|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
82.156.128.190
In Reply to: RE: Kingston 16 gb - no problems with installation posted by rickmcinnis@dogwoodfabrics.com on February 10, 2011 at 07:57:25
Hi Rick,Well let's spread the word than: MOUNT AN SSD IN YOUR cMP !!!
:-)However unfortunately my price/performace ratio question was unanswered in this thread:
http://www.audioasylum.com/forums/pcaudio/messages/8/84277.htmlI wanted too now what is the most costs-effective solution for sound quality:
- a small capacity SSD with big external HDD (eSATA/USB/firwire/NAS)
or
- a large capacity SSD?After a while two respectable inmates responded: Ryelands and Gstew.
But the thread didn't give me any clues about what is the most cost-effective route to go.In my setup I don't hear any sound quality differences when I use the LAN too connect too a NAS.
But I only own (fairly good) mid level quality gear.
So I had hoped for responses from inmates with real top noth high quality gear,
which would be able too tell from there own experience, what sounds better.But Ryelands has not responded yet about what his experiences are with a small SSD he recently purchased. I hope he will report soon.
Mark
fully optimized cMP2 PC -> ESI Juli@ -> Van den Hul Optocoupler MkII-> Lavry Black DA10 -> XLR Mogami Gold -> Klein & Hummel O300
Edits: 02/10/11Follow Ups:
Ryelands has not responded yet about what his experiences are with a small SSD he recently purchased.
That's because, for the last few days, Ryelands has been listening to another system, one he's building for a friend. The SSD, which was very easy to install, had only been auditioned for a few hours though I did finally switch over for (hopefully) good last night.
Yes, I think it sounds better though it's not one of those the-scales-fell-off-my-eyes differences that others report. Even so, for the modest £35 cost for an 8GB device, I think it was worth doing.
BTW, the PSU I'm using to drive both the HDD and the SSD is a pretty good linear design, not one of those "Granite Digital" switchers (Yuck). Whether that matters, I can't say.
I'll see how things go and, if there's anything worth mentioning, I will.
Hi Ryelands,
Thank you for your feedback.
Pitty that it didn’t bring a major improvement in your setup but only was worth doing because of the modest cost.
I read that you still have an HDD connected too the MoBo.
Did you also try the SSD without any HDD being connected too the MoBo?
Mark
Some background about the power supply in my setup:
I never used the ‘Granite digital’ switchers in my system.
My current power supply situation is:
* ‘Dirty’ linear: PSU HQ-Power PS1503 SB powers:
- P4
- picoPSU PW-200-M which powers the USB bracket (wireless mouse/keyboard) + SSD
* Linear PSU: EA PS2012-05 powers:
- picoPSU 160-XT which powers the P24
fully optimized cMP2 PC -> ESI Juli@ -> Van den Hul Optocoupler MkII-> Lavry Black DA10 -> XLR Mogami Gold -> Klein & Hummel O300
Did you also try the SSD without any HDD being connected too the MoBo?
If using the SSD, I don't have an HDD connected to the mobo. My setup is basically the same as yours - music server > LAN > Audio PC - the differences being that I copy data from the server to the PC before, not while, playing and that the (headless) Audio PC is controlled from the server.
Before making a categorical judgement, I'd need to compare the same OS running on an HDD and an SSD in relatively quick succession and must in any case get used to the AQVOX driver, which certainly did make a significant difference.
To do all that properly means listening to the same limited selection of music over and over. That's something I'm not always very motivated to do . . . though perhaps I "bloody well ought" to be. :> )
Dave
You guys (you and Rick) are making me crazy. I have some hardware and software coming that will make re-loading windows very easy. My 1st step is to get familiar with same in an effort to implement the Ryelands algorythm (i.e., loading the music file onto a flash drive in real time running cplay). I have been experimenting with offline upsampling and 'manual load' fd playback for the last few days and these items while not an ssd gives me some idea of what is possible beyond where I was. And now I see that the potential of cmp^2 playback is perhaps way beyond where I had been. I have been upsampling 6-7 of my favorite music files with sox learning how to trim, upsample, reduce gain, compare various sox parameters so feverishly the last few days (-v versus -s as an example) that my wife keeps asking what I'm doing. When I expain she says '...yea right!...' Anyway comparing cpu usage with task mgr shows that running off a fd with offload upsampling reduces cpu usage down to zero (or very small but zero as shown in task mgr) and only peaks to 15% or so when swapping data from fd -> memory. I assume running off an ssd probably reduces this even further. So after I try fiddling with my new sw/hw I may very well buy that kingston 8 g ssd. After all it is so cheap (which is what I will tell my wife) that I cannot pass up the opportunity.
