical reasons. But now I guess it has to do with evolution and backwards compatibility ect, ect.
Already A few months ago I also red the formfactor.org text on the ATX power-up sequence ect, ect, So I understand why it’s difficult (Impossible ?) to power-up from battery’s.
When google-ing a also stumbled opun this (for me) very informative text on how a ATX PSU works at a conceptual level
http://www.xbitlabs.com/articles/coolers/display/psu-methodology.html
It explains why all ATX PSU have troubles with load balancing between 3,3 v, 5 volt and 12 v lines and have irregular distribution of the output voltages when the load is unevenly distributed, ect. It all comes down to that a ATX PSU only has one single voltage regulator but it has to regulate 3 different voltages lines. The bottom line of the paragraph is: an ATX can only regulate one voltage good, the other 2 voltages are regulated within an ‘acceptable range’ with some ‘trics’. Literally the text says :
“this regulation is far from perfect and gives rise to the most common problem of computer PSUs – irregular distribution of the output voltages.”Sow how can one feed the P24 pin with ‘good’ +3,3, +5 and +12 v dc power?
- 2 ATX PSU: one master PSU and one slave PSU. It’s easy to implement.
Each of the ATX PSU is only doing one voltage. The loads on each line will not influence each other. But to do with the 3rd power line?- 3 ATX PSU’s ? each doing there own voltage?
Than I will have 5 !! PSU’s supplying the cMP2 memory player.- Pico PSU’s?
How is load balancing done by the pico-PSU’s?
Since a pico-PSU already is fed with 12V DC, it may be more easy for a pico -PSU to balance the 3 loads. But that is just more of a question. I really have no clue on how a Pico PSU works and how it is is balancing the loads between + 3,3V + 5V and +12V. May be pico-PSU’s do a beter job in load balancing. I don’t know.May be a better way of supplying power to the P24 pin could be:
- a dedicated ATX PSU’s for each voltage on the P24 pin
- or a dedicated pico-PSU’s for each voltage on the P24 pinBut since adding extra caps to the power lines on the P24 lines only gave a small improvement. I ask my self: is it worth the trouble to better the power-supply on the P24 pins with extra dedicated ATX PSU’s. Because there may be only a small potential SQ improvement to be achieved.
When adding 2 of 3 dedicated extra PSU’s might not being worth the trouble (and costs) How about using a linear powered 12v pico-PSU (instead of an ATX PSU)?
GStew reports that his pico-PSU powered by a linear power supply sounds better (than powered by an ATX PSU?) So what I also would like to know is:
is it a significant improvement in SQ when switching from a (good) ATX PSU (with smoothing caps on the powerlines) to a pico-PSU ?Can anyone report on the SQ improvement between an (good) ATX PSU and a pico-PSU?
Robert thank you for your response.
Al 3,3V, 5 V and 12 V lines are bundled together behind the P24 connector.
So paying special attention to feeding lines 11/12 and 23/24 is less relevant. One would only connect 2 PSU’s to 1 voltage.LynxL22 dig i/o -> Lavry Black DA10 -> Mogami Gold -> Klein & Hummel O300
Edits: 05/26/09 05/26/09
Hi Mark,wow, what a post! ;-) As usual, let me try to add my 1.8 cent (inflation takes its toll):
Interesting - optimizing P4 and P24 was my start, to get the basics right, and there's sooo much more to do (DAC, amps, HF shielding etc.) ;-)
You're right, battery power isn't always totally practical, but (i) you can get it organized well, respecting load cycles and using a good charging concept, and (ii) as a perfectionist I myself am way into impracticability too often anyway ;-)
So your goal is to achieve an as-good-as-possible power supply for P24 with power derived from AC mains (by the way, you don't want ANY AC traces in your digital pure DC environment, that's also what the caps take care of). I went to great lengths to really pin down which change in PS does what, so I compared
- Antec EarthWatts vs picoPSU powered by Antec EarthWatts (without any caps)
- Antec EarthWatts vs picoPSU powered by battery (again without caps)until i found out that once power is up I can easily switch the main rails (3.3V, 5V, 12V) to whatever alternative PSU I favor, so I compared
- Antec EarthWatts vs picoPSU (powered by Antec EarthWatts or by battery) vs "pure" battery without caps, and
- the same WITH caps.Let's cut a too long story short: Once you add caps (and I would like to repeat that for what I have understood the capacitance I have detailed before seems to suit the purpose best) it's really getting close to irrelevant what you use "before" them as a source.
Always keep the square wave in mind: the most important task is to keep power supply as clean and quick as possible to keep the rise time between 0 and 1 ("no power" and "power", changing ever so quickly) at a minimum! IMHO this can best be done with caps sized the "right" way, plus cables from the caps to the board which are short, have a lot of internal conductors (that brings speed, e.g. I use AWG 6 which is 133*AWG27) and a very low internal resistance ( <.5Ohm/1000ft, just to be safe - that's in the region of an average PCB's traces).
One more comment on the number of wires: you may have seen from my P24 measurements that current draw on 12V and 3.3V lines is quite low (always < 500mA), whereas current draw on 5v line is quite hefty with up to 4-6A (almost high current in a PC environment). That explains the number of wires - 2 for 3.3v and 12V usually is more than enough, whereas BIG current is drawn across the 5V line (MUCH more in a gaming PC or such of course), so getting more wires in parallel reduces the current flow, thus temeprature and internal resistance and so on - you get the picture. The GND ones just complement.
Also one more comment on ATX power-up sequence: As said elsewhere, I tried to build some module that would do this power sequencing and correct ramp rates for me, but (although it would be very elegant) I have put that back for the time being: I simply power up either with picoPSU or Antec (haven't decided yet finally) and when power is up swith off the "special" lines (-12V, Power_Good and 5V_standby) and switch over the main lines to the clean power sources. Keeping the picoPSU powered by battery or linear is not an option since it is a switching PSU in itself (although a rather good one when powered itlsef with "clean" power) - it's definitely not doing load balancing but is just as much a switching PSU to maintain the various power rails as your average ATX PSU, with all its drawbacks too well known.
In your case - I'm sorry to say that - a "clean power source" for every rail indeed would be one linear PS per voltage, there's no way around that if you're heading for the optimum - you simply HAVE to keep them separate with strict and exact power regulation on the proper voltage. (If we're talking about extremes, I'm providing a separate battery PS for every voltage rail for every component, just to be safe, I don't want to think about interference of any kind. There's not many voltages involved anyway, so a quite ok charging concept for all of those is doable. And by the way, you say in your last line "One would only connect 2 PSUs to 1 voltage" - yes, that makes great sense, I have in some places connected 5 or more batteries in parallel to get an internal resistance of equal to or below 2mOhm - think of the square wave, speed and impulse is key!).
My recommendation really would be: Don't think about 2 / 3 / many ATX PSUs (they're switching PSUs, that's crap for our purpose, forget about them!) - start up with one (ATX or pico, doesn't matter), and when done, turn off the "special" lines and switch to your 3 linear PSUs for the 3 main voltage rails that feed the proper caps, then we're talking :-) Maybe a bit of overkill, sure - but isn't that what we are here for, striving for excellence and sonic bliss?! :-)
Oh, almost forgot to mention: Modification on the P24 12V rail (as detailed above, I used switches in setups with and without caps to determine exactly which mod leads to which sonic result) yielded minimal sonic results, 3.3.V was clearly perceivable, and 5V is HUGE - sure, that's the main powerhouse! Interestingly enough, the sonic effects weren't "more/less bass" / "more / less transparency" or such, but less distortion, less fatigue (I thought that was a myth and lack-of-words before I clearly heard it myself), a sonic picture clearly and perceivably more "right" and correct, all in all absolutely worth the effort, for me one of the major contributions for a sonic experience just "right" and "natural" - I'm sure you'll find out yourself! :-) Enjoy!
Best,
Robert
Edits: 05/26/09 05/26/09 05/26/09
OS-CON type caps (made by Sanyo, Nichion, and many others) have extremely low ESR (under 20mOhm at 100kHz). These caps are used on high performance motherboards for the same reasons proposed here. They do seem somewhat difficult to source though...
Edits: 06/13/09
Hi Robert, Gstew,Thankx again for your information. It’s now more clear for me which routes I might want to follow in optimizing the dc delivery to the P24 connector
you wrote: "Let's cut a too long story short: Once you add caps (and I would like to repeat that for what I have understood the capacitance I have detailed before seems to suit the purpose best) it's really getting close to irrelevant what you use "before" them as a source."
1. So (high quality) caps are still a viable option.
But I made calculations on the cost for low ESR caps from Mundorf. It costs about 222,75 euro (12V VP4) and 321,50 euro (+5 V at P24). I think the costs involved are to high versus the resulting effect. It has to be done let costly. Low ESR caps (model’s NSP or NPC) from NIC Components Corp. (ESR 7mOhm) cost about 1 dollar.2. Looking at Gstews contribution:
A Hybrid solution (mixed pico / ATX ) also springs to my mind:
- the picoPSU (with extra caps) fed by a linear dc 12V supply for letting the pico deliver only the 5 volt.
- the ANTEC (with extra caps) for delivery 3,3 V and 12 Volt.
(don't know if the ANTEC wil start but i'll give it some dummy load on the 5 volt, so it has someting to chew on)
3. Your recommendation to use 3 linear PSU's.
But than there will be 5 PSU boxes (!) in my setup.
It’s going to look real messy in and around the Zahlman-case.
:-(I think I start with Gstew’s suggestion.
A second linear PSU that is feeding dc 12 V into a pico-PSUAnd after that I think I will try to make a hybrid solution and see how that impacts SQ.
I also going to look for super low ESR caps. But at a more cost/effective proposition. As you explained, I see the need for low ESR.
LynxL22 dig i/o -> Lavry Black DA10 -> Mogami Gold -> Klein & Hummel O300
Edits: 05/27/09 05/27/09
I used Greg's plan and built a Daniels derived 12VDC Linear coupled to the Pico w/Oscon caps.
Also have a GStew modified Juli@ with regs, BG caps and BNC out. Made a large difference on the positive side.
Have the Hammond L193 inside and AR power cables. No fans.
Sounds clean and sharp.
RayBan
Mark,glad that the information shared was helpful for you.
However I have to comment on your calculation for pricing the suggested capacitance:
For P4 (based on how I have implemented it) you would need 5 x MLytic SI 2.200 63VDC + 1 x TubeCap 47uF + 1 x MCap RXF 1uF. looking this up online in Mundorf's price list, I get to 5 x EUR 7.99 + 1 x 24.90 + 1 x 5.49 or EUR 70.34 in total. Although this is not exactly cheap, it is (i) way below the value you have come to, and (ii) IMHO well worth the money spent :-)
For P24, you need 2 x the above plus 5 x MLytic SI 10.000 63VDC + 1 x TubeCap 47uF + 1 x MCap RXF 1uF. This amounts to 2 x EUR 70.34 plus 5 x EUR 16.90 + 1 x 24.90 + 1 x 5.49 or a total of EUR 255.57 if you want to add capacitance to all 3 rails, or EUR 114.89 for the 5V rail only.
Remember, you don't have to put caps to every red or orange or yellow line, these are just parallel wires that re-connect "underneath" P24 anyway - connecting all red and all orange and all yellow cables to the positive pole to the 1uF cap for the respective voltage rail is enough. (I guess that's how you came to these high values.) And please do have a look at the link Greg provided re cap comparisons, the inner structure and architecture of the caps is just as important as the relevant parameters like low ESR etc. as clearly shown and measured there, I fully support that.
Cheers,
Robert
Edits: 05/27/09
Hi Bertel,
I toke a while but after my holiday two weeks ago, I was able to pick up again the process of optimizing the power supply to the P24 connector.
As a start I ordered a mini-box a 12 V Mini-Box PSU, model: PW-200-M. and a linear bench / lab PSU capable of putting out: 12V DC/ 6 Amp max
As for a start I put some Audyn smoothing caps on all 3 voltages (3,3/5/12) with I had laying around. But the don’t do much. With or without, I can’t hear any difference. I will have to order real good quality caps as you pointed out already in an earlier post. :-)
I also did some current measurements before the 12 V Mini-Box PSU PW-200-M on the 12 Volt DC line that feeds the 12 V Mini-Box PSU PW-200-M.
I can’t copy your current measurements into the P24 connector. The current into the 12 V Mini-Box PSU PW-200-M on start up on my multi-meter varies between 1.6 amp and 2.1 amp max. And during play back it’s a constant of 1,28 amp – 1.35 amps. Depending on FSB speed, grafics, soundcard used ect.
I also ordered a picoPSU (90 watts) yesterday. Some inmates report that the 12V line on these smaller picoPSU’s are just 'passed through' in these models.
The PW-200-M isn't just passing through the 12 Volt line. If I vary the input voltage up and down, the PW-200-M stabilizes it to 11.93 volts. So the PW-200-M isn’t just passing through the 12 Volt line.
If the smaller picoPSU’s realy just pass through the 12 Volt line, than they have a different concept than the PW-200-M has. Since the smaller picoPSU's are real cheap (37 euro’s) I ordered one. I’m curious if this different setup also makes a difference in Sound Quality
If you plan any new current measurements in the future, please post the results if you like. You’re readings are almost twice mine. May be my multi-meter (4 month old but just a cheap one of 75 euro’s) isn’t working properly.
LynxL22 dig i/o XLR AES/EBU -> Lavry Black DA10 -> XLR Mogami Gold -> Klein & Hummel O300
Hi Robert,
grrr :-(
you'r right.
I mis-calculated that way.
Thank you !
At 70 euro's, I indeed think costs are worth the effect.
Thank you for correcting
LynxL22 dig i/o -> Lavry Black DA10 -> Mogami Gold -> Klein & Hummel O300
Robert,
My current setup uses a Power-One 12v / 1.7A supply for the P4 (fitted with several Black Gate 1000uF caps), a un-modified Pico 120 psu (currently changing the caps on another pico and putting together a linear 12v to replace the switcher) and a Power-One 5v 2A supply for the HDDs.
I am about to change it putting all the psu's in another case underneath the Antec case. So having read the various posts it seems to me that I should try your idea of switching the pico lines off and switching in three other linear psu's. So would I be correct in saying that for the P24 supply there would be effectively 4 psus, a 12v to power the picoPSU, another 12v, a 5V and a 3.3V psu? The first 12V (on the picoPSU) would stay connected with the 12V / 5V / 3.3V switching in and out.
Where do you have the switches? I assume that the picoPSU would not be plugged directly into the board but into a socket that feeds into the switch.
I have tried batteries on the P24 but it is the charging that made me change back to the Switcher/picoPSU; if changing over to batteries then I would like to leave the batteries in situ and when the pc is switched off they would connect to a charging circuit. For the moment I have too many other projects to pursue this idea - I am sure that a search of the web will give me what I need to know.
I also need to power my Juli@ from batteries - this one is high on my list as I have to break out the SPDIF and I2S signals to send directly to the DAC.
And of course then there is the time to listen to the music!!
Thanks for your information.
Alan
I am certainly impressed by your industriousness on the power supply saga. So I take it a 12 volt 2 amp supply is ok for p4? Also is the 12volt supply adjustable?
theob, the Power One linear supplies have a trimmer pot on the board - never tried using it though!
Alan,
your setup for P4 and the HDDs already sounds very good (maybe you might want to consider adding an additional cap like my 47uF Mundorf MLytic SI to the P4 line for supplying the current peaks most quickly plus a 1uF type for shifting all remaining AC fractions and interference into oblivion...? ;-) )
Regarding PSUs for P24 and position of the switches: Ok, I'm afraid I need to answer that again a bit too detailed... ;-)
Concept/layout:
- I haven't decided yet what I finally will use for powering up the system before switching over, the Antec or the picoPSU.
- You are right, when I use the picoPSU, I plug it into a P24 extension cable to get all the required lines "out of" the picoPSU.
- Once the system is completely powered up, I switch of all "special" lines (-12V, Power_Good, 5V_standby) with a 3-pole toggle switch right "before" the P24 connector.
- The capacitance detailed above is also sitting "before" the P24 connectors, the 1uF cap being the last in line and connecting all red / yellow lines for the 5V / 12V rail with its positive pole and a number of (in my case 4 for the 5V rail and 2 for 12V) black ones with its negative - you get the picture?
- 3.3V is different - I use a pack of 5 LiFePo batteries that supply exactly 3.3V for a long time and at an internal resistance of 2mOhm - no need for any regulators and caps here, direct is best possible :-) (I use an identical one to power Juli@ as per Alfred/sonics)
Switches:
- The switches for 5v and 12V are "before" the caps, i.e. between PSU and caps, thus their interference is irrelevant (I believe).
- Well, there is a switch between the battery pack for 3.3V and the P24 pins... Up to now it's still an ordinary toggle switch, but will upgrade to some "audiophile" switch (like for choosing source in a pre-amp, not decided yet which, internal resistance is the key)
- Another upgrade to come is getting rid of the P24 extension cable and soldering low resistance cables (thick ones with many wires for speed plus copper mesh for keeping HF interference out) directly onto the caps on one side and into gold-plated Molex P24 pins (got them thru Digikey) on the other. Well, you can't get too extreme, can you? ;-)
- All in all, this allows a very "pure" environment when everything is switched over or off - decent capacitance supplying clean power as quickly as possible to the various lines completely separated, fed by exactly regulated and also quite clean voltage from batteries. Probably sounds like overkill for most, but since current and voltage is basically everything and the starting point for any signal, it's my preferred point of interest, and in audiophile terms it absolutely pays for me :-)
Ah well, your thoughts about the number of linear PSUs to power pico: Well, you could get along with 3 linears for 3.3V, 5V and 12V, using the 12V one also to power the pico from the start and switch off the pico when you have switched everything over and the pico is idle then. But remember to size this 12V linear right - the 12V line only draws <200mA but needs to provide 4-6A while powering everything through the pico... I (of course) use a seperate battery... ;-) But surely that's not necessary.
Charging the batteries:
- I disconnect the 3.3V LiFePo packs and charge them elsewhere. not a hassle for me, once charged the last for days so that's a nonissue for me
- All the other batteries are in fact charged in situ (I like that) by manually connecting the required chargers. Not an elegant concept, I know, but I like to have the chargers completely out of the way for "serious" listening sessions, I don't have too much time for them unfortunately anyway (1-3hrs every second day on average) so again for me that's pretty manageable - and my "pure extremist" approach (no switch is the best switch) is supported :-)
BTW: Breaking out i2s from Juli@ to DAC - interesting! What DAC are you using? I'm going Buffalo32S at the moment... ;-)
Best,
Robert
Robert - I will digest your words over a glass of wine this evening - but as a quick response if you look for 'rebirth of the juli@' you will see all the details for the I2S breakout - as soon as I have done this (slightly differently as i will remove the Analogue board) I will take a photo and post.
I am also putting together the Buffalo 32s - hopefully I should have it in the system (using SPDIF) tonight or tomorrow. I am really looking forward to using this dac.
Alan
Robert,
I think I understand what you are doing and where, one question springs to mind - you say that you are using a 12v battery for the ATX24 which you switch to once the pc is up and running. Is this one SLA or is it several (is it even an SLA?) also (OK I cannot count) with the 3.3v you do not use any regulator as per sonics - what about the 12v?
Since I have a few spare Jensen / Black Gate / BHC capacitors I think I will first of all attack the setup as it is (i.e. no change of psu) and then look at the switching off / over of the supplies.
All of this will be after I have the Buffalo 32 settled in - all psus for that now up and running just a question of a few connections (and possibly putting it in a box).
Alan,not sure if I understand your question right - I probably was unclear on whereto I switch once the PC is up and running, right?
Let's assume for powering up the PC, I have the Antec Earthwatts connected to P24 (caps are in between on 5V and 12V lines). Well, "connected" isn't really precisely what it is - every line currently is going through a switch, two switches in total, one for the "special" lines (-12V, Power_Good and 5V_standby) and another one for 3.3V, 5V and 12V and the respective GND wires.
When powering up, all switches are so that every cable coming from the Antec has connection to its Pin on P24. When everything is up and running, I switch off the "special" lines. I then switch over the three main lines to their battery sources:
- For 3.3V I have a LiFePo battery pack (5 cells) directly connected, i.e. the wire from the switch that connects to the orange cables is connected to the positive pole, another wire that is now connected via the switch to two black lines goes to the negative pole of the pack. No regulators in between, no nothing - the pack is charged at exactly 3.3V and has a capacity of 2.3Ah so that they can keep exactly this voltage for a long time (due to the quite low current draw on this line of ~ 300mA) and give full advantage of their low internal resistance.
- For 12V I use a set of three preconfigured NiMH packs at 6V each (just so easy to handle and charge, of course you could use any 6V batteries that you prefer), so 18V in total, and feed them into a voltage regulator to bring them at exactly 12V. I know that most probably consider that as overkill, you usually use a 12V battery for a 12V circuit, don't you - but a 12v battery almost never is at exactly 12.00V, whereas I prefer to feed the digital circuits on the mainboard with an exact and always identical and reproducable voltage of 12.00V. (BTW I use the same setup for P4 to get to exactly 11.93V which has been reported to be the best sounding voltage by Bernd and Theo and others). Again, positive wire from switch to positive output of the regulator, negative wire to negative output.
- For 5V I currently use 4 8V VRLA Pb batteries (can't use the NiMH ones or other "weak" batteries because for 4-6A and regulated you need really strong ones, and these are). Each goes through one regulator (same as above) each to get to exatcly 5.00V. Again positive wires to positive pole etc.
Is this explanation halfways clear...? I promise to take a few pictures and post them - once I tidy up the still very experimental mess... ;-) That'll take a few weeks.
And you're right - what I described (except from the 3.3V line) all takes place "before" the capacitors, so they are really the most important part re quick and clean supply of decent power to the demanding circuits. By adding the spare caps you'll most likely will get substantial improvements already.
Cheers,
Robert
Edits: 05/28/09 05/28/09 05/28/09
Bertel,
I am trying to put together a diagrammatic representation of the various psu mods using batteries and switches that you have posted - I always find this helps me understand things. I will send you a copy before posting - to make sure I have understood.
I am trying to source the batteries and I thought I had a reasonably priced source for the LiFePo4 batts until I checked and one of them is dead! So after much searching through all the Radio Controlled Planes/Helicopters I think I found some decent ones - Graupner / Racing batteries at http://www.acemodel.co.uk/batteries/a123-lifes04-/1100mah-3-3v-single-cell/prod_199.html although they describe them as LiFeSo4!Are these OK? If not where do you get yours?
I can find many sources of 6V NiMh packs - but the only source for 8v VRLA batteries is China! Given that this is supplying the 5v work-horse line then the other styles of batteries do not have enough power. I suppose I could always use one of those Marine batteries - big enough to run a boat!
I have to say that searching for batteries has been quite an experience - I might just take up Radio-controlled Planes as a hobby - well it would get me out in the fresh air.
Any help/sources would be appreciated.
Alan
![]()
Alan,
Radio Controlled electric airplanes and helicopters is my other main hobby, so maybe I can help. Where are you located?
Greg in Mississippi
Greg,
I am in the UK in a town called Farnham (Surrey); I made my comment about Helicopters and Planes in jest - however I have now started looking at the various planes available. It seems that there is quite a selection from ones to use indoors to hi-performance outdoors types.
Near where I work is a very large playing field owned by the Army and there are quite often people flying planes - despite being close to one of the largest Army Barracks in the UK!
Regards
Alan
look at that handsome guy. that is a big plane!
Robert,
Thanks for the clarifications - I will spend some time now planning what to do - I have quite a lot of 6v SLA's sitting around not doing a lot, same as the capacitors; just need a few more regulators.
On the Buffalo side I have been wondering about how to put it into the PC without any interference - this would make the I2S leads as short as possible.
Photos would be great - but I would not like anyone to see my PC as it is - just a heap of wires - transformers open to the world - so when yours is all neat and tidy please publish.
Alan
how difficult was the buffalo32 to assemble? which ps did you get for it? nice.
theob,
The Buffalo 32 board itself comes built/tested all that is needed is to solder in the connectors - left out just in case you want to hard-wire it.
I bought three psu's so that I could drive each analogue channel separately - they come with all the components identified - there are documents on the website as to the board layout etc - very easy to build, the only thing that has to be done is to twiddle the 20-turn trimmers to get the voltages set as required.
All in all - very easy - hopefully it all works when I connect them tonight.
I am very excited for you--having ability to go I2S out to the Buffalo. Please let me know how it sounds. I am very interested in going this route too, just waiting for better time to finance it (2 trips to finance 1st).
I bet it is so.
I think Mr. Nugent is one of the important voices in the world of PC AUDIO. I do not own any of his devices but I am always looking over his shoulder. I am always interested in what he is working on. He is committed to the USB and we, mostly, are not. Nonetheless, no one has done more tinkering and thinking than he has.
Look at this thread:
http://www.audioasylum.com/forums/pcaudio/messages/5/52276.html
Nothing like improvements at the absolute beginning of the chain to make the biggest difference.
I went ahead and bought the package of drive and software instead of seeing what I get from the ebay seller. For an extra eighty five dollars I like knowing I am getting what the software writer intends.
If it is worth it looks like I will have to re-load all of my files, but if it sounds better it will be worth it.
I doubt this is going to romance me away from the turntable but it could make a significant difference.
Bye,
Rick McInnis
Hey Rick - I'm not trying to diminish what Steve is recommending there, but I just want to point out a couple other things of note. When Spoon (of dBpoweramp) chose the Teac drive as the best one for the RipNAS, he was not saying it is the best drive period. His criteria was to find a drive that was an IDE laptop drive with slot-loading. So of those drives, this Teac was the best. However there are desktop drives that are just as capable. I have a Plextor Premium and Plextor 760 which he said was equivalent to the Teac.
Also, in Steve's comparison when comparing the different Led Zeppelin rips, he said they had different checksums, so of course they sounded different. And apparently to him the Teac results sounded better. But let me just warn you that simply re-ripping your music with a different drive is not going to necessarily be a jump in SQ. If you have damaged CDs (as I suspect Steve's LZ disc is), then the Teac may give you better accuracy (and different checksums) as your current rips (and indeed sound different). But just to share my experience, my CDs are flawless, no fingerprints or scratches or anything, and I have compared some rips between EAC and dBPoweramp and on different drives and they end up matching AccurateRip and have identical checksums - and they sound exactly the same.
I'm not trying to come off as a skeptic here. YMMV. But if you already have a decent desktop drive and have correctly used dBPoweramp, then you may not be gaining too much with the Teac (except perhaps to better rip damaged CDs).
Edward,
If only everyone commented the way you do. Of course, the apologies are appreciated but unnecessary when one makes reasonable points in a reasonable way.
I do not doubt I probably jumped too quickly. Nonetheless, I was needing a new CD R drive so at worst it is not as if I have bought something I do not need. Maybe I did pay too much ...
I will see if I can tell any difference and report my findings. We are all too anxious to find the golden goose of digital audio. I was not expecting a dramatic change just hoping for another increment but how great it would be find something that would make a leap similar to cMP!!! Even one-quarter as much ...
Thanks for tempering my intemperate enthusiasm.
Rick McInnis
... but I keep getting a 'not enough ram error' in cplay. Before I put my old kingston 1 gb ram back in is there some criterion for determining not enough ram? I got one 6 minute file converted that plays ok but that took 9 loadings. Is 10 the limit?
.
...and once I got the hang of it the sox upsampler is not bad. It's not on par with SRC but it's good. Also I found that I was more successful converting to flac files than wavpcm-> wav. But let me say sox is good, very good, but I would have to say that SRC is really a lot better in upper mids, highs. There is more texture, nuance, details in a full orchestra with SRC. With sox the same piece is smoother, warm but an almost 'cover-up' of details occurs. Some may prefer sox to src because it is warmer (and I know those who are willing to throw the baby out with the bathwater but I'll take the baby and bathweater every time).All in all though I prefer sox at 192 versus src at 96. A very good choice as an alternative to the metallic(in my case) src.
Question though on the sox parameters how do phase choices (other than I) affect sound? Just curious if anyone has tried.
Edits: 05/20/09
I have listened to several files with both M (minimum) and I (intermediate) phase called out in the sox conversion command and I believe I prefer M. With M the resulting sonics are closer to SRC wrt to getting the initial ictus or impulse of the sound right. Plucked strings, fast percusion and imaging within a large orchestra sounds better or more realistic to me. Reading about minimum or linear phase suggests that minimum has less pre ringing but more post ringing in a asymmetrical way. Linear phase is more balanced wrt to pre and post ringing but has more pre ringing ( than M) which is a bad thing imo. I is in the middle but I think pre ringing is the more important parameter.Has anybody else tried this?
Edits: 05/21/09
theob,
I have not tried it but I will give it a spin over the weekend - very interesting findings. Have fiddled about with any of the other settings?
right now with sox I use gain -1 (or -2) rate -v -s -M 192k and I put data in flac files.
with cplay I use 192 rate, dsp buffer at small. I'm still running juli@ analogue outs.
I'm really liking sox-ed files at gain -1 (or -2) rate -v -s -M 192k. when I go back to cplay 96 rate 146 upsampler and listen to original 44khz files it sounds more etchy to me in the highs. maybe I'm getting used to sox-ed files. Its very nice but if I continue to convert my files at my current rate I'll be needing another 2.5" drive soon.
Obtw when running flac 192 files I'm getting under 10% cpu usage (key to avoiding metallics).
Hey Theob,
Forgive me for jumping in and not reading alot of the prior posts, but it seems like this is not a fair comparison as you are really comparing an offline upsampler against a realtime one.
It would be interesting if cplay would allow for creating files using src, or comparing sox when it is integrated for realtime upsampling.
probably not a fair comparison but a real either/or for me. are you saying one is intrinsically better than the other?
As you've noticed, offline upsampling is much less CPU intensive.
YES, and IIRC that might be a big deal with Theo and his brightness.
Anyhow, it would be nice to do a fair comparison either way.
that is sounding really good - looks like you are back to enjoying the music!
:)
A Cautionary Tale :
Hi All (...and especially you Juli@-with-BNC-sockets-added folks)
Last fall my Juli@ underwent a modest transformation, as can be seen here -> http://www.audioasylum.com/forums/pcaudio/messages/3/39956.html
Having compared the S/PDIF out from both connections - new BNCs vs the stock breakout cable - and declaring the BNCs a clear winner (...is that a pun ?), I turned my attention to other things.
The final step of that listening test was completed only last night ... ie: remove the breakout cable !! - Duhhh !!
I was listening with my teenage daughter, Jane (no... not her real name) and we both instantly heard the improvement... and she is not encumbered with "audiophile" filters to cloud or colour her perception.
Jane hasn't heard the cMP since the release of B22, and it's become a tradition to celebrate a new version of cPlay with a performance of Tubular Bells II from Mike Oldfield. Every time there is a new cPlay, we come to know and appreciate this recording better.
(1992: WEA CD 90618 - producers: Trevor Horn/ Mike Oldfield/ Tom Newman. One of the most stunning examples of Studiocraft, in the service of Music, that I know !!)
We began with B22-@-121/192KHz (my P4 can't do 144 SNR), just to calibrate our ears. Track 08 - Weightless ... very nice.
Next up was B25-@-121/192KHz (same track). MUCH better... remarkable that TB II still has more secrets to be revealed.
OK, with so much opinion on both sides of the NOS 44.1 vs 192 upsampling, we listened again (B25-@-121/44.1KHz).
Nope... not in this system. Sixty seconds into the song, Jane turns to me making a sour-puss face ("I don't like it"). Neither did I. Missing in action was the incredible sense of width & depth - confined now to the space between the speakers (and no longer reaching back into the neighbour's house !!). Gone was the sparkle of the HF, the minute dynamic contrast of some percussive lines, and the separation of instrumental voices we heard before.
Quick !! back to 192... and The Magic returned. At this point I noticed the pigtail still hanging out of the Juli@. Hmmm... "OK Jane, just one more test before we get to hear the whole thing" (remember the impatience of youth...).
AGAIN... within the first sixty seconds of play, the benefit of losing the breakout cable was plain to hear. Jane said the "different layers" were more distinct, and that all the parts "were easier to hear". The "audiophile ears" agree.
Yes, this simple tweak should have been made back in the fall... but in a way I'm glad it happened like this. It's another lesson in support of the old maxim: EVERYTHING MAKES A DIFFERENCE.
Cheers,
Grant
That's not a Toy... IT'S A TOOL !!
Uh-huh. Just "left it there"... of course "there" being in a corner on the floor where you never go unless something's fallen there - or broken.
Out of sight, out of mind. And the Mind is too full of other noise ! Gee, it's like getting a new version of cPlay for free... er, well... ahead of the rest of the group. Or am I just catching up ? Yes, THAT's what the fuss is all about. Well done and three Loud, Appreciative Cheers for version B25 !!!
Better quit now.
That's not a Toy... IT'S A TOOL !!
Grant,
Are you saying you had connected the BNC to the board, YET, had left the breakout cable attached?
I can only assume this is the case, yet, I am amazed you left it "there" for so long! Yet, I know the feeling, who has not done this same thing in the audio eccentric's feverish must listen to it now after all of this work syndrome. Just leaves the pleasure of hearing another veil lifted. Not in the ARABIAN NIGHTS sense ...
Still trying to muster the courage to get a copper mat.
Bye,
Rick McInnis
Hello cics and cplay users,
First of all, let me say that cplay is the best audio player I know.
Thanks cics for all your hard work.
That said, there is one problematic side to cplay, the visual side.
Even a slight aesthetic improvement without a complete gui makeover
will make a big difference. Since I understand cics is very busy, I have
found a software "Resource Hacker" :
http://www.angusj.com/resourcehacker/
that allows one to change icon buttons inside executable files.
Tiny step, but, one step for man ... :)
Anyway, cics, if you need any help I am willing to contribute my
programming skills and little time I have for this great project.
.
http://en.wikipedia.org/wiki/Hidden_Markov_model
As a way to improve upsampling, I was wondering lately about using similar
patterns in the track, to thicken the original sample rate.
Here are some small steps even more worth taking:
1. In cMP, show the samplerate of each cuesheet so one knows whether to change the cPlay playback rate: either by showing the number or using different colors for different rates in the cuesheet names
2 In both or either cMP or cPlay, allow enlarging the window or decreasing the font size to allow more of the album, title or cue title to display
3. In cMP, since the artist displays already in "All", substitute or provide option for Composer display instead of "Artist"
Dear Theo and others who have heard Theo lament his metallic affliction,
I heard it today.
I installed 25 today in the SSE4 version since cics seemed to say this was his best work. I have never installed this version (instruction set) ever before.
At the same time I am trying to lower the Vcore voltage and probably too quickly since the machine has not been running for six weeks previous to yesterday. I, of course, have disabled EIST. Did that yesterday and was able to run at .80 volts without a problem.
Today, I took the voltage down to 0.7625 (or something like that) and I get the blue screen. I turn the computer off and attempt to re-start and there is nothing. I re-set BIOS and incrementally increase the voltage until the machine will run. Something like 0.80 and request a selection and I hear the sound he has been speaking of. It is horrible.
I re-start and increase the voltage to 0.825 and ask for the same selection (all within "start cMP") and it is back to normal until I accidentally hit a button on the mouse and it begins again. I re-start and all seems to be fine so far. Though I did not try to antagonize the machine. I will let it run constantly for the next few days and see if it settles down.
NOW, I am wondering if it has something to do with SE4? Does changing the instruction set take some getting used to for the CPU? Excuse my processor mysticism, but I have never had this happen before.
I cannot remember all of the details of Theo's travails, but I wonder if there is any connection with this?
Could disabling EIST have anything to do with it? Since I have just done those two things, they are the differences in a previously trouble-free set-up I cannot help but wonder. Again, one wonders if a system gets used to a certain set-up and rebels when "these things" happen? Of course, it sounds ridiculous to me too but in the world of audio, it seems, anything can happen. When little atomic squiggles can produce astonishing sound what is one to say?
More than anything I wanted to assure Theo he is not damned or cursed or any of that other stuff OR I am, also.
Other than that, I do find 25 SSE4 to be very attractive and superior to the sound I was getting last night with the previous iteration and the other instruction set, even though, in my case, it required a higher voltage. Still very little information extending beyond the outside of the loudspeakers. Within, it is very good. Like a loving microscope on the recording. Tremendous detail/precision in imaging within the scope of the lens. This is sound one could easily live with as long as they do not ever listen to a better turntable. blah, blah, blah. I know ...
It has occurred to me that cics has never told us what this mysterious abbreviation stands for. Dear cics, any chance of letting us in on the secret?
As always my best wishes to cics and his family who must be given credit for allowing him to tinker and tweak this amazing device so all of us may enjoy listening to the CD.
And best of luck to Theo for keeping the metallic wolves at bay.
Rick McInnis
thanks rick. I began going metallic with cplay 2.0 15 and above, only exception was 18 which never went metallic. Eist on or off never mattered. Going lower on voltage and speed seems to have an effect. I can only go to .85 volts (set in bios) which reads .80 something in cpuz. When I go lower than host clock control of 157 I get instabilities. And of course running 44 sample rate is no problem. I seemed to think it was related to the revisions associated with asio 'refinements or changes' (I'm saying associated with not caused by).
Also playing music files off of flash drives helped a lot. I believe it is somehow associated with power supply or power line noise. I guess if I ever go to ssd's or more battery produced power supplies will that probably help too.
Anyway I have concluded that 44 is better sonically in my system than 192 so its a moot point now. But if I could get sonics of 25 with non metallic performance of 18 I would be a happy enough camper to try 192 again.
Theo and all,
I can play through entire albums and nothing goes wrong. I can stop a selected album and ask for a new one and all is fine.
It seemed to happen (and it was only twice) when I accidentally hit one of the mouse buttons. After I re-started all was fine.
I think it can only be because we are running our machines at such a low level. There is really nothing else that could explain it. Considering the above it would be hard to believe there is any stress on the power supplies. My ANTEC must be loafing along. If anything one wonders if there could be a problem from the PS running below the level it was designed to handle.
Since I am needing a new MD for the home computer I will try the board recommended by abysstw.
As far as the upsampler is concerned, for many of us, including myself, if the computer does not do the upsampling your DAC chip will. Most of the current available devices contain these chips. Not using the upsampler just passes the task to the internal DAC which, in my case, does NOT do as good a job.
Those having DACs with chips that do not do this could very well experience better sound without the upsampler, but I have no idea.
When this happens it sounds so awful I figure if we could find a way to make it repeatable at will it would become the next rap sound sensation only to find its way across the spectrum of POP music and then Phillip Glass could attempt to formalize it and we would ultimately hear it in concert halls around the world.
Don't say I didn't warn you.
Bye,
Rick McInnis
Hi theob,
You may want to try out the GA-EG45M-UD2H.. The sound is beautiful, and my metallics problem completely went away after I replaced my old GA-G31M-S2L with the new board. I'm able to go down to .75v without any resets, hiccups, glitches or whatsoever. With my old motherboard, going below .85v was not possible without allowing the machine to "warm up" for at least a couple hours.
outstanding I went thru the review of this mobo (link shown else where under pc audio) and this thing looks like the real deal. could you get all the bios settings in cics specs (plus obviously the new ones associated with voltage, speed and timings)? how low a host clock control did you get?
do you know if cics is going to try this?
it looks very promising!!!
The BIOS of the EG45M is quite flexible. It offers some new settings, but a few of the old ones are missing.
I'm not sure whether cics plans on trying it personally, but he seems to think highly of the new motherboard (see cPlay 2.0b25 release post and link below). Greg (GStew) says that his GA-EG45M-UD2H has already arrived, so you may want to hear his thoughts first.
Even with my current (probably non-optimal) BIOS settings, the new setup sounds significantly better than my old one with the recommended settings applied.
If you do buy an EG45M, please spend some time trying out the various settings, and help the community realize the MB's full potential.
I'm still in the process of loading & customizing it's setup... I did discover that you cannot activate Windows XP AFTER you've done most of the customizations (through the autoruns disables and eliminations), so I started to reload Windows XP again 9pm last night.
I did listen to it partly optimized in a garden-variety case with a stock non-optimal PS and a not-separately-powered Juli@ for awhile yesterday afternoon. The sound was promisingly-good. I hope to have it fully setup by Wednesday evening.
One trick I tried that worked out well was to disable things like the on-board network card BEFORE loading Win XP... that way it doesn't find them during the installation, doesn't load the drivers, and at worse reduces the number of things you need to remove later... and at best makes for an overall cleaner install.
don't understand, hanssatink said that he did not have to reload windows. pls say more about this.
And while the new motherboard did function without a new load of WinXP, I figured that the cleanest and likely best-sounding setup would be with a new, fresh, and clean install.
With the trick of disabling un-needed devices in the BIOS before loading Windows, I'm getting an even cleaner install than my 1st one. Plus with the experience of loading a cMP the first time behind me, this time is going rather quickly.
So you can use it without a fresh install and re-customization... but for the cleanest installation and the greatest probability of the best sound quality (and lack of problems, like your 'metallics' issue), a fresh install is best.
Greg in Mississippi
I'll probably do a clean install and customization just to be sure I'm getting the best possible SQ. I also plan to further reduce the windows footprint with XPLite this time around.
thanks for explanation
Isn't it the real world that windows only loads device drivers wich are presented by the hardware. Example; when i disable the onboard soundcard, usb or lan in the bios, windows won't load the drivers after a restart. Sound quality is imho not influenced by the size of windows installation, but the sum of processes running. Enable/disable hardware before or after installation make no sense i suppose.
Honestly, I don't not really sure that's the only impact of non-used drivers. If they show up in Windows somewhere (like in the Device Manager), they at least add to the size of the Registry. Other impacts... I don't know, I'd be interested in input from those more versed in Windows internals than I am.
One thing I have learned from doing this project is that there is a lot about Windows that I don't know and that things I didn't think made a difference do. So I'm playing it safe by making my system as clean as possible.
Greg in Mississippi
did you have any trouble fitting everything in? are you using the thermalright fanless cooler? It looks like it could be a tight fit.
No trouble at all. I use the Coolermaster Gemini (see pic). Not sure whether the thermalright fanless cooler will fit. Can you post the dimensions of your cooler?
![]()
Thermalright SI-128 SE Main Features:
Technical Specifications
Model number: SI-128 SE
Dimensions: 125 x 145 x 91.5mm (LxWxH)
Materials: Copper base, 8mm copper heat pipes, and aluminum fins
Finish: Nickel plated
Weight: 510g (heatsink only)
Intel support: Socket LGA775
AMD support: Socket AM2
Recommended fan: all 120mm
The gemini seems to take up alot more space than the SI-128SE, so I think you'll be fine. Not 100% sure though.
Oh - I am glad someone else is getting the metallics - not really - but it is good to know that it is not something that we are all doing wrong.
I know that on my setup the metallics are only at 192khz and it always starts as cPlay is reading the next track. Since I have my hard drives on a separate psu it can only be either a fight between cPlay and Juli@ for processor time or for power.
I will try the new mobo but I want to experiment a bit more with the Juli@ sound card psu (Sonics battery mods) to eliminate the possible psu conflict. I have also wondered about the screen and keyboard being permanently attached (especially the screen as it is also being updated at the start of each track and I have noticed it skip the display of a second or two - e.g. it jumps from 0.02 to 0.4)
Having said all that I am currently running at 96khz very happily.
thats fairly similar to my experience.
Refer to ReadMe.txt for important details & specifications.
Change Log (2.0b25 release):
- ASIO output interface refinement
- Minor DSP refinements
- Fix for soundcards using latencies of odd samples (rare)
No further releases are planned.
Please REMOVE previous versions before installing cPlay 2.0b25. Normal (SSE2), SSSE3 B9 and SSE4 B9 versions are available. Running cPlay on a CPU not supporting the required instruction set will cause cPlay to exit immediately.
For cMP² users, review these BIOS changes (underclocks FSB, CPU & disable additional USB items). New settings are in effect allowing for lower power consumption (below 20W). Recommended Gigabyte mobo GA-G31M-S2L is no longer available, use either GA-G31M-S2C, GA-G31M-ES2L or GA-EG45M-UD2H ( advanced settings ).
Tried SoundForge Seattle for V25 and all downloads were corrupt, switched to San Jose-success.
WOW, what a difference between V22 and V25. Soundstage is immense!! Bass somewhat subdued but still there.
Listening to Alan Parsons and can really, really hear the difference.
I recently made big changes in my Cmp machine: Linear PS, Pico PS, filtering caps, Hammond L193 and a Greg Stew modified Juli@ so changes weren't subtle. But still going from V22 to V25 is stellar!!
RayBan
Can't agree more with the magnificence of 2.0b25!
Did you modify the Pico itself, by chance? I found that replacing the stock caps with a BG NX 1000uf/25v improved SQ noticeably. Adding "output filter caps" to the 3.3v and 5v rails seems to also have a significant effect on SQ.
I added 470 Oscon's.
Having made so many changes at once (not recommending anyone do so) masks the ability to judge which had the most (any) impact either way.
RayBan
Hi Cics,
When I was building my cmp2 rig, I got to a point where there was a dramatic improvement and a 3d flesh to things. But I made more changes and some where along the way that improvement was lost a bit.
But v25 brought it all back.
This is a very big leap forward.
Nice job!
Hi Cics,
So I have settled on b25 SSE4 as well. Needless to say the resolution of B25 SSE4 is the highest, surpassing wavelab, Samplitude and Saw Studio, Musicality can be adjusted using DSP buffer. Again if resolution is your objective of cplay then I think you have done 300%. I support the idea of taking a rest for the moment. After all it's been a full year of hard work.
I am very very satisfied with my current system. I will go back to study how to write a playlist program.
(FYI My PC/Audio configuration can be found at http://canhtpcbeatcd.blogspot.com)
Thanks again cics for a wonderful year of surprises after surprises. Greetings from Hong Kong.
First I downloaded both cPlay 2.0b25 and cMP 1.2. Then I installed cPlay just to make sure it worked like the earlier version 2.0b22 I briefly tried. Needless to say I was amazed at how well version 2.0b25 sounds, way better than anything else I've heard.
Now the rub is I've never used cMP before and I wasn't sure whether you install cMP before cPlay or vice versa. I tried cMP first and found that I had few computer controls and I had to unplug the computer to even reboot. So then I install cPlay first and then cMP which seems to work but I'm not sure if that's the correct sequence. I thought I read "The art of building Computer Transports" but I must be tired since the document wasn't clear and I thought it to be out of date with a focus on foobar and not cPlay.
Please forgive my eagerness to achieve quick success but I'm excited and would be ecstatic if I knew the correct sequence for cPlay, cMP and how to bring a starter library of about 300 WAV and FLAC files into cPlay so I can relax and enjoy the rest of this weekend.
Z
cMP's user manual explains all the operational details needed (see Installation & User Manual (18 pages) - be sure to select the second PDF document). Sequence (easiest):
- Install Soundcard & ASIO drivers
- Install cPlay - test
- Install cMP - test in XP Mode thereafter switch to cMP Mode
cMP works on video resolutions of 1027x768 or higher. Pressing 'X' exits cMP (there's different behaviour depending on mode you're in - all explained in manual).
I had pretty much given up on ever seeing a response. So I did some exploring, found what I needed from earlier posts and worked things out to my satisfaction. The biggest impediment to the installation was actually finding the Installation & Users Manual, since as some earlier posts indicated that links to it were broken. Even though I eventually found it on my own, thanks for re-posting the link with an explanation for others that may have similar difficulties.
I believe any perceived difficulty is somewhat of a good thing because it forces you to learn. Similar to other DIY efforts where one's own sweat and trials and tribulations can lead to greater rewards than what one could ordinarily find or afford in commercial audio.
The biggest problem I have so far is with the cMP and cPlay websites. Its just too slow to navigate quickly. If there is a better way please let me know.
The other somewhat minor problem I have is the apparent need to re-rip my CDs that were ripped to the AIFF format for iTunes on a Mac since cPlay does not support AIFF. Thus I am using EAC to rip to WAV files since I also need to create cue sheets for cPlay. In the meantime I can still use cMP and MediaMonkey to play all my AIFF, FLAC and WAV files on my XP computer music server or just use cMP/cPlay to play the few dozen CDs I've ripped so far.
Use the Windows exe version (Mac is experimental).
Command would be something like this:
sox "input.aiff" "output.wavpcm"
Then rename the output file to "output.wav". Thereafter, use Alan's (AudioAl) playlist utility to create cue sheets.
and there was me thinking theob was having a laugh (Men In Black??) with his 'best of the best' - nope - no joke - this is astonishingly good. Running at 96khz I am getting music which is hard to describe (to do justice to it) - at the moment I am listening to a Janacek String Quartet, The Kreutzer Sonata, the power and emotion is unbelievable, the bows hitting the strings is visceral. Earlier I listened to John Lill playing Beethoven Sonatas not only with power but with delicacy; Eleanor McEvoy was standing a few feet away and the bass playing was superb.
Like all the others I can only thank cics for his hard-work and his willingness to share his findings with us all - very unselfish and generous despite our many 'complaints'.
Have a good 'rest'.
cPlay 2.0b25 is the most delicate/detailed w/o hardness mids to upper mids I have ever heard period! I am listening to J P Rampal's Japenese Melodies for Flute and Harp and I am thunder struck. This is beatiful!!Clearly best of the best.
Also I continue to try different dsp buffer settings and I have gravitated back to auto. But if I have something that sounds a little hard or etchy I reset dsp buffer to small and it is greatly improved (I'm talking 16/44 files). But if small buffer size is used where it should not be then bass, depth of field perception and bass dynamics suffer. When I use the small setting to improve a marginal file I simply raise volume 1.5 db or so and it brings back some of the losses in bass and dynamics.
But I also note (per appleteapot) that the small setting is overwhelmingly beatiful on female voices (try some of channel classic pieces).
Again thanks cics!!!
Btw I'm using sse4.
Many call hobbies pastimes. I have been listening to cPlay 2.0b25 for over an hour and it seems like less than 10 minutes. I can only state that one: I have the right past time and two: it is enhanced when listening to cPlay 2.0b25.
Edits: 05/08/09
Reason: SQ in B25 will be difficult to improve on, especially SSE4 B9 version.
That is very good to hear! You do deserve a rest--your recent burst of creativity has been dazzling. I'm still trying to get a handle on 25 (so far its all been good).
Edits: 05/07/09
Thank You, Cics! It is a wonderful player and it is a wonderful hobby. I think I understand You.
Serge.
P.S. Hardware will change and times will be different. There is always new morning after night.
I think we’ve been spoiled with cPlay, at least, I perceive it to be the most detailed and clear sounding player (in my system and environment).
I would also like to thank for all the work U are doing and also express my anguish if the development of this outstanding player is over.
I’ll be using it for some time to come!
Thank you for all you have done for me in terms of helping me build cmp^2. You are a scholar, a humanitarian and a great audiophile. Please don't keep away from this website. Your comments are always brilliant.
(Please say it ain't so).
The following are my current settings.
1) CPU Vcore at 0.75v
2) CPU Termination Voltage at 1.05v
3) CPU Reference Voltage at 0.669v
The .703v in my previous post was a typo.
I'm sorry to hear that there are no further planned releases of cPlay. Then again, most of us didn't expected another 25 releases after cPlay 1.4. So we can always hope, right?
Just looked at the GB manual.
![]()
![]()
Suggested settings:
- CPU Clock Ratio=6x
- Fine CPU Clock Ratio=0.0
- CPU Host Frequency=150 [CPU-Z to show core speed of 900MHz]
- MCH Frequency Latch=150
- System Memory Multiplier=2x [CPU-Z to show Memory speed at 150MHz, 1:1 ratio]
- DRAM Timings Selectable=Auto [CPU-Z to 3-3-3-?]
- CPU Vcore=0.85000V or lower [Make sure "CPU EIST Function" is disabled in Advanced BIOS Settings]
- CPU Termination=1.05V (or lower if available - this is the same as FSB "De-overvoltage" at -0.15V)
- CPU Reference=trial n error (this replaces GTLREF)
- MCH Core=lowest possible setting (the northbridge voltage?)
- DRAM Voltage=1.800V (or if possible, lower - test this)
- CPU Clock Drive=700mV (found under Advanced Clock Control)
- Other: Disable PAVP Mode (found under Advanced Chipset Features)
cics,
Thanks for the suggestions. I tried these settings today and here's my results:
CPU Clock Ratio=6x - Same as my previous settings
Fine CPU Clock Ratio=0.0 - Not changable, but automatically set to 0.0
CPU Host Frequency=150 [CPU-Z to show core speed of 900MHz] - I tried this and went back and forth a few times between 150 and my previous setting of 133. I marginally preferred 133 with it providing a bit more midrange presence and body, but need to spend more time comparing the two with lowered CPU Vcore voltages.
MCH Frequency Latch=150 - The only options available are 266Mhz, 200Mhz, 333Mhz, 400Mhz, Auto. I'm currently using Auto. When I used 200Mhz before, 2.66C was the lowest System Memory Multiplier available and it did not achieve a FSB:DRAM ratio of 1:1. When I tested it before, I preferred the Auto/2.00B setup over the 200Mhz/2.66c setup.
System Memory Multiplier=2x [CPU-Z to show Memory speed at 150MHz, 1:1 ratio] Options are Auto, 2.50A, 3.00A, 2.00D & B, 2.40B, 2.66C, 3.33C, 4.00C, I have set to 2.00B
DRAM Timings Selectable=Auto [CPU-Z to 3-3-3-?] - I'm using 3-3-3-5 with Kingston ValueRAM
CPU Vcore=0.85000V or lower [Make sure "CPU EIST Function" is disabled in Advanced BIOS Settings] - I had previously disabled "CPU EIST" and gotten down to .75625v. I went down three more steps over the day today, but got an interesting instability... the system locks up and I get a high-pitched squeal out of the speakers. Luckily, it hasn't blown a tweeter yet! I'm back up to .75000v now and checking to see if it stays stable there. BTW, I was not getting this before... either lowering the CPU Vcore further or setting the CPU Clock Drive down to 700mV caused this.
CPU Termination=1.05V (or lower if available - this is the same as FSB "De-overvoltage" at -0.15V) - I was already at 1.050V, the lowest available setting
CPU Reference=trial n error (this replaces GTLREF) - I was currently at 0.669V and tried 0.654V, the only lower setting available. I marginally preferred 0.669V, but I could not articulate why. Difference was subtle.
MCH Core=lowest possible setting (the northbridge voltage?) - I was already at 1.100V, the lowest available setting
DRAM Voltage=1.800V (or if possible, lower - test this) - I was already at 1.800V, the lowest available setting
CPU Clock Drive=700mV (found under Advanced Clock Control) - I changed this. I did not listen to the different it made, will have to go back and do that later. Also, if my instability behavior stays, I'll set this back to 800mV to see if it fixes it.
Other: Disable PAVP Mode (found under Advanced Chipset Features) - I had disabled this when first setup this board. I also set DVMT Memory Size to 128MB. the lowest available setting.
The major changes from my previous setup were setting CPU Host Frequency to 150 (from 133) and setting CPU Clock Drive down to 700mV. The Bios settings do not allow an MCH Frequency Latch of 150.
All of these settings still result in the following:
In the BIOS...
CPU Freq shown as 800Mhz normal, 133x6 current
Mem Freq shown as 533Mhz normal, 267Mhz current
And in CPUZ...
Core Speed 1192.6
Multiplier x6.0
Bus Speed 198.0
FSB 794.8
DRAM Freq 198.7
FSB:DRAM 1:1
3-3-3-5
That the CPUZ Core Speed is still shown at 1200 or so makes me think that there are still some limitations in the BIOS. I very interested in seeing what the updated BIOS Gigabyte is providing Abysstw will do. I'd also like to have the option to disable Spread Spectrum... These two things should give us the same level of flexibility we had with the GA-G31M-S2L BIOS.
Sonically, from memory, my impressions are that this board adds a nice sense of 'solidity' and micro-dynamics to the sound up and down the spectrum, but with the current settings, the treble and fine details were better with the GA-G31M-S2L. The GA-EG45M-UD2H is not bad, but the GA-G31M-S2L was magical for me. But this is based on long-time-recollection (three weeks!) which is notoriously faulty, I really need to put the GA-G31M-S2L back in the system and listen a bit to make a valid comparison.
So progress is being made... and many will prefer this Mobo the way it stands. It's still a mixed-bag in my book... but the things it does well are so good it's worth working with it!
Greg in Mississippi
Hello,
I just started building a memory player and I bought the GA-EG45M-UD2H.
I did a little testing yesterday with all the settings just to see if I could underclock everything with success.
Here are my BIOS settings, with my results in CPUz.
I started (just like GStew) with the settings listed for the GA-G31M-S2L : (see post below)
http://www.audioasylum.com/forums/pcaudio/messages/3/36343.html
Then, I whent in the MB Intelligent Tweaker and entered GStew's settings :(see post below) http://www.audioasylum.com/forums/pcaudio/messages/5/51752.html
At first, it didin't work (multiple boot) so I had to change somme settings in order to be able to boot :
CPU Host Frequency: 133 MHz
DRAM Timing Selectable: Manual
CAL Latency Time: 3
tRCD: 3
tRP: 3
tRAS: 6 (With my 512MB ValueRam, when I had tRAS to 5 the BIOS would not boot)
CPU Vcore: .9v (still testing this setting)
CPU Termination: 1.100v (still testing)
CPU Reference: 0.688v (still testing)
CPUz 1.49 tells me :
CoreV : 0.864V
Bus Speed : 196 MHz
Rated FSB : 785 MHz
MEMORY : 3-3-3-6 with Command Rate 2T
Just to let you guys know, I couldn't disable USB2.0 because my mouse wouldn't work without it (strange!)
I don't have the Juli@ plugged in yet and I don't have the Granite PSU's.
others : CPU E7200
psu : Antec EarthWatt 430
RAM : 512 MB ValueRam
BIOS Rev : R3
I'm not sure which setting will make a SQ improvement because I didin't configure windows/cMP/cPlay yet!
Thanks
Etienne
Etienne,
Two things...
First, the CPU VCore setting is one that is best to sneak up on... your .9 volt is a safe starting point. After you have your machine fully configured and running cMP, you can drop it by a step or two, boot the machine, run it for an hour or two, then if it seems stable, drop it by another step or two and repeat. I'm at .75625v now and at the edge of instability... I got some minor, occasional playback glitches when I first set it this low, now that it's been running a few days it's pretty stable and I'll try one more step down soon.
Second, I experienced the same issue with the USB mouse with USB 2.0 shut down in the BIOS. But after I loaded the Touchpad software for the Zalman touchscreen, I could shut down USB 2.0 and the mouse would work ok. If you have the recommended Zalman case with touchscreen, this is a good path forward. If not, I'm sure there is a driver that can be loaded to make it work, but I haven't spent time trying to track it down.
Good luck. Let us know how it works out!
Greg in Mississippi
P.S. Theob & all, I'll be posting a few impressions of the sonic differences soon. Now that I've got the new board as configured as I think it can be with the currently available BIOS settings and it's had a week or so to break in, I want to put in the old motherboard and listen a bit before publishing detailed impressions. But in short, it sounds very good in some ways, but I think additional BIOS options and/or configuration settings will be required to get the best performance out of it.
great cant wait to hear your sonic reaction vs other gigabyte mobo!
I went to set up the old board and found that my spare Thermalright cooler was a different model. I wanted to have the cooler already mounted to try and minimize the time and effort to swap between them.
I have a second cooler of the same type as the one I have alreay on order and it should be here Friday... hopefully I can do the comparison over the weekend.
Greg in Mississippi
P.S. I'll also post some comments on the latest BIOS settings suggestions from cics tonight.
thanks
Good post! I'm interested in your sonic reaction once you get fully onboard with cmp^2.
Vcore 0,75?
What processor do you have?
.
Since yesterday i use a GA-EG45M-UD2H too with E7400
The normal CPU core voltage is 1.24
I am now at 1.1 volt. Don't know yet what the lowest possible is. 0.75 doesn't work.
What is the normal voltage of the E7200?
Great Board by the way. It burning in at the moment :-)
How is it sounding? Any installation issues? What about support for a crt type display (looking at back panel it does not appear to have the normal multipin connector)?
Use this board with a E7400 @ 0,9 V and 7x multiplier (1,8 ghz)
The boxed intel cooler is very quiet. No need for expensive cooler. rest on the board is passive cooled.
Installation is simple. No jumpers at all. This board has a DVI and HDMI connector. Video quality is excellent. After HW installation i'll started up with previous XP on harddisk. I had luck. XP started and i could easily update chipset drivers and other drivers. No need to install a fresh XP and do all the tweaking again.
Bios settings are huge and complicating (for me) I followed the guide from Cic's. Most settings are equal. Disabled as most as possible to keep a running system. Maybe i'll find time to post all the settings. (seperate thread?)
The sound is very clear, open and airy with Cplay SSE4. Highs are little bit metallic, but out of the box not bad en very promissing. The system is now burning in for a least 48 hour.
One great benefit of this board is the self-repairing Bios. When the system doesn't boot anymore because of tweaking the bios, you don't have to reset the Bios. The pc starts a few times on itself and then booting again. So playing with the bios settings is easy.
I spent some time tonight working with the BIOS settings of my GA-EG45M-UD2H.The overall state of my cMP is that it is a fresh Win XP install (SP1) that is mostly customized/optimized through component removal, configuration, device & services disabling, AWE, and Minlogon (I'll post some updates and tips based on my experiences doing that later, likely next week).
I've based my target BIOS settings on the ones listed for the GA-G31M-S2L board.
In the MB Intelligent Tweaker (M.I.T.) settings page, here's the best settings that I've gotten so far:
Robust Graphics Booster - Auto (default)
CPU Clock Ratio: 6X (lowest available)
Fine CPU Clock Ratio: 0.0 (default)
CPU Host Clock Control: Enabled
CPU Host Frequency: 133 MHz (I set this as low as 110, but it sounded over-etched at anything lower than 130 and I may ultimately prefer it higher than 133)
PCI Express Frequency, all the Advanced Clock Control settings, and Performance Enhance all left at default values
(G)MCH Frequency Latch: 200 MHz
System Memory Multiplier: 266 C
DRAM Timing Selectable: Manual
CAL Latency Time: 3
tRCD: 3
tRP: 3
tRAS: 5
Advanced Timing Control Settings left at default values
CPU Vcore: .7625v (I started at .850 and have been slowly setting it lower, trying to get to abysstw's .75v or cics' magic .74375v)
CPU Termination: 1.050v
CPU Reference: 0.669v
MCH Core AND DRAM Voltage left at default values
According to CPUZ 1.49 (not sure I want to install 1.51 with it trying to make ASK.com my homepage!), this gives me a Core Speed of 1188, Bus Speed of 198, and Rated FSB of 791. This compares to a core speed of 900, Bus Speed of ????? (sorry, I didn't record this and can't easily check it anymore), and FSB of 600 for the GA-G31M-S2L board.Not bad, but I have the feeling it's not really optimized yet. Especially since I can't get it down to match the Core Speed and FSB of the previous board.
Thoughts?
Also, the setup above is with the Mushkin memory. With either Kingston memory (ValueRAM or HyperX), I can't get good settings. When I put one of these modules in, the board does the multiple-reboot that signifies it is adjusting the BIOS settings because they didn't work. After it is done and finally comes up, it ends up with Core Speed at 1600, Bus Speed at 266, and FSB at 1066 and the RAM timings have been set to something like 5-5-5-12 (sorry, I din't record those). Thoughts on why this is happening too? Maybe I need to do a 'Load Failsafe Defaults' and 'Load Optimized Defaults' before putting one of these modules in?
TIA!
Greg in Mississippi
Edits: 05/12/09
Try setting MCH Frequency Latch to Auto and Memory Multiplier to 2.00D. IIRC, that lowers memory speed to 200Mhz and FSB to 600 or so.
How does it sound like with your current settings?
I think the UD2H probably prevents bus speed to go below 197MHz. Right now, my CPU Host Frequency is set to 200 since the core speed wouldn't go much lower than 1.2G anyway. Do you hear sonic improvements by lowering the CPU Host Frequency, even if the core speed doesn't (appear to) change?
I'll call Gigabyte to confirm whether the 197MHz "limit" is intended, and if there's a way to circumvent it.
***** EDIT! I WAS WRONG!!! The FSB stayed at 800... I could not read the small print on the Zalman screen. See more info in post below! *****Just before leaving for work this morning, I tried your trick of setting MCH Frequency Latch @ Auto and Memory Multiplier @ 2.00D (2.00B worked too) and it did get the FSB to 600. I didn't look at the memory speed, but did notice the Core Speed went up a bit to 1206 or so.
I didn't get a chance to listen this morning and only did a bit of listening when I was messing with settings last night. The sound showed promise, with good solidity to the sound along with mucho details, but I wasn't totally blown away yet. I did think I heard a difference with various Host Frequency settings, with settings below 130 sounding too etched, but as you indicated, the Core Speed did not vary as I made those changes, so it may have been imagined.
I should have a little time to compare the Freq Latch @ 200/Mem X @ 266 vs Freq Latch @ Auto/Mem X @ 200 this evening, along with listening to various Host Frequencies from 200 and below AND I'll try to get the Kingston memory working with a BIOS reset to Failsafe Defaults, then Optimized Defaults, then set it up again.
Much to do yet.
Greg in Mississippi
Edits: 05/13/09
Have you tried using the Gigabyte EasyTune6 tool instead of BIOS to make the changes? I talked with Gigabyte's technical support, and the engineer said the program may solve our "problem". However, I couldn't install the program on my fully optimized cMP machine, so it'll have to wait until I have the time to reload XP.
I suspect that you'll notice SQ improvements from the new settings. Memory Multiplier @ 2.00 brings FSB:DRAM to 1:1, a parameter that - IMO - is critical to SQ.
I didn't try the EasyTune6 tool and my cMP is well beyond where I can easily install a program like this too.
I'm skeptical that it'll do anything for us that we can't do with the BIOS settings, but I'm willing to be wrong on this. I'm just not likely to be willing to go through the work to get it installed on my machine at this time.
Did you get any more details on what it might do to help us get the machines optimized?
On the setup with FSB:DRAM @ 1:1, I'll listen tonight and let you know.
Greg in Mississippi
I didn't get anything else that's helpful, unfortunately. She asked me to take screenshots of my BIOS (I assume she means photo...) and EasyTune (cpu-z doesn't qualify) for them to see the "discrepancies" and find a solution. I know, sounds like customer service BS to me as well. I highly doubt that EasyTune would display core/bus speeds that differ from cpu-z.
Anyway, I'll call them again later today, and see if I can get anything helpful.
LOL... Cust Support BS indeed. They really don't know how to do what we need! (IMHO)
AFAIK, the main issues I see are:
1. Setting the processor speed down below 1200 or so.
2. Setting the FSB down to 600... I was wrong, your trick with MCH Frequency Latch @ Auto and Memory Multiplier @ 2.00D didn't set the FSB down to 600, it stayed at 800, I just couldn't read the numbers on my small Zalman screen!
3. Getting access to some of the missing options such as disabling Spread Spectrum.
Maybe cics can work his magic and get a BIOS update released!
The picture of a rabid audiophile is me sitting in front of my cMP at lunchtime eating a sandwich and checking BIOS settings!
I did listen for a couple of minutes... still promising, but not quite there yet! I'll try to get the other memory sticks to work this evening.
Greg in Mississippi
More time tonight tweaking the GA-EG45M-UD2H BIOS settings.
I started out focusing on two setups... First, the one that worked the best last night:
Settings:
CPU Host Freq 133 MHz
(G)MCH Freq 200Mhz
Sys Mem Multi 2.66C
Giving:
Mem Freq 800 / 356
And in CPUZ:
Core Speed 1188.6
Multiplier x6.0
Bus Speed 198.2
FSB 792.2
DRAM Freq blank
FSB:DRAM blank
The other settings were the same as the baseline settings from last night.
And the one that gave the best settings today based on Abysstw's advice:
Settings:
CPU Host Freq 150 MHz
(G)MCH Freq auto
Sys Mem Multi 2.00B
Giving:
Mem Freq 800 / 300
And in CPUZ:
Core Speed 1187.5
Multiplier x6.0
Bus Speed 198.0
FSB 791.4
DRAM Freq 198.0
FSB:DRAM 1:1
The second was better in objective terms... Better definition, more silence between notes, good composure when the music got complex, just as expected with the 1:1 ratio, but the highs were a bit etched. The first was more listenable with a softer and more delicate treble and I found my head nodding in time with the music more with it, but it was not as defined.
Remembering that I thought I'd heard sonic differences with various CPU Host Frequency settings even though the Core Speed didn't change with differnces in the Host Frequency, I tried a Host Frequency of 133 in the second setup. Hmph... The strengths of that setup was retained, but the highs lost much of the etched-ness and it was overall more listenable. That's what's playing now and while I still don't think this is the ultimate setting for this board, I could listen to it long-term now.
I'm very curious to see if others can duplicate my results or if it was my imagination.
I didn't have time to work out the issues with the ValueRAM and HyperX tonight... That'll have to wait until Sunday or later as I’m away for business for a few days.
Greg in Mississippi
Back home after being in the DC area for the weekend, I played with the cMP settings just to unwind a bit.
1st, I did two of the last customizations remaining... I checked to see if the soundcard was sharing an IRQ (it didn't) and then disabled the un-used USB controllers. BTW, this produced a noticable improvement in transparency... well worthwhile.
Then I tried to compare the ValueRAM vs HyperX vs Mushkin memory sticks. With the BIOS settings I was using early last week, both the ValueRAM and HyperX would not get a clean boot when installed. Now the ValueRAM worked ok... but I still had BIOS errors with the HyperX. I didn't take the time to further troubleshoot the HyperX, but did go back and forth between the ValueRAM and the Mushkin... and I preferred the ValueRAM. I heard better definition up and down the spectrum, but most noticable in the bass. Tonal balance was a bit different between the two, with the Mushkin sounding a bit more forward in the mids and lower treble.
I would judge that whether you liked one or the other could be very dependant on what one considered important and what worked best in one's system... and since they both cost less than $20 ea, if you are so inclined, get one of each and compare yourself.
Maybe this weekend I'll spend more time with the HyperX. In the meantime, I'm very curious what the BIOS update will bring.
Greg in Mississippi
I just convinced Gigabyte to give me a "special" BIOS version. I'll let you guys know if it fixes the problem.
Hello Abysstw,
I know I'm replying on an old post but:
any news on that special BIOS for the GA-EG45M-UD2H?
Did you actually receive it?
Are you able to get bus speed to go below 197MHz?
What about spread spectrum?
Thanks
Etienne
abysstw,
Any news on the how the 'special' bios worked?
Inquiring minds & such!
Greg in Mississippi
That's GREAT!
See if they can give us the option to disable Spread Spectrum at the same time!
Thanks!
Greg in Mississippi
P.S. Contact me via my Asylum profile so we can talk Juli@ mods.
nt.
Keep us posted as the mobo burns in relative to sonics. did you use the easy tune facility or utility to adjust speeds/voltages?
Did the voltage adjustments manually in the bios.
It does not seem to be available in the UK (advertized but not stocked) so I will have to be a late adopter.
Glad that you like it :)
I think the normal voltage of the E7200 is 1.1v. I'm out of town right now, so you might want to reconfirm this with other inmates.
I compared a 512Mb stick of HyperX to a 256Mb stick of ValueRAM in my supercharged cMP2 Saturday evening with a rematch Sunday evening so I could test it out with b24 too. To try and cover all the options, I ran the HyperX at various timings, 3-3-3-7, 3-3-3-6, and 3-3-3-5, along with the manufacturer's recommended 4-4-4-12 and a wild-card 4-4-4-5.
In short, I liked the HyperX best at 3-3-3-5, but still prefered the ValueRAM at 3-3-3-5. The HyperX was very clean and didn't do anything wrong, especially at 3-3-3-5, but the ValueRAM gives me deeper and darker silences between notes and a wealth of texture and density to the sounds that added to the sense of realism along with a greater sense of presense and vibrancy in the midrange
As for the 44.1 vs 192, what I heard provided me with more insight as why I prefer 192 in my setup... it's that wealth of texture and density to the sounds which is present with 192 and diminished with 44.1 (and it grades down pretty linearly as you select the intermediate sampling rates between these two). The metaphors that come to mind... "average recordings start to sound like direct-to-disk"... "it can be so involving that I'll stop what I'm doing and just listen to the music"... "For the first time, I'm hearing all of each note... before now, the beginnings, ends and much of the fine textures in the middles were missing."
In my setup, 44.1 is very, very good. But in my setup, 192 is exceptional!
Greg in Mississippi
Reading the specifications of the ValueRAM they generaly need less power than Hyper sticks, as peer Cics recommandation the lower power give the best sound.
You can check on www.valueram.com, the specifications for valueram and hyperX is available for all part number.
trend in system architecture and whether one finds upsampling preferable.
I am still in analogue mode so I have no opinion to offer on #24. When I get tired of listening to my LPs and want to hear stuff that is only available on CD, I will return to cMP.
Nonetheless, with my system there was much more ease to the sound with the upsampling. My system needs no additional sharpening and if anything would benefit from intentional softening. I do not doubt there is something like this at work.
Based on the fact that my system is very high efficiency (edgar horn titan II with largest round horn and TAD 4001 and TAD 2001 (w/ small horn)as super tweeter) my guess is that similar systems would also sound "better" with the upsampler. My experience with direct radiators of low to middle efficiency would lead me to believe that owners of these speaker would not mind a little additional "sharpening" of the sound. If the upsampler smooths the tops of those sawtooth waves, those remaining tips could actually be euphemistic with the direct radiator and its comparatively more difficult job of loading a room. All of this is nothing more than conjecture but it would be interesting to see if there is any discernable trend.
Now I will get deluged with reports of 44.1 partisans telling me of their horn loudspeakers!!!
Bye,
Rick McInnis
Rick,
As always, good to hear from you!
Well... I can provide some contrary data here... my system is based around a pair of fairly in-efficient planar magnetic (and fairly rare) Eminent Technology LFT-IVs.
For me with my setup and these speakers, the additional detailing provided by the 192 setting adds sharpness to the beginning of sounds... it adds the initial impact and rise time that is present in real life (jingle your key ring), but so often missing in reproduced sound. I don't hear extra hardness, but the beginning of each note is more clear and distinct with the 192 setting... adding 'micro-dynamics' and giving my planars a taste of most horn-based systems do very naturally.
My 2 cents!
Later!
Greg in Mississippi
P.S. Don't be such a stranger!
Hey Gstew,
My magnepans dont suffer from bright treble, but they do act as radio antennas! Thus it is possible for rf to get into the amp and modulate the signal. This kind of sounds like the "digital" sound. Putting the chokes really helped add warmth and also a bit of top end resolution.
For about $12 you could really improve your emminents if they are anything like the maggies. I certainly found a great improvement:
http://db.audioasylum.com/cgi/m.mpl?forum=mug&n=139415&highlight=octamom+dawnrazor&r=
very interesting! do you think this concept might work with electrostatic tweeters (like martin logans)?
I really dont know. BUt I am finally going to hear Al's system on Friday and I can ask him for you.
Have you tried powercords on the Logans?
About 5-6 years ago I made some powercords and speaker cables for a guy who had some logans. He was supposed to use the cords on his pre and his cdp (hey, it was about 6 years ago...) and I got this call from him and he was frantic.
Apparently it was too much trouble to get behind his pre and cdp so he just tried the cables on his Logans.
The difference apparently was pretty big and shocked him (but not in the dangerous way). He said something like "...but they aren't even in the signal path, how can they make a difference?"
Anyhow, you may have already tried some, but if not give it a shot.
d
I have gotten good results with pcs and conditioners with my Logan Requests: improved clarity, detail and tighter bass. I haven't choked 'em, though.
thanks yes I have experimented with pc's on logans, not very much but some. I use the hammond inductor on many of my circuits as well as a running springs ac conditioner on my panels.
I do use an inductor, cap filter on my woofers and they improve the bass a lot. so I am always interested in these type of low cost/high impact mods.
oh, well.
Greg,
Nonetheless, a planar speaker is unlike other direct radiators because of its ability to reproduce a large wavefront without all of that frantic motion of an un-fron loaded cone.
I am an admirer of Bruce Thigpen's thinking and products. I have maintained an ET-2 arm for twenty years keeping up with his upgrades, not yet getting the carbon fiber arm wand.
I have never heard his loudspeakers though I am very familiar with how they work.
I came to horns from electrostatics. The large planar loudspeakers are excellent. I have a room that would overwhelm any of them, or at least, anyone of them I could afford.
SO, I do not think we are far off in our comparative assessments. Those speakers will follow the waveform and therefore, no surprise to me, you would benefit from the smoothing from the up-sampler.
Thanks for pointing out this discontinuity in my thinking.
Now if only we would hear from some other cMP users.
Bye,
Rick McInnis
I must confess I've been running ET arms since 1986 or so starting with an ET-1. While my vinyl source is not currently setup, I have a tweaked ET-2.5, use a high-pressure pump with serious smoothing tanks, and regulate the pressure to about 20 lbs. I do want to try the large storage tank / pump off setup when I get it running again.
Then on the planar side, I've been running Acoustats, Magnapans, Martin-Logans, and now ET speakers pretty much exclusively since 1982. The only speakers that have tempted me away from them were a pair of Spica TC-50s and I still have a pair of them around. I got the LFT-IV's for a song about a year ago... they replaced a pair of Gallo Ref-3's and I strongly prefer the LFT-IV's, even with an almost 4x price differential.
I never had a pair of direct-drive amps even tho I've lusted for them for years, but did have a pair of original CLS's and still have a pair of highly tweaked 1+1's... but my heart is currently with ET speakers and in addition to the IVs, I also have a pair of LFT-VIs and LFT-VIIIs.
TOO MUCH GEAR!
Greg in Mississippi
I promise not to tell Barney Frank how profligate you are. My, my, my, you should be ashamed!!!!
Sounds like you are having fun. OH, I did it again. I am really going to get you in trouble, it seems.
Maybe you should make it correct by calling your collection a museum of planar loudspeakers? The world does need one, after all.
Bye,
Rick McInnis
PS if it was not OK to go "off-topic" I would have had most of my posts deleted.
I had an ET 2 tonearm for about ten years. I drove it off my sears air compressor which had a 20 gallon reserve. I would start it up (sounded like a lawn mower in th house--wife hated it) then I would set the regulated pressure at some value (say 3 or 4psi) and it would run for 4 or 5 hours. Man that was great fun and wow did it sound good that way. Sigh and then I sold all my original vinyl stuff (vpi, ET2, Koetso cartridge) and even half of my vinyl. Now I have an old tecniques sl210 integrated table, pivoting arm and some mm cartridge. Haven't listened to it in 4 months.
I have Martin Logan Monolith III's biamped.
Your comment on Eminent brought back some memories and hence this post.
I, too, use a BIG air compressor. I have got a manifold to optimize this so I do not get 4 hours out of my large tank, more like thirty minutes but it is far enough away as to not be a problem. High Pressure and high flow make for a STIFF bearing.
I had a pair of MARTIN LOGAN CS-1's that were powered by a dedicated tube amp (Joe Curcio designed and built) that was once part of the original ACOUSTATs. No transformers between the output tubes and the stators.
When I got a large listening room these speakers did not have a chance unless I wore them like headphones. They are classics, though.
Sorry you got rid of all that nice stuff!!! My LPs are a part of me even though some are borderline unlistenable. They are my biography.
forgot to mention I heard Bruce Thigpen's fan generated subwoofer down in Las Vegas 2 years ago. Tremendous ability to reproduce bass down to a couple of hertz (he had a scope there to prove it). But it didn't seem to have very good dynamics. Maybe it needs more work. But he is very innovative and his stuff will be among the classics. Love to talk to him --- very bright young man.
Rick,
If you mean confess to owning a horn-loaded speaker then I may own up to having a pair of Gedlee Abbey (clone) speakers which have a 12 inch waveguide. For the most part I would describe these as efficient and transparent speakers; they are powered by a Peter Daniels Patek-style gainclone amplifier.
They have shown clearly differences in the cMP setup and until recently I have always preferred upsampling at 192k - easier to listen to / better detail. However I have of late been running at 44.1k and I have been pleased with the sound; with cPlay 24 I am not sure (after one evening) whether I prefer 44.1 or 96 or 192. Whatever the first listen to 24 put a smile on my face and that was the initial listen at 44.1.
I will let the system settle in for a few days before deciding which I like.
On the analogue side I have not played any records for about six months being so busy with setting up cMP2; it might be a good idea to just leave the system as it is but I have not started modding the Juli@ card and its power-supply.
So much to do so little time.
Refer to ReadMe.txt for important details & specifications.
Change Log (2.0b24 release):
- ASIO startup refinement
- ASIO output interface replaced
- Minor DSP refinement
Please REMOVE previous versions before installing cPlay 2.0b24. Normal (SSE2), SSSE3 B9 and SSE4 B9 versions are available. Running cPlay on a CPU not supporting the required instruction set will cause cPlay to exit immediately.
For cMP² users, review these BIOS changes (underclocks FSB, CPU & disable additional USB items). New settings are in effect allowing for lower power consumption (below 20W) as well as kernel changes. Recommended Gigabyte mobo GA-G31M-S2L is no longer available, use either GA-G31M-S2C or GA-G31M-ES2L.
due to major turntable modification in progress.
I cannot say with specifics what 24 is doing in comparison to its immediate predecessors, but there is no question this sounds very good but (here comes the big surprise) it is missing that enveloping musical cocoon effect that good analogue does so well. The image between the speakers is precise in three dimensions but there is not much of anything to either side of the speaker. Of course, this could be my DAC's fault. I am just reporting what I hear. I hope I would be considered the last person with a desire to savage cMP. I believe these are the tired old digital shibboleths and I am not saying anything that hasn't been said a thousand times.
Nonetheless, as fine as cMP is, it still is not the missing link that I so sincerely want it to be. cics has made immense strides in making digital music reproduction enjoyable; don't get me wrong I can listen to this and not be miserable, it is just the whole time I am thinking, "I can't wait to get that turntable working again".
Something I discovered in the last few weeks: with analogue I was interested in listening to unheard music which, for some bizarre reason, had not seemed interesting to me while on a steady diet of digital. I was buying lots of CD's but they were of music I had on LP.
With a CD of something I was familiar with on LP I assume my brain was able to fill in what wasn't there (upsampler cannot predict what isn't there at all) and I was able to enjoy the sound since it did sound MUCH better than the old THETA JADE and GEN Va by an order of magnitude. But as the old story goes one still cannot put Humpty Dumpty back together again.
Another thing is that I want to listen REALLY loud with digital and find nothing attractive about this with analogue. There is much greater dynamic range with the LP and one compensates with digital by turning it up LOUD to get some impact, I surmise.
To be fair, I have MUCH more money involved in the analogue set-up and when one makes a value judgment, ESPECIALLY considering cics's saintly generosity, it damn well BETTER sound better.
I am glad to have such a good back-up system. What else can one do for recordings never released on LP?
Even with all of the tweaking cMP requires to get started it is nothing compared to what is needed to get a top performing LP set-up.
And digital is quiet, one cannot dismiss this. It is a fine attribute.
Maybe cics has one incredible trick up his sleeve, and maybe there is a DAC that can change water into wine. No one will be more effusive in their praise of such a thing if it appears. But, excuse me for repeating it and I know it is said too often for many, but there is still a LONG way to go. Analogous to the line about the more you know the more you find you do not know ... I think this is happening with cMP and other digital advancements. After the glow wears off, even though one can sense a great improvement from what has come before, the "latest device" seems to point out even more clearly than you perceived previously just how much ground there is to make up. The audio tortoise and the hare ...
Cordially,
Rick McInnis
(admitted analogue partisan)
Hi Cics,
After your suggestion on the "DSP buffer" setting I start to notice the sonic difference for the 3 "DSP Buffer" settings (Small, Medium, Large). When I set to "Small", the resolution is highest, and when I set to "Large", it is very musical and comfortable to listen to. So I use Small buffer to listen to general pop, medium for male vocal, large for female vocal. This gives me alot of power to adjust cplay to suit the music :)
Appleteapot! Thank you for bringing this up. I discovered more music in my files. I won't say which setting I like best but it has moved my system closer to what I expect out of my music. There's more experimenting to do ,but here is yet another thing to adjust to make cplay better. Hey I know it was always there -- but I just discovered it.Thank you again (and obtw thank you cics!)
Just like the old vinyl days -- adjusting the cartridge vta for every record. And I mean that in a positive way.
Edits: 05/06/09
Then I wonder whether I can tweak the Lynx buffer along with DSP buffer...
At the same DSP buffer level, The lower Lynx latency is always better, sounds more accurate. So I still stay at 32 at the moment.
Hi Cics is it possible to change the DSP buffer setting so numeric values setting say 0-1000 so that we can freely tweak the character of the sound?
At small I get a very very fine grain in the highs whichs makes the highs very very finely separated --- great air around everything almost a swimming liquidity. At 1st I just loved it and was going to set it at small and forget it. But, alas, there is a tradeoff. The bass becomes weaker and the soundstage moves forward and dynamics suffer. It sort of like a 'tube' filter. So I use it sparingly where I can or where I need it.
You are right it is so nice to have. I tested only with 16/44 files. Small is killer for 24/96 or higher data files (I always match sampling rate on cplay to the data now).
Thank you very much...B24SSE4's musicality is comming back!! Tears are in my eyes... To me, previous verions resolution is so high that I can hear background noise.
B24SSE4's background noise is not as high (resolution is lowered a bit?) However, musicality has improved and I can enjoy the music more, especially female vocal. I then listen to male vocal immediately, tears in my eyes when I listen to Paul Potts. So to me this version is a very very pleasant change.
(But just watch out in subsequent versions, too much smoothness will start affecting pops and "strong stuff". )
Thank you so much!
Edits: 05/03/09
+1 but there is more for me. I loaded 24 last night (is there a prize for 1st downloader?) and it was ok. I liked it but didn't get excited. This am , in my 5:00-7:00 powerline sweet spot I began to here more into the music. I must say that I had made a couple of changes as well. Yesterday I put my audio pc on hockey pucks versus on the floor--always a good start for isolatating the source. Anyway it sounded better with 23, little fuller in the bass, more air in mids and up. After loading 24 I felt intuitively I could do more. So I reduced my hockey pucks from 4 to 3 and then put an aurios bearing (the floater type) between my case and the pucks. Wow what an omg moment! It was like the fog lifted. I could see into dense classical orchestrations. Highs were less bright, bass tighter and deeper, mids were glorious. I have many household tasks to do today but I don't want to start. I want to listen to everything I have.
I feel 24 (sse4) is more transparent but not in a 'in your face/offensive' way. I know some of this is coming from my better isolation but w/o the details/smoothness of of cplay it wouldn't be there.
An example of what I'm hearing: one of the telarc samplers has an excerpt from the 4th movement of beethoven's 9th and where the male bass comes in is very dramatic. I could always hear air around his voice but now it sounds as if there is a cushion of air all around him. Yes I sense I can hear behind him. Sounds weird but this is my impression.
There are other examples of voices where I can now clearly understand words better and have a sense the weight of these voices/chests radiating in space.
Just simply wonderful, thank you cics--and all this at 44khz.
i can't say it better in english.
:)
you keep talking I appreciate your words, especially on mods. wish I could speak german like my kids. my wife speaks italian, spanish, and some french and english, seems like I am the only one language talker in the family.
In the last version ( 2 0b24_ssse3b9) the sounds are a bit closed .
I think the soundstage of v23 is better than v24
Cics, please check into this.
I removed original release post until files can be downloaded.
riboge wrote:
"Link to download is not working. Cics, please check into this.
It's often several hours before the cPlay links work but they usually seem to sort themselves out. I've no idea why . . .
D
I had been on the fence about this, liking either way but recently I conducted some listening tests when my power line was the cleanest (5:00am -7:00am) and after I went completely fanless. What a revelation! You know when you change something and everything just falls into place how you want to listen to all your faves? I have tested for 3 days but today I finally concluded I'm a 44khz guy. Initially I thought 192 was airier in the highs but after several days listening I conclude 44khz is better defined in the bass, mids and while not as deep in the highs as 192 is definitely more true to the actual sound of music in the hall. 192 has a sort of impressionistic feel to the sound, a little out of focus but pretty. 44khz is the real thing and obtw I'm listening to cplay 2.0b 23 sse4 which by definition has no upsampling period. I often thought that my 24/96 files sounded better @ 96 than upsampled to 192 and could not seem to reconcile that with the fact that I believed 192 was better for 16/44 files. Now with 44 khz with cplay 23 I finally (at least for now) conclude that right here is my best sonics. At other times of the day (when the powerline is noisy) I think I prefered 192 because, being a little out of focus, it masks the noise induced/subtle hardness. 44 at these times still sounds good but not preferred to 192. But in the am, when I do my serious listening, 44 khz is much much preferred.
Getting back to making some breakthrough on ones system, when this happens for me I really just put on some music and listen all the way through w/o flitting around to favorite cuts or passages. And that's what I have done for 44khz the last few days. Yes it does solve my metallica issue but I am still going to try some type of ups particularly one that can allow battery operation for several hours (so I can get that morning sound all day long). Btw does anyone know of a UPS that will allow this (very feasible I think with cmp^2 operating at 20watts or so)?
Cics: juli@ analogue outs now is so far superior to juli@ digital outs through the benchmark even with its better dynamics in the bass. Juli@ is better for resolution from the mids up. I guess I have to try the Sabre like Alfred (Alfred how does it feel to be on the leading edge of dac's, power suplly mods et al).
Anyway just thought I'd share this experience. Those who have not tried 44 setting on cplay 23 please do, you may be surprised as I was.
theob wrote:
Btw does anyone know of a UPS that will allow this (very feasible I think with cmp^2 operating at 20watts or so)?
The 50/60 Hz regeneration circuits in most UPS devices are very poor quality (at least for audio, which is not what they were designed for) and the power they supply is likely to be significantly noisier than the mains. The technique has been tried before but never AFAIK successfully.
Dave
thanks dave
Theob,
Your post is a great testament to how we always need to challenge our assumptions... and as a corollary, check our settings.
At the end of last weekend, my cMP did not reboot after putting in some Juli@ changes. So I had to reset the bios and then reapply all the bios updates. That was a good thing... I'd missed that the preferred setup could go lower in CPU voltages with EIST disabled and I got it all the way down to .74375v. So I got good sound out... but the next few days and then after I did a further Juli@ mod, it didn't sound quite up to snuff.
I went back and checked the bios settings and saw that I'd missed setting the CPU Host Frequency. Set it right and the magic was back... mostly. Then I ran CPUZ and saw that my RAM settings were on 'Auto' and had defaulted to 3-3-3-7 instead of the 3-3-3-5 that seemed to work so well in my setup with the ValueRam.
So just to be sure, I compared a number of different settings options...
1. RAM timings 3-3-3-7 vs 3-3-3-6 vs 3-3-3-5 (I'd been using 3-3-3-5)
2. Juli@ outputs 1/2 vs 3/4 and having monitoring on or off (I'd been using 3/4 with monitoring on)
3. All the sampling options from 44.1 to 192 (I'd been upsampling to 192, which is why I'm putting this in your thread, Theob!)
I ended up with was 3-3-3-5, Juli@ outputs 1/2 with monitoring off, and upsampling to 192.
While I hear a slight increase in bass and midrange solidity with the lower sampling rates (and it seems to transition pretty evenly as you go up from 44.1 to 192), I get an increase in treble details at 192 with no loss of focus. And with the replacement and add-on power supplies in my setup, the bass and midrange solidity already starts out pretty darned good, so this increase in treble details makes a greater difference in my setup and to my ears.
My lessons from this week's exercises are:
1. As new versions of cPLAY come out or I put additional modifications in to my cMP, it's worthwhile to re-check the effect of the various optional settings.
2. If something does not sound right, check the settings. Having the CPU Host Frequency not set correctly AND my RAM timings different than what I consider optimal made a significant difference in my cMP's sound.
3. The sound quality of a cMP is MUCH greater than the sum of it's components... hardware, software, and settings. They all play together to produce what we hear... and if they are not set to the optimum mix, you will hear something that is significantly poorer sounding. Those people who run cMP/cPLAY on a laptop with 1/2 of the optimizations are not hearing what this setup can do. Heck, I was not hearing what this setup could do with one of the most highly-evolved cMP's around and just two settings off!
Greg in Mississippi
have you tried no upsampling for prolonged periods of time? maybe with all your mods 192 is better. but did you have any trouble getting 150 host clock control? the bios resetting itself to auto host clock control and resetting