You guys plus Dave are bad influences!
maybe the fact that I am using the internal CPU in real-time to do the upsampling, etc. is why I am hearing such a difference?
At the same time I wonder if those who were not enamored of SOX and SRC might have a better impression with the SSD installed?
It could well be that if you do the processing before, and your CPU usage becomes minimal there might not be the same difference as I am hearing.
If it turns out to be an either/or thing - one can say it is much easier to just have the music on the HDD and play it instead of all of that work.
Now, the logistics of getting two, otherwise similar, cMP setups in the same room with SSD and real time upsampling and regular HDD and offline upsampling.
Being basically a lazy man who just wants to listen to music while sipping rye whiskey, I hope the SSD/real time approach is, at least, almost as good!
I do look forward to hearing your impressions with your new implementation, Theo.
Bye,
Rick McInnis
Interesting questions...now Mark listens 44k-> 44k no upsampling on the ssd and his cpu usage runs 0-1%. When I run offline upsampling off a fd I get 0-15% cpu usage (I only get the spikes at transfer from fd-> memory but there are 4 times as many transfers because of the higher native resolution). But If I run no upsampling off a fd I get 0-4% cpu usage agian with spikes @ transfer points (but far fewer transfers). Maybe that is the secret. Did you try no upsampling?
I'm starting to believe that whenever you lower power demands and burst current draw you get better sonics. I guess I gotta try an ssd.This is really facinating.
I just tried 44-> 44 off a fd and it is very close to offline up sampling off a fd sonically. I think I like upsampling better by a smidge though.
listening to it now would not tell me much,
I think the "power" aspect is important but it cannot explain everything.
When the disk drives are powered by their own supplies that argument seems tenuous to me, anyway. I can see the power argument being persuasive when it comes to power to the MB.
One would assume the more activity of the processor would, in turn, require more activity from the drive containing the OS and the software. The faster the drive can respond to requests it only makes sense this would greatly effect the function of the system. At least to my way of attempting to understand what goes on in the computer.
We have learned that speed is not important for the CPU or the memory but maybe where speed makes a decisive difference is with the drive containing the OS, the cMP/cPLAY software, or both?
It seems to me, going back to cics's idea of jitter, - this could be a source of a form of jitter if there is a lag in instructions required by the system?
Bye,
Rick McInnis
Even though the drives are powered from the outside they still generate emi/rfi on the mobo perhaps when called upon to work. Hey I'm just speculating.
Hi Theo,
Okay, here some good influence than:
It looks like you have been behind your computer too long and too much last few day’s.
Next Monday it’s 14 February: Valentines day.
Surprise (!) your wife with a romantic diner for two and a nice bouquet of red roses.
1 for every year you know her.
So far for my bad influence.
May be I’m doing something wrong or my setup is not revealing enough, but in my setup there is hardly any gain in sound quality through up-sampling. It just sounds very, very slightly, somewhat different. I would not even dear to call it ‘better’.
Please first be seated than (with your seatbelts on) and your feet firmly on the ground, when you listen too your cMP with a SSD inside for the first time. Compared to up-sampling in my setup, it was a massive improvement in sound quality. Bigger than all previous PSU improvements. “Bordering on the absurd” as Rick put it so nicely.
Listening ( I listen only 44.1 -> 44.1) indeed hardly does use any cpu capacity.
Typically only between 0% and 1%.
Even with use of the network function, it only rises too 2% very briefly when cPlay loads a (part of a) wav-file through the network in too memory.
I will do some more listening: comparing using an USB flash drive against using a network connection.
Using the LAN is very easy for me compared too the Ryelands algoritm. At least: that how it looks too me.
Let us know how you progress.
(and if she liked your Valentine surprise)
Mark
fully optimized cMP2 PC -> ESI Juli@ -> Van den Hul Optocoupler MkII-> Lavry Black DA10 -> XLR Mogami Gold -> Klein & Hummel O300
Post a Followup:
FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
This post is made possible by the generous support of people like you and our sponsors: