Computer Audio Asylum

Music servers and other computer based digital audio technologies.

Return to Computer Audio Asylum

Message Sort: Post Order or Asylum Reverse Threaded

Page: [ 1 ] [ 2 ] [ 3 ]

The art of building Computer Transports

196.25.255.195
Posted on April 5, 2007 at 08:00:10
cics
Audiophile

Posts: 1320
Joined: November 9, 2006
 What started out as a $1,000 experiment in storing & playing CDs from PC turned into an amazing step forward in music reproduction.

Initially, I was happy to sacrifice some sound quality over PC convenience (as my 18 month old daughter was causing havoc with all our CDs - fingerprints, scratches, hidden just about everywhere...). Like others, my CDs were copied to black CD-Rs.

Welcome to a new paradigm. I was pleasantly surprised over how good a computer transport sounded. It quickly surpassed my AA Prestige SE player's internal Philips drive. This was recently upgraded to a Sony which is a step up from the Philips. Nonetheless, the computer delivered much better performance. Having the pleasure to listen to high-end transports (Wadia 270SE, Esoteric P03, MBL's $20,000+ 1621A, full DCS stack), I am comfortable in saying that Computers have outclassed them.

My experience in IT and a background in physics certainly helped in getting to an optimal setup together with learnings from the web including Audio Asylum. I kept notes on all tunings and changes that were necessary. As this list grew, I decided to write a paper. Interestingly, this paper led me to question many assumptions which in most cases led to further sonic improvements. Its amazing how much can be done when creating a dedicated computer transport. It cost me ~$1,500 (excluding digital cable).

I have reworked this paper to be a full guide on building computer transports. In keeping with the spirit of open source and freeware, this paper can be found at the web link provided, otherwise goto http://www.imageevent.com/cics. There are many screen shots and details on important BIOS settings, etc.. Its in pdf format and is 2.7MB large.

Your feedback would be appreciated as I'm sure more can be achieved. Paper is at version 0.1.

 

Hide full thread outline!
    ...
this is worth at least $600, posted on April 5, 2007 at 09:37:18
bobp


 
Cics, congratulations, this is an outstanding work. Because people tend to disregard free advice, I'll put a price on it. It can save at least 20 -30 hours of testing and tweaking time to someone starting from the scratch. Depending of person's billing rate, it can save from $400 upwards. (I am playing with PC servers for two years and know from experience what time waster it can be.)

There is nothing I would change. I would add:

Under 3.2

PC case - Case is an important part of PC server and has to be chosen with the utmost care. Minitower case provides better air circulation and is preferable to desktop case, but it is an eyesore in audio environment, unless someone has Theta Citadel or Thorens amps :-)

If desktop case is needed to match the rest of audio equipment, it has to be chosen with special care to air circullation:

- choose multi compartment case with walls separating motherboard, Hard drive(s) and Power supply areas.

- choose case with two 120mm exaust fans. Large fans can be set to
operate almost silently when powered by 7V (i.e +12 and +5 instead of +12 and O)

- choose a case with bottom and top perforations to allow for convection cooling (like valve amps)

Before installing ANY components into PC case, dampen the case thorougly to prevent ringing from vibrations. Dynamat or EVA kits are the most effective. Make sure not to block any perforations on the case.

-Use sorbothane gaskets under the exaust fans, under power supply, and under all screws, including Hard drive. Consider using Vibrapods instead of provided feet. Also, consider placing 3" thick maple board under the PC case. Because of the applied materials, it is more economical to go with the steel case than Aluminum.

-PC case is sometimes not rigid enough for audiophile use. Brace it with Aluminum square tubing. An audible difference.


3.5. Use good quality AC power cord to your PC. At least 14ga, with quality plugs.

 

Re: The art of building Computer Transports, posted on April 5, 2007 at 12:31:05
ASB
Audiophile

Posts: 36
Location: Istanbul
Joined: September 1, 2005
 Thanks cics.

Great document and will be very useful to everybody interested in PC Audio.

Thanks again.

 

Great Post - thanks!, posted on April 5, 2007 at 13:38:02
navman
 I am soon to build a PC based transport, and this has me really excited. It would have taken forever to get all that info on my own.

 

Bobp, you're right, cases are important and affect sound. More..., posted on April 5, 2007 at 20:39:15
cics
Audiophile

Posts: 1320
Joined: November 9, 2006
 I use a standard mini-tower an a solid platform. Don't let it lie on carpets.

I've been silent on this as I plan to test the Zalman TNN-300. It has many design features I like: fanless PSU and CPU, rubber mounting for hard drives and (as you correctly point out) a good chassis (heavy aluminum plating). Unfortunately, Zalman has not yet offered a bracket for Intel Core 2 Duo processor (which would definitely work with Zalman's heat pipe technology) and has a flimsy stand (spikes would be perfect). I also want to confirm if PSU has active PFC (specifications seems to suggest this).

If someone could check this out it will be great. Costs ~$800.

With standard cases, try to avoid more fans as they create more vibrations and consume power. I chose Biostar mobo as it runs cooler than most mobos. Also, many functions are disabled (thus less power and heat). Overall design seeks to lower power consumption.

If you have high ambient temps and need more cooling, use PSU fan. The recommended Enermax has dual fans (with minimal vibrations) and offers variable speed control. Dial this to max (no impact on sound). Case air is pulled through PSU and expelled at rear.

 

Re: Bobp, you're right, cases are important and affect sound. More..., posted on April 5, 2007 at 22:17:55
bobp


 

It's not difficult to find a good case when you have $800 to spend.

But when you have $150, it is more of a challenge.

Antec 2400 is my favorite.

 

Some additional suggestions, posted on April 6, 2007 at 01:50:41
Christine Tham
Reviewer

Posts: 4839
Location: Sydney
Joined: December 29, 2001
 On the premise that power supply fluctuations cause jitter (because it causes the ground to bounce which affect the precise timing of logic transitions), you probably want to avoid any motors or fans.

For example, you can build a totally fanless PC (no fan in power supply, passive cooling etc.) and solid state hard disk (eg. high capacity compact flash or similar). The actual music files can be stored elsewhere (ie. on a server or NAS).

Another area to look out for is EMI/RFI. For example, you could try various methods of shielding the audio circuits. Otherwise they can also add jitter.

You can also reduce generated EMI by using a slower CPU, or underclocking. If the BIOS supports spread spectrum (the best BIOS I have seen so far applies spread spectrum not just on the graphics bus but also memory and PCI), you could try enabling them.

Your recommendation to use optical for digital out would only make sense if you are using a reclocking external DAC, because otherwise optical output adds a lot of jitter (due to the translation between electrical to optical and back to electrical).

 

Spread spectrum, posted on April 6, 2007 at 04:53:23
gbeard
Reviewer

Posts: 646
Location: Indiana
Joined: April 21, 2000
 Hi Chrisine,

I saw this option in my BIOS. What is its effect if enabled?

Thanks,
Gary

 

Flash, posted on April 6, 2007 at 07:56:25
jbmcb
Audiophile

Posts: 1348
Location: Southeast Michigan
Joined: August 7, 2001
 >For example, you can build a totally fanless PC (no fan in power supply, passive cooling etc.) and solid state hard disk (eg. high capacity compact flash or similar). The actual music files can be stored elsewhere (ie. on a server or NAS).

Careful when using flash disks, depending on your OS and configuration you can kill a flash drive very quickly, as it's only good for a few hundred thousand writes before it fails. This is fine for music files and pictures, not so much for OS temp files, the registry, the MFT, or swap.

Linux has no problems with a read-only boot drive, and even has special filesystems available that reduce the likelyhood of killing writable, bootable flash devices. Windows tends to hate it. There's a utility called BartPE that lets you build bootable windows CD's, you could probably use it to build a read-only filesystem that could boot from flash.

If you are going for a fanless solution, check out the new low powered processors from Intel and AMD. They vary their clock speed depending on load, so they only use as much current as needed to perform a given task. These are probably well suited to using a fanless heatsink solution.


/*Music is subjective. Sound is not.*/

 

Don’t play music from USB drives and Network, posted on April 6, 2007 at 10:52:57
cics
Audiophile

Posts: 1320
Joined: November 9, 2006
 I want to clarify this subject further on why they should be avoided. This also applies to Windows on USB flash drives. My insights are as follows:

1. Best design is when soundcard has lions share of PCI bus use.
2. All soundcards (USB, Ethernet or PCI) operate of PCI bus.
3. USB drives and Network Drives (which also connect to PCI bus) goes against this design principle as data traffic now competes for PCI bus use.

PCI read or write transactions take place between master and slave devices (soundcard, CPU…). All devices connected to PCI bus can act as master or slave. During a read or write transaction, master takes control of bus and STOPS all other devices from transacting. This is where a conflict arises – soundcard wants more data for output BUT PCI bus is busy with a data transaction (from USB drive or Ethernet) thus forcing soundcard to wait. This all happens at frantic speeds but for ultra low latency playback, this conflict will have impacts.

Hence, I prefer internal laptop SATA drives where mobo maintains dedicated bandwith for data flow. A further improvement is to use SATA RAID 0 which does away with IDE – this is my current setup using 2 Seagate 5400.3 160GB drives.

 

Check out something called "EWF", posted on April 6, 2007 at 15:10:09
Christine Tham
Reviewer

Posts: 4839
Location: Sydney
Joined: December 29, 2001
 It's a driver originally written for Windows XP Embedded, but you can make it work with normal Windows XP. It "protects" your flash drive by making it "read only" - all writes are buffered to memory, and then discarded when you shut the PC down.

As a bonus (because the flash drive never gets written to) you never need to defragment it, and it's also protected from viruses :-)

 

Re: Spread spectrum, posted on April 6, 2007 at 15:18:44
Christine Tham
Reviewer

Posts: 4839
Location: Sydney
Joined: December 29, 2001
 depending on the BIOS (most only support spread spectrum on AGP or PCI Express, not memory or PCI - so far i have only seen one motherboard that allows spread spectrum on the latter) it can spread the generated EMI across a wider frequency spectrum, which can reduce the "damage" on audio circuits. Note it doesn't actually reduce EMI, just spreads it "thinner".

If you can, try "underclocking" - that will actually reduce EMI, but be careful, you may create a non-bootable PC (in which case you need to reset the CMOS).

You could also try EMI shielding, for example, wrapping the audio card and ribbon cables using aluminum foil (grounded). Be careful about creating short circuits or unintentionally trapping heat.

 

Re: Check out something called "EWF", posted on April 6, 2007 at 15:27:38
Dawnrazor
Audiophile

Posts: 12794
Location: N. California
Joined: April 9, 2004
 Hi Christine,

How would a mere mortal do this.

I have shared your prior posts on EWF with any "computer geek" I could from professional IT people to my "I'm a PC builder/gamer" friends.

No one seems to get this.

None of them have heard about the write limitations of flash drives, don't seem to believe me when I mention such limitations, and don't seem to be able to work with EWF.

What do I tell them to get them to do this?

 

Not necessarily, posted on April 6, 2007 at 15:31:34
Christine Tham
Reviewer

Posts: 4839
Location: Sydney
Joined: December 29, 2001
 Your advice of keeping the PCI bus relatively "free" to avoid contention with the sound card is good. By the way, the goal is not necessarily to have "ultra low latency playback" - it's to have low jitter. In fact, typical solutions to reduce jitter often increases latency (ie., memory buffering). Fortunately, latency is not a problem when you are just listening to music (it is however an issue if you are watching video at the same time).

However, depending on the motherboard chipset, USB may or may not go through the PCI bus. For example, on newer chipsets, USB support may be handled through a PCI Express bridge rather than PCI.

Also you don't need to use USB devices for flash. For example, you can use a compact flash card and a a CF to IDE bridge and connect it directly to your IDE port.

A good soundcard should have some internal buffering and also act as a PCI bus master. So hopefully it should not be impacted by other PCI devices (provided it gets allocated a dedicated PCI interrupt line). Generally, the only device that tend to cause problems for PCI sound cards is graphics cards that hog the PCI bus (by having excessively high PCI bus latency). Fortunately, this should no longer be a problem with PCI Express x16 cards. Also, motherboard Gigabit Ethernet chipsets tend to bypass PCI as well.

 

Google is your friend, posted on April 6, 2007 at 15:33:41
Christine Tham
Reviewer

Posts: 4839
Location: Sydney
Joined: December 29, 2001
 This link below, which I found via searching, may be useful.

 

Re: Google is your friend, posted on April 6, 2007 at 15:43:33
Dawnrazor
Audiophile

Posts: 12794
Location: N. California
Joined: April 9, 2004
 THANKS.

When I did use google, it only took me to Microsoft sites, and they were useless.

Thanks a bunch!!!

 

By the way, if you do follow the instructions ..., posted on April 6, 2007 at 15:49:53
Christine Tham
Reviewer

Posts: 4839
Location: Sydney
Joined: December 29, 2001
 Be careful about replacing Winlogon with Minlogon if you intend to use the PC in a networked environment. My advice is don't do it unless you absolutely know what you are doing and understand the consequences.

And an easy way to avoid Windows File Protection errors if you do want to use Minlogon is to replace not just the copy in the windows directory but also the "safe" copy (best to search for all occurences and replace all of them). You need to do this after booting Windows in "Safe" mode.

Finally - a warning when using the hibernate once reboot many feature. Many pro audio soundcard drivers (unfortunately, including mine) don't support hibernation.

You should be able to get EWF working reasonably easily. If in doubt, try to get it working on a normal (non flash) hard drive first. Once you are confident (and you can turn writebacks on and off with no problems), then follow the instructions to create a bootable flash drive.

I use a 4GB compact flash card with a flash to IDE bridge, so I don't have to do anything special (most flash readers mark the media as "removable" which creates issues with installing XP - there are various ways of getting around this - easiest is to install on a normal hard disk, then image copy to a "removable" flash drive).

Good luck! The reward is you may find that your system sounds much better when you take out the hard disk. Mine sounded noticeably "cleaner" and more "dynamic" (previously I was using a laptop hard disk).

 

BTX motherboards, posted on April 6, 2007 at 18:02:25
laznos
 My music computer has a motherboard with a BTX form factor. They are designed for more efficient heat dissipation. While not a "silent PC", I can't hear it from 8 feet away (where I sit), so it may as well be silent.

Unfortunately, they have been discontinued but you can still find the boards and new computers with them.

 

Re: By the way, if you do follow the instructions ..., posted on April 6, 2007 at 19:21:12
Dawnrazor
Audiophile

Posts: 12794
Location: N. California
Joined: April 9, 2004
 Thanks for the added info C,

I am already a bit confused.

THe nice link refers to a trial version, but it is only good for 120 days. Is it good enough to just use it once, or do I need to buy something?

 

Re: Check out something called "EWF", posted on April 6, 2007 at 19:56:20
Bubba


 
There's been a lot of discussion on optimizing Windoze in the forums at MP3car.com They're not concerned with ultimate audio fidelity so much as quick booting, but I have seen ASIO mentioned, and there've been discussions on running from flash memory.

 

Re: Not necessarily, posted on April 6, 2007 at 20:28:27
cics
Audiophile

Posts: 1320
Joined: November 9, 2006
 There is a connection between latency jitter (variations in latency) and transmitted jitter. I found that reducing latency helped in overall sound quality. Why this is so I don't understand.

On USB, yes if mobo has dedicated USB bandwidth then you would avoid conflict described earlier.

On IDE, I found its best to avoid this interface as it lends itself to electrical noise pickup. Also, using Windows IDE drivers is not an optimal choice. However, I'm keen to use CF technology implemented as a disk drive. No moving parts and even less power! Price is high and storage capacities low but as most pc harware goes, this will improve very quickly.

 

Re: Not necessarily, posted on April 6, 2007 at 21:11:53
Christine Tham
Reviewer

Posts: 4839
Location: Sydney
Joined: December 29, 2001
 *** There is a connection between latency jitter (variations in latency) and transmitted jitter. ***

That could well be true. But I'm not sure what you mean by latency - do you mean bus latency, or audio delay latency?

As I've pointed out before, *increasing* audio delay latency (through buffering) actually *reduces* jitter.

*** I found that reducing latency helped in overall sound quality. ***

If you mean bus latency, then possibly. Although ThomasPf and I disagree on this - I feel it could make a difference, Thomas seems to think (and I'm paraphrasing him) as long as the buffer on the soundcard never becomes starved, PCI bus interrupt processing should have a negligible effect.

If you are interested in reducing bus latency, you may want to consider using Vista with a WaveRT audio driver - Microsoft promises that WaveRT has much lower latency than either WavePCI or WaveCyclic (which are the traditional models for audio drivers used in XP).

*** On IDE, I found its best to avoid this interface as it lends itself to electrical noise pickup ***

That's probably due to EMI generated by the IDE cable. You can take various steps to reduce this. For example, you can try shielding the IDE cable.

Since you are interested in latency, are you aware that some SATA drivers generate more interrupt processing latency than IDE? There have been cases of SATA drivers causing audio/video stutter. So in that respect, I'm surprised that you consider IDE to be non-optimal.

*** I'm keen to use CF technology implemented as a disk drive. ***

I'll be keen to find out whether you notice any differences switching to CF. I've noticed an improvement getting rid of my laptop hard drive and using a CF card. I've also noticed an improvement applying shielding to the audio card.

Pricing is pretty good - I bought a 4GB CF card for around $100 and you don't really need more than 4GB (provided of course the music is stored elsewhere on the network).

 

The "trial" period is for Windows XP Embedded ..., posted on April 6, 2007 at 21:18:06
Christine Tham
Reviewer

Posts: 4839
Location: Sydney
Joined: December 29, 2001
 ... however, all you need is the EWF driver (I think it's on 3 files which you need to extract).

The driver itself, once you install on normal Windows XP, does not carry any time limitation.

Let me know via email if you get really stuck trying to extract the files, I can probably help you out there.

 

Thanks, posted on April 6, 2007 at 21:46:52
Dawnrazor
Audiophile

Posts: 12794
Location: N. California
Joined: April 9, 2004
 THat is such a nice offer, one I hope I don't need to take you up on :).

 

Re: Don’t play music from USB drives and Network, posted on April 6, 2007 at 22:45:28
artcheng@netvigator.com
Audiophile

Posts: 388
Joined: November 14, 2003
 Interesting! Is this the reason why playing the music through Ramdisk gives better results?

 

Re: Flash, posted on April 6, 2007 at 22:51:41
artcheng@netvigator.com
Audiophile

Posts: 388
Joined: November 14, 2003
 I will wait for this

http://www.theinquirer.net/default.aspx?article=23425

 

Re: Don’t play music from USB drives and Network, posted on April 6, 2007 at 22:55:40
artcheng@netvigator.com
Audiophile

Posts: 388
Joined: November 14, 2003
 Dies firewire operates via PCI bus as well?

 

Re: Flash, posted on April 6, 2007 at 23:08:27
artcheng@netvigator.com
Audiophile

Posts: 388
Joined: November 14, 2003
 More options here

http://www.dvnation.com/nand-flash-ssd.html

 

Yes., posted on April 7, 2007 at 02:04:46
cics
Audiophile

Posts: 1320
Joined: November 9, 2006
 .

 

Could be - not sure., posted on April 7, 2007 at 02:21:59
cics
Audiophile

Posts: 1320
Joined: November 9, 2006
 Test this by not using Ramdisk and force Foobar to play from system RAM. If sound is same as with Ramdisk then this could very well be the reason. Otherwise its Ramdisk.

Set Foobar's full file buffering to say 800,000KB (this must be larger than your .wav file). Run your .wav file with foobar as program (you may need to right-click on your .wav file and use 'Open with'). Foobar will load entire .wav file into RAM (10-15 secs) and start play. Let me know what you get.

 

Re: Could be - not sure., posted on April 7, 2007 at 02:32:01
artcheng@netvigator.com
Audiophile

Posts: 388
Joined: November 14, 2003
 Will try. Unfortuately, I am in the process of upgrading to a new PC so this will take awhile!

 

Re: Not necessarily, posted on April 7, 2007 at 03:19:40
cics
Audiophile

Posts: 1320
Joined: November 9, 2006
 On latency/jitter stuff I suspect its a combination of both. I'll need to do more research. For now I want to test the Zalman TNN-300 case - it does a few things right and PSU may even better Enermax. I use a basic mini tower case with heavy non-metallic object on top (the added inertia helps). For EMI shielding, I use Quantum Physics noise disruptor on soundcard (other methods risk short circuiting).

IDE/SATA. Yes you can buy shielded IDE cables and it would help. I still prefer SATA as you can disable IDE. I found playback latency to be important - SATA latency has no impact (as long as Foobar is never starved of data and drive has at least 8MB cache). Makes sense on video stutter as video application demands factors more data from disk. SATA RAID 0 would solve this.

BTW - On toslink, yes DAC should do reclocking - I use AA Prestige SE. I like your point on under-clocking. It got me thinking on how to do this permanently with Core 2 Duo processor. This processor is great as it optimises for low power (and therefore lower voltage, lower RF radiation, harmonics) - current suggested setup in paper does this. But during playback, clock frequency hops to higher levels to its max of 1.86GHz. Fortunately these (voltage) transitions occur infrequently. But getting it to stay at its lowest level permanently is best - I'll post a solution.

 

Setting for permanent underclocking, posted on April 7, 2007 at 03:49:01
cics
Audiophile

Posts: 1320
Joined: November 9, 2006
 Thanks to Christine Tham for highlighting this. In paper, Core 2 Duo CPU is set for low power use and this results in lower clock speeds (underclocking). This can be improved by keeping processor permanently set at lowest clock rate. Currently, CPU does step up but is mostly at lowest level.

In section 4.4 (Power Schemes), just set Power Schemes to 'Portable/Laptop' with all else set to 'Never'. Windows sets CPU to permanently underclock. Note BIOS setting of 1.325V is not used and no BIOS changes are necessary.

Technical Background: this setting keeps CPU V-core (voltage core) to lowest level of 1.13V). This gives constant lower clock speed of 1.6GHz which means less RF radiation (& its harmonics) and uses less power. Previous setting has V-core between 1.15V and 1.28V with clock speeds ranging from 1.6GHz to 1.86GHz. You can see all this happening by installing supplied mobo 'doctor' utility (program installs as ITE Smart Guardian). No need for clocking utilities but if you want to confirm clock speed then use RM Clock Utility (see link).

 

Setting LCD/plasma screen resolution, posted on April 7, 2007 at 09:14:37
cics
Audiophile

Posts: 1320
Joined: November 9, 2006
 Under 4.3.b there is an important point to note on Screen Resolution which was not explained. This is deliberately set to 1280x1024. Whilst this goes against principle of reducing video traffic (as lower resolutions are better), its set to work best with LCD or plasma.

All LCDs & plasmas have a native (or default) resolution. Latest model LCDs is 1280x1024. Plasmas will differ. Windows Screen Resolution must match this native resolution. If your LCD panel has a different native resolution than change Windows to this native resolution, e.g. 1024x768 or 1280x768. Recyle power on computer (reboot is not enough) for change to effect LCD/Plasma.

Why? Both LCDs and Plasmas put back noise into your AC circuit which can be audible in a high resolution system. The most offending piece inside these monitors is the Display Processor. It detects input resolution and scales it to native resolution. This is a video DSP chip. Display Processor is bypassed by setting your computer to the native resolution.

 

Re: Setting for permanent underclocking, posted on April 7, 2007 at 09:51:44
Dawnrazor
Audiophile

Posts: 12794
Location: N. California
Joined: April 9, 2004
 Thanks for the wonderful documentation on all this!

WOuldn't it just be easier to start with a slower cpu in the first place, or is there an advantage of underclocking a faster processor than just using a slower processor?

 

Re: Setting for permanent underclocking, posted on April 7, 2007 at 12:28:04
cics
Audiophile

Posts: 1320
Joined: November 9, 2006
 Key things needed from CPU are: low clock speed, low power consumption and excellent throughput (workrate). Its the last bit that makes it very difficult. So while you can get low speed & sometimes with low power, without throughput, correct 24/96 upsampling suffers. SRC upsampler is process intensive. For 24/192, processing loads grow much more and its always good to have plenty of headroom to do this, hence Core 2 Duo.

Some of Intel's Core 2 Duo for mobile looks very good too but its more expensive. Would need to buy a laptop (which has some advantages over a desktop but suffers in other areas).

 

Re: this is worth at least $600, posted on April 7, 2007 at 15:42:51
aljordan
Audiophile

Posts: 1252
Location: Southern Maine
Joined: November 4, 2003
 Hi bobp,

Have you found a desktop like case that supports 120mm exaust fans? I have only found tower style cases that do. I built my media pc using a Silverstone Grandia GD01 case. It is one of the larger of the HTPC style cases, but it has room for 80mm exaust fans. It is interesting that it will allow 92mm intake fans but only 80mm exaust fans.

If youknow of any good looking HTPC style cases that allow for 120mm fans, please forward them.

Thank you,
Alan

 

Re: underclocking, posted on April 7, 2007 at 16:09:02
aljordan
Audiophile

Posts: 1252
Location: Southern Maine
Joined: November 4, 2003
 Hi cics,

For underclocking or undervolting a Core 2 Duo chip, try a program called RMClock by Rightmark (see link below). For underclocking an AMD 64, try CrystalCPUID. The standard version of RMClock is free.

RMCLock is easy to use and excellent. You can specify four different modes of operation and customize the CPU frequency and throttling of each. One mode, called Performance On Demand, will automatically jump between the modes as needed. You can also lock to a specific power saving mode if you prefer.

I have found that the RMClock utility makes a cooler running system. In combination with the Speedfan utility, fan speeds can automatically be knocked down according to system temps, allowing for a quieter system as well.

 

Re: underclocking, posted on April 7, 2007 at 16:12:23
aljordan
Audiophile

Posts: 1252
Location: Southern Maine
Joined: November 4, 2003
 oops, just saw that you already found it!

 

I'm not sure I understand you, posted on April 7, 2007 at 19:37:45
Christine Tham
Reviewer

Posts: 4839
Location: Sydney
Joined: December 29, 2001
 *** I still prefer SATA as you can disable IDE. ***

Since SATA has higher bus latency than IDE why not disable SATA?

*** SATA latency has no impact ***

In that case, why worry about PCI bus usage at all?

I'm not sure why you would be worried about USB and Ethernet PCI usage (when these devices typically don't impact the PCI bus) but then say a device that has latency high enough to potentially cause audio/video stutter has "no impact"?

*** SATA RAID 0 would solve this. ***

I'm not quite sure how using RAID 0 can reduce the bus latency of a SATA chipset? Wouldn't it actually *increase* latency (because the chipset now needs to do more work for each I/O request)?

*** On toslink, yes DAC should do reclocking - I use AA Prestige SE ***

If the DAC is reclocking and resamping the SPDIF input, then why optimize the PC at all (since any jitter in the input signal to the AA will be discarded by it's resampling)?

Also, why upsample in Foobar using SRC? You'll end up triple filtering the audio (SRC upsampling, and another layer of resampling by the AA Prestige, followed by final filtering in the DAC itself). Each filter will add an additional set of artefacts to the signal.

 

120mm fans, posted on April 7, 2007 at 22:27:31
bobp


 

Alan,

I have tried several cases and finally settled with Antec NSK 2400. It is less than $100, steel and plastic but it is built with audio server in mind. An outstanding piece. Two side 120mm fans.

It seems that the guy from Silentpc.com worked as consultant for Antec.


There is one souped up model with display for $100 or so more, perhaps the best deal.


 

Re: The art of building Computer Transports, posted on April 8, 2007 at 05:10:52
fmak
Audiophile

Posts: 13158
Location: Kent
Joined: June 1, 2002
 Your paper is interesting and you have obviously spent a lot of time 'optimising' your system as have I.

I too at first find my fanless Commell system with Lynx AES 16 card 'as good as' my dCS 972/954 setup. However, having bought an all class a capacitorless adc/dac (Universal 2192) and clock (Apogee Big Ben), and integrating the compter versus fixed transport boxes into the same replay chain (Palcette Audio constant impedance balanced passive) into Accuphase A50 amp, I can now confidentaly say that my fixed boxes are better. The problem with the computer system seems to be due to the poor clocking from the sound card (this is generally true no matter what is said; the XOs used are are not great)and noise and radiation within the system. This is inspite of the claim by Lynx that their clock is superior.

With the Big Ben as Master clock and dejitter playing into the dCS 972 at 4x upsampling, I am getting 'better than analog' playback
in terms of imaging, superior s/n and superior music. The Lynx system is now quite inferior, although I have not yet clocked the card using the Big Ben.

The problem with computer audio is the multiplicity in setting up parameters. There are many claims on what the optimal settings are, and a lot of hot air about jitter. But there has never been meausrements to back these up, except for some measurements of sound cards and the slim devices streamers in Stereophile. There is also no ransparency in what software or their plug-in do. For example, Foobar can be made to sound quite different dependig on what version of the ASIO plugin is used. I am also not convinced that SRC sounds better all the time than SSRC.

So, my issue is with validity and generalisation of the superiority of computer audio. I do not dispute that this can be made to sound good, as I have done using many of the techniques that you mention. But I maintain that your conclusion about them being superior to high end systems is invalid; you have merely chosen to optimise your computer system, not optimise both in parallel and compare as you develop. If this is not true, please post what you have done in comparing systems side by side with the same replay chain and I would like to learn from you.

ps: I am a professional engineer qualified in acoustics and measurement and control, and computer literate

 

Re: The art of building Computer Transports - some questions, posted on April 8, 2007 at 05:28:13
aljordan
Audiophile

Posts: 1252
Location: Southern Maine
Joined: November 4, 2003
 Hi cics,

Thank you for the post. Could you speak further about some of the points that I pulled out of your paper?

"Critical that one module is used. 2 x 512MB is bad. Multiple modules often cause blue screen crashes"
Why is it critical that a single module is used? Is it only to avoid the blue screen?

"Remove all unnecessary cables within PC - no external USB connectivity hubs, firewire, and unnecessary sound cables."
Do the internal cables cause some sort of interference or do you suggest removing them to reduce clutter and help air flow?

"Remove keyboard. Only connections should be the VGA cable and mouse (either system or USB)."
What affect does keeping a keyboard hooked up have?

"Remove CD ROM Drive. Apart from high EMI, it uses IDE connectors. For CD ripping, use home computer and copy files via USB (stick or drive). For CD based software installations, connect CD Drive temporarily (and enable IDE Channels as described in BIOS setup)."
Does a DVD/CD drive still emit EMI when no CD is in the drive and it is not spinning? Just a note, SATA optical drives are becoming quite common.

"Select Shadow Setup. Disable Video BIOS Shadow"
Video BIOS Shadow copies the ROM to a normally hidden area of RAM and write protects that region. It allows for much faster access to the ROM code. How or why does disabling it improve sound quality?

Also, can you speak to how or why setting lower latency in the sound card drivers achieves far superior results in sound quality?

Thank you,
Alan

 

Re: I'm not sure I understand you, posted on April 8, 2007 at 07:55:02
cics
Audiophile

Posts: 1320
Joined: November 9, 2006
 Of all latencies in computer, I find reducing playback latency to have a large impact on sound. Sound delivery happens in realtime, hence preference for soundcard having lions share of PCI bus.

On SATA/IDE, SATA latency may be worse than IDE (but biggest latency is at hard drive irrespective of interface @ ~5-6ms). Of course, you raise a valid point why this latency is less important?

My thinking: disk read latency happens only once at start of play where Foobar through Windows (through interface, data bus, disk controller, cache and finally disk) establishes a data read stream. Since this sequential read stream rate (@ 42.9MB/s for Seagate 5400.3 160GB) is factors more than what Foobar needs there is minimal effect (as disk data is available to Foobar ahead of time –Windows/disk controller prefetch & block reads occur). For example, on my setup, an hour long CD (600MB) takes ~10 seconds to load into RAM, yet playback takes an hour.

Let me preempt your question here, 600/42.9 = 14 seconds, how do you get to ~10 seconds? RAID 0. File data is stripped sequentially across drives thus on read, you get both drives performing sequential read in parallel. In theory you should get double throughput but some serialization occurs and throughput improves by ~30-50%. RAID 0 does away with Windows IDE drivers (and all its overheads including SATA in IDE mode) which I found beneficial.

Most optimal is when NO disk reading occurs during playback. By this I mean entire CD .wav file is read into RAM (using Foobar’s full file buffing). There is a bug when playing .cue files where foobar ignores file buffering – playing .wav files, buffering occurs (and 5-15 seconds elapses for playback to start). Hopefully this bug gets fixed in next release. This way, disks go into idle state consuming 40% less power (meaning less EMI, etc.).


*** If the DAC is reclocking and resamping the SPDIF input, then why optimize the PC at all (since any jitter in the input signal to the AA will be discarded by it's resampling)? ***

There is good sonic improvement. Like all DACs I suppose the one I’m using is not perfect. Others have found this too with DACs that do internal jitter management.

*** Also, why upsample in Foobar using SRC? ***

To bypass DAC’s inferior 16/44.1 -> 24/96 upsampling. DAC further upsamples to 24/192 but most ‘damage’ occurs on former. DAC sees 24/96 feed as native input (with music information to 22.05kHz only).

 

Re: The art of building Computer Transports, posted on April 8, 2007 at 08:07:11
cics
Audiophile

Posts: 1320
Joined: November 9, 2006
 A similar dialogue on why should computer transports be superior happened at Audiogon. I'm not going to repeat here - see link.

 

1 x 1024MB vs 2 x 512MB, posted on April 8, 2007 at 08:22:56
cics
Audiophile

Posts: 1320
Joined: November 9, 2006
 Yes, to avoid blue screen crashes. Also, reduce power consumption & EMI/RF radiation - you get audible improvements.

 

Remove unnecessary cables within PC incl keyboard, posted on April 8, 2007 at 08:30:09
cics
Audiophile

Posts: 1320
Joined: November 9, 2006
 These cables add unwanted EMI/RF to computer. Most of these cables are not shielded and therefore act as antennas. Keyboard and things like USB hubs consume power (albeit little, every reduction helps).

 

DVD/CD drive still emit EMI when no CD, posted on April 8, 2007 at 08:40:05
cics
Audiophile

Posts: 1320
Joined: November 9, 2006
 Yes, because power lines are active to your drive and some work is done. Also, IDE data bus and interface is prone to EMI/RF. Data bus acts like antenna.

On optical SATA, I suspect similar problems will occur regarding power. SATA cables are shielded.

 

Disable Video BIOS Shadow, posted on April 8, 2007 at 08:45:28
cics
Audiophile

Posts: 1320
Joined: November 9, 2006
 You want to reduce video chipset to system memory traffic and free up precious memory in Windows. Adding more memory is a bad idea as more power is used (and of course this comes with added EMI/RF).

 

Why setting lower latency in the sound card achieves far superior results?, posted on April 8, 2007 at 08:59:11
cics
Audiophile

Posts: 1320
Joined: November 9, 2006
 Its been raised by Christine Tham as well - see comments.

Basically: There is a connection between playback latency jitter (variations in latency) and transmitted digital audio jitter. I found that reducing playback latency helped in overall sound quality. Why this is so is not yet understood. You'll find many comments on ASIO being better than KS - playback latency drops with ASIO (assuming for KS, soundcard playback buffers weren't reduced).

 

Re: The art of building Computer Transports, posted on April 8, 2007 at 09:08:04
fmak
Audiophile

Posts: 13158
Location: Kent
Joined: June 1, 2002
 Nobody is asking you to repeat yourself. You made ascertions. I am asking you to clarify whether you have made direct comparisons between high ens and your computer.

It seems that you have not. I have, and I do not agree with your generalised blanket statement.

 

Re: The art of building Computer Transports, posted on April 8, 2007 at 10:50:51
cics
Audiophile

Posts: 1320
Joined: November 9, 2006
 On direct comparisons, do you mean something like this (as stated by kana@audiogon):

"Unless you test all the transports you listed and your PC transport with the same DAC, your statement that PC transports lead to a new level of digital audio performance is just a guess."

My response (I have expanded a bit):

Not sure of that methodology being the correct one, as systems that I auditioned were well setup and optimized within entire ‘system’. There is much to do with cable synergy, power conditioning, power cords, room acoustics, to preamp or not, etc..

Nevertheless, many people are finding computer transports to yield better performance – this is not something new. Look at the Nova Physics Memory Player (link below), which is a computer transport but costs $10k. Its getting rave reviews, e.g. Positive Feedback or Stereo Times. Side by side comparisons are done (as in direct comparisons) against Esoteric P03 and Zanden Stack where the Memory Player was preferred. My problem with the Memory Player is its proprietary architecture (no open standards like .cue files and choice of playback software, e.g. foobar2000 or winamp) and cost (similar performance can be had at much less).

 

Re: The art of building Computer Transports, posted on April 8, 2007 at 11:25:18
fmak
Audiophile

Posts: 13158
Location: Kent
Joined: June 1, 2002
 No. You say you have listened to dCS etc. You don't say if you have compared a dCS system with a computer as transport, keeping the rest of the system the same.

It is the attention to detail which ultimately determine the fidelity of a system; clock, cable, etc all come into it.

You cannot say that any 'decent' sound card in a computer system with all your tweaks is better than high end systems as a general rule

 

Re: I'm not sure I understand you, posted on April 8, 2007 at 13:09:03
KeithC
Audiophile

Posts: 94
Location: Sale England
Joined: December 2, 2002
 Two immediate points:

1. Raid - unless you have a premium raid card, raid support is software based.

2. Not all disk activity will be related to audio file i/o.

 

Re: The art of building Computer Transports, posted on April 8, 2007 at 14:14:21
cics
Audiophile

Posts: 1320
Joined: November 9, 2006
 I see. The methodology you are proposing is not correct to determine superiority of computer transports over high-end traditional CD transports. I’ll explain later.

Methodology proposed: keeping rest of system same whilst CD transport is changed to computer and comparing results. I have summarized this in my initial remarks introducing this topic.

Here is the details. Extensive testing and comparisons over several months was done against the AA Prestige SE internal transports. Initial Philips transport (as used in dCS Verdi La Scala & Encore and EmmLabs transport) was upgraded to a Sony (which I believe is also being used by dCS, EmmLabs and Wadia replacing Philips and Teac VRDS).

My findings on this were posted on Audiogon. Here are the relevant comments:

"... Philips offered overall good sound with SACD being its strong point. The Sony has improved CD playback by a good margin with more detail and a blacker background making for even better musicality. On SACD, its pretty much the same. It also operates faster and offers text messages on SACD (Album, Title, etc.)...

Using a Computer Transport, feeding 24/96 via toslink (glass fibre) to player (DAC), I get even better results! Improvement over Philips drive was very obvious but less so with Sony. Computer over Sony drive yields more clarity on low level information and tighter bass. However, setting up such a computer transport was not easy (and cost me ~$2000 including $450 toslink cable).

Things got very interesting when I upgraded interconnects to Synergistic Research's new Tesla Apex (from SR Absolute Ref). Active shielding is done by Quattro unit powered by SR Abs Ref Power Cord. ...

As I suspected, improvements between transports widened significantly. Now, the computer transport yielded much more information over the Sony (and the improvements are very obvious). Things like dynamics (both macro & micro), fleshing out inner detail of instruments, bass and imaging are just splendid. I am amazed at how much music is stored on CDs! ..."

On the example of dCS, here’s a recently received email quote from a Purcell / Elgar owner who shares his experience on computer transports as follows: "The result is wonderful and better than any transport that I have heard.".


Whilst this establishes computers relative strength of over CD transports, as a methodology to suggest computers are superior to highend transports is not enough. Instead, I prefer the additional approach where high-end CD transports are auditioned in their optimal respective systems (with the requisite attention to detail you speak of). Hence initial remarks on this topic: “Having the pleasure to listen to high-end transports (Wadia 270SE, Esoteric P03, MBL's $20,000+ 1621A, full DCS stack), I am comfortable in saying that Computers have outclassed them.”

 

Re: The art of building Computer Transports, posted on April 8, 2007 at 22:41:17
fmak
Audiophile

Posts: 13158
Location: Kent
Joined: June 1, 2002
 Sorry, I respect your views but the methodology you post is Voodoo stuff. One simply cannot take a Philips transport and use Toslink to validate differences. Also, what is the optimal setup for a particular box? There are as many variables as as in your computer setup.

It seems to me that there are just too many variables in your paper to make valid ascertations in general.

I have over several years developed my fanless low power computer system in parallel with my hifi. My conclusion after many iterations is as follows.

'An optimised HD system can be better than a non optimised high end system but an optimised high end system is even better. The problem with an HD system lies in inferior hardware and software limitations over which the user have little control'

The Audiogon stuff you posted doesn't allow anyone to make a valid judgement. Also jitter is not properly understood in postings on computer ausio and this is a major issue not caused merely by drives and interfaces, but also power supply and data transfer issues.

Finally, you must know that a cumputer with a fan degardes the effective signal to noise ratio from say, -105 dB to perhaps -70 to 80 dB unless you isolate the computer. Even then, the vibrations in an HD can be very significant by way of microphony!

 

Hardware RAID is implemented, posted on April 8, 2007 at 22:49:12
cics
Audiophile

Posts: 1320
Joined: November 9, 2006
 Recommended Biostar mobo implements SATA RAID at hardware level. Check out the VIA chipset it uses (link below). No need to get additional RAID hardware if using this mobo.

Once BIOS is set to SATA RAID, Windows installation CD will not recognize disk. VIA chipset driver for Windows needs to be supplied on floppy.

 

Design goal is to elliminate all IO overheads, posted on April 8, 2007 at 23:04:53
cics
Audiophile

Posts: 1320
Joined: November 9, 2006
 Recommended Windows settings as set out achieves this by:

1. Having no paging file (no VM)
2. Stopping all unnecessary services (task scheduler, disk indexing, etc.)
3. Remove unnecessary programs (many sw installations do sneek in startup actions)
4. Remove Windows components (disk indexing, messenger, etc.)
5. Only have foobar in startup
6. Perform boot only prefetch (and deleting all contents of c:\windows\prefetch)
7. Install no virus protection sw

Test this and let me know what you get. After bootup, with Foobar started but not playing, there should be no disk activity whatsoever no matter how long system idles. Only disk activity would be the disks themselves (Windows has no control over this), e.g. parking drive heads and thermal recalibration.

 

It is working but upsampling doesn't seem to buy me anything., posted on April 9, 2007 at 01:42:42
Ugly
Audiophile

Posts: 2942
Location: Des Moines, WA
Joined: August 22, 2006
 the best for each I could get in my system was: ASIO - 32bit 48kHz, kernel streaming - 32bit 96kHz and direct sound - 32 bit 192kHz all with SRC. I couldn't tell any of them apart from 32 bit 44.1kHz sound quality wise, but I am thinking this has to do with the fact that I am sending my audio from PC to my reciever via toslink and my reciever has this "24bit 192kHz Digital Remastering" feature so my reciever is essentially making all the upsampling in the PC redundant. That would be cool by me though since doing 32bit 192KHz pretty much made the thing nearly useless for other stuff since the sound got all stuttery if I tried anything much. At any rate the whole upsampling thing didn't buy me a darn thing.

I really dig Foobar and EAC though and am totally getting a mic and preamp this week so I can try this system and room correction filter convolved with music via convolver plugin for foobar. That is some sweet stuff there that I finally decided to pull the trigger on trying this stuff when your guide led me to try foobars upsampling how-to so thanks man.

I spent all weekend getting started on re-ripping my CD's onto my hardrive in lossless compressionless images using EAC which is also new to me from your guide. I have to admit that these EAC ripped files played through my PC and optical link to my reciever sounds mighty nice compared to the lowend old skewl CD player with its analog RCA wires from high school that I never use any ways. This DRC stuff might make it worth going for the best quality rips even though they take up some room.

PS which one is best between ASIO, kernel streaming, and direct sound and why?

 

Re: It is working but upsampling doesn't seem to buy me anything., posted on April 9, 2007 at 02:20:29
artcheng@netvigator.com
Audiophile

Posts: 388
Joined: November 14, 2003
 I thought a single toslink cable cannot transmit anything higher than 16/44? Or am I confused with ADAT?

 

Re: It is working but upsampling doesn't seem to buy me anything., posted on April 9, 2007 at 07:19:32
Ugly
Audiophile

Posts: 2942
Location: Des Moines, WA
Joined: August 22, 2006
 Beats me but that would help explain the huge CPU load if I am upsampling and then downsampling back to where I started all in the same move. It would also be an alternate way of explaining why I can't hear any difference.

 

I completely ignored the case and fans and couldn't be happier, posted on April 9, 2007 at 11:13:18
pburke
Audiophile

Posts: 2558
Location: Madison Wisconsin
Joined: September 9, 2002
 you guys are soooo not thinking outside the box...

I have a pretty regular case, 80mm noisy fans, etc - but I just put the darn thing in the basement, remote KVM to the listening room via 50 foot Cat5 cable and things couldn't be quieter here.

My entire listening box was less than $600 and it's working just fine on a 3 year old CPU (at 2% CPU with foobar running) and a fully isolated battery powered DAC (fiber optic USB extender)

Basically, I decided to forego the PC tweaking - get the bits from it and put the money on isolating it from the listening room equipment.

just food for thought... don't have to do it that way, but I can imagine worse solutions.

 

Re: It is working but upsampling doesn't seem to buy me anything., posted on April 9, 2007 at 22:18:41
Ugly
Audiophile

Posts: 2942
Location: Des Moines, WA
Joined: August 22, 2006
 I did some reading and found out my motherboards codec is stuck at 48kHz spdif output rate, I assume they were talking about 16bit data but that wasn't perfectly clear. I did however note that my codec also supports multiple sample rates on its spdif input.

Ill let the codec do the upsampling and keep those host cpu cycles for something better.

 

Re: Setting for permanent underclocking, posted on April 9, 2007 at 23:49:20
swarmer


 
No need to buy a laptop. You can buy core duo and core 2 duo mobile processors and compatible Micro-ATX motherboards on newegg.com (among other places, I'm sure). It does limit your motherboard selection a bit. Should be almost as fast as the desktop core 2 duo, but cooler & lower power. That's what I would try using for this type of project. Apple's Mac Mini is an example of a desktop system using a Core Duo chip originally intended for laptops. Yours would be much better optimized though, with a bigger case (cooler), no optical drive, and a PCI sound card or USB sound device. CPU performance capabilities would be worse, but probably not by very much. RAM speeds and front side bus (FSB) speeds also would be slightly worse, but would this matter for these purposes? I doubt it.

 

Re: Setting for permanent underclocking, posted on April 10, 2007 at 00:35:13
swarmer


 
By the way, have you thought about moving the power supply unit outside the case to reduce interference? i.e. an external power brick. I don't know how this would work out... just a thought. This would reduce heat inside the case too, so the fans might not have to spin as much or as fast. It may even be possible to go fanless if heatsinks are used. Replacing the hard drive with flash might help with that too.

 

Further reduction in Video hardware acceleration, posted on April 10, 2007 at 05:19:14
cics
Audiophile

Posts: 1320
Joined: November 9, 2006
 Learning to permanently underclock CPU (at constant 1.13v core voltage), allows for an optimization I originally wanted but wasn't totally successful with. Permanent underclocking (or undervoltage) has made this optimization possible.

Currently, in 4.3.c, video acceleration is one notch before 'None'.

Change this to 'None' (leftmost setting) and leave 'Enable write combining' unchecked. In a well tuned system there should be improvement. You should notice a 'slow down' on video performance (e.g. doing rapid scrolling) as mobo video chipset is now disabled. Basic video functions are done by CPU so avoid busy visuals during Foobar playback, e.g. disable spectrum analyzer.

Technical Background: Previous attempts resulted in higher v-core (1.16v-1.28v) with more frequent voltage transitions. During playback, v-core was weighted towards 1.20v (vs 0.15v). So whilst benefits of disabling mobo video chipset is good (eliminates its emi/rf radiation and overall power consumption drops), CPU gained 10-20% more load that resulted in greater power consumption (as v-core was higher). Basically, no sound improvement.

With permanent underclocking, CPU power consumption remains unchanged (as v-core stays at 1.13v) with extra workload. Hence, disabling video chipset becomes beneficial. You can see this using supplied mobo monitoring utility (ITE Smart Guardian) by changing power scheme back to 'Home/Office Desk'. Remember to set power scheme to 'Portable/Laptop' when finished.

 

Re: Setting for permanent underclocking, posted on April 10, 2007 at 05:41:49
cics
Audiophile

Posts: 1320
Joined: November 9, 2006
 Thats why I like Zalman TNN-300 case. See link where I commented on this.

It has a good fanless PSU installed on side of opening door. Heatsinks are used for cooling. Inside door is largely solid metal and should help with interference.

 

Re: I completely ignored the case and fans and couldn't be happier, posted on April 10, 2007 at 07:22:20
artcheng@netvigator.com
Audiophile

Posts: 388
Joined: November 14, 2003
 Yes, the music server can be placed far and away. But I think these guys is talking about the PC that function as a tranport which house the sound card.

 

Re: Don’t play music from USB drives and Network, posted on April 10, 2007 at 22:04:19
swarmer


 
You can get a SATA to eSATA adapter for 10 bucks or so (search google), and use an external HD over eSATA.

 

Computer case/chassis recommendation, posted on April 12, 2007 at 05:48:09
cics
Audiophile

Posts: 1320
Joined: November 9, 2006
 Paper is silent on what case or chassis to use. Clearly, normal computer cases are not ideal (poor drainage & damping of vibrations, dependence on multiple fans, PSU EMI radiates within case, etc.).

Good news: Home Theatre computers (HTPCs) are rapidly growing. Manufacturers like Thermaltake and Zalman offer HTPC cases that address these issues. Zalman’s TNN-300 certainly looked good on paper (complete fanless design using heat pipe technology, rubber mounts for drives, fanless PSU mounted externally, solid aluminum chassis). It’s also a 2nd generation design. So I bought one – if you don’t try you’ll never know.

Zalman TNN-300 cost me $700 (a net increase of $350 after adjusting for Enermax PSU and a standard computer chassis). I had it burn-in at high power before using it. Installing everything is a full day’s effort.

Does it make a difference? Well if results were unclear, I would’ve said: been their, done that, got the t-shirt and gone back to my trusted old mini-tower with non metallic heavy object on top … Well it’s a different story! Is it better? YES, and its fanless PSU betters the excellent Enermax Noisetaker II.

Firstly, this chassis is well designed, well built, compact and looks nothing like a PC. Its high-tech look fits the part of a supremely talented computer transport. It has a totally different way of installing. Take time to study this. Installation notes:

1. All components install perfectly with the exception of Intel’s socket 775 fanless CPU cooler – manual is unclear. It seems Zalman purposely did this as many Intel 775 CPUs have high temperatures beyond specifications of this chassis. This is not the case with Core 2 Duo CPUs. (Solution: install 775 brackets, then Copper block with smooth end fits flush on CPU layered with thermal grease. Thereafter fit heat pipes etc.).

2. Do not install USB hub and Zalman’s supplied Multifunctional Multimedia Center. Zalman’s remote control and built-in IR receiver is not needed.

3. Clumsy rear-mount thermal blocks (under mobo) is unnecessary as Biostar mobo runs cooler (even more so when BIOS optimizations are done).

4. Downside is for soundcards with IO modules fitting into 5.25” bay. Front panel has a silly lip that covers part of bay – cut holes to allow Soundcard’s IO module to fit perfectly. (Remove front panel before doing this).

5. On first power up, Biostar mobo picks up different CPU voltage and for safety reasons, resets BIOS to defaults. Bypass this problem by resetting BIOS as suggested and also set CPU voltage to ‘Startup’ under ‘Performance Booster Zone’ – although this is the default, setting it again does away with mobo warning on power up.

6. On RAID 0, make sure BIOS SATA mode is set to RAID, otherwise system will not boot as no valid boot area is found.

7. BIOS boot priority may need to be changed as well. Once BIOS is corrected, system starts up. It’s a good idea to check all optimizations.

This computer transport has a great deal of weight on its side which is a good thing as added inertia prevents mechanical vibrations from impacting sound.

What a pleasure. No fans whatsoever (on PSU and CPU), no vibrations and its almost totally noiseless (laptop SATA drives can just be heard). The Biostar mobo complements this chassis beautifully. Zalman’s stand does a good job of dampening vibrations (but spikes would’ve been better). After an hour+ of standalone play I was surprised to see CPU temperature grow slowly and eventually maintain same levels as before (~42 Celcius). It’s been playing continuously for 14+ hours without any problems.

Installing the Zalman TNN-300 with suggested 2 additional optimizations (‘permanent underclocking’ and ‘no video hardware acceleration’) together with RAID 0 yields tangible improvements in sound. Every CD is laid bare in front of your ears with the finest and subtlest information presented oh so effortlessly. It’s very musical. Once you start play you’ll struggle to stop – it’s that good. Average quality CD recordings sounded good & enjoyable but now they’re in the zone (and that’s playing softly). Wow!

This Computer Transport will set you back ~$1,850 - a bargain when you consider more improvements are on the horizon (e.g. non-mechanical low power laptop SATA drives). If you’re shopping for a high-end transport or turntable at any cost, first look here – you won’t be disappointed.

 

Point of note: Steel cases will generally be better for electromagnetic disturbance shielding than aluminum cases., posted on April 12, 2007 at 11:42:37
Ugly
Audiophile

Posts: 2942
Location: Des Moines, WA
Joined: August 22, 2006
 This is due to the high iron content. Iron is an excellent conductor of magnetic fields. Aluminum is a poor conductor of magnetic fields. So while Al provides great E field blocking it will do nothing for H fields. IMO IBM went with steel for more reasons than cost.

 

Not sure if steel over aluminium would make a sonic difference, posted on April 12, 2007 at 14:09:23
cics
Audiophile

Posts: 1320
Joined: November 9, 2006
 It's more to do with no fans and better PSU that is located externally preventing internal mayhem.

BTW - did you get your answer on DS, KS or ASIO (see section 4.7.g)?

 

Re: Not sure if steel over aluminium would make a sonic difference, posted on April 12, 2007 at 16:54:05
Ugly
Audiophile

Posts: 2942
Location: Des Moines, WA
Joined: August 22, 2006
 "BTW - did you get your answer on DS, KS or ASIO (see section 4.7.g)?"

Thanks for pointing that out. While I did read through it it was a few days ago way back when I was still a Foobar newb ;-]

4.7.g nicely explains the KS ASIO tradeoffs but doesn't mention direct sound. Any opinions about that?

 

Re: Not sure if steel over aluminium would make a sonic difference, posted on April 13, 2007 at 02:59:57
cics
Audiophile

Posts: 1320
Joined: November 9, 2006
 Stay away from DS - Windows will mostly likely interfere. If your soundcard forces this, I would suggest getting another.

 

More XP tweaks:I'd like you to hear your thoughts., posted on April 13, 2007 at 12:10:46
texastea006@yahoo.com
Audiophile

Posts: 58
Location: gulf coast
Joined: March 22, 2006
 Hello all. I have some additional XP settings that I thought I’d offer up and see if anyone agrees on an improvement in sound. I applied all the common XP tweaks a year ago, but these I don’t think I’ve read about so I figured I’d post them. Nothing scientific, just using my ears.Try them one at a time if possible:

Device Manager:System Devices:
Disable system speaker. Disable Windows UAA High Defenition Audio (if you have it). If it asks you to reboot, you can say no and still hear the difference. I disable other devices not in use. i.e. network adapter, DVD DRIVE, monitors..Anyone know of any others under System Devices that are safe to disable?
(Side question: What does disabling items in device manager actually do? Just stop data communications?)
next…but note that these steps may be annoying to try to use windows this way if you actually do more than listen to music on this pc. I’m going to see if I can adapt to navigating with the file explorer..
My Computer-Tools-Folder Options:
View tab – I have checks only on “Show hidden files/folder”, “launch folder windows in separate process”. “Show both parts but manage as a single file”. Everything else is unchecked.
At the top set all folders to “Like current Folder”.
General tab. Use classic desktop and folders. Open each folder in its own window. Single click to open/underline icon consistent with browser.
Again on the view tab set all folders to “Like current Folder”.

I can hear improvements with all these settings. Anybody disagree with any of them? Any other uncommon settings I might not know about?

Also, is it safe to remove and replace power connection from the dvd drive while computer is on? I think people say that it helps.
Mike.

I use foobar asio out with rme digi soundcard .

 

Disable system devices, posted on April 15, 2007 at 09:52:49
cics
Audiophile

Posts: 1320
Joined: November 9, 2006
 Thanks Mike,

Just for clarity, this is the change I did:

Control Panel > System > Hardware TAB > Device Manager > expand System Devices > right-click 'System speaker' > Disable

You're right there's an improvement (one less item on PCI bus).

Also try disabling 'Terminal Server Device Redirector' (which is further down the list) this will force a reboot. Its a low-level change and in my system I get improvements. Redirector seems to add overheads.

Collectively, there's more air. Brilliant!

Don't have Windows UAA stuff. On folders/explorer, I don't use it but will try it (you never know with Windows).

 

Safe to remove & replace power from ROM drive while computer is on? Definite NO., posted on April 15, 2007 at 09:57:45
cics
Audiophile

Posts: 1320
Joined: November 9, 2006
 Yes - you must remove this drive for best sound (not just the power connections, also IDE bus cable). Keep away from IDE interface.

 

What does disabling items in device manager actually do?, posted on April 15, 2007 at 10:09:35
cics
Audiophile

Posts: 1320
Joined: November 9, 2006
 My take: Its more than data comms. You free up important hardware resources & bandwidth (on PCI bus). If device is linked to a specific chipset, you get less power consumed. Associated Windows processing also stops meaning less overheads on CPU.

 

"Al provides great E field blocking it will do nothing for H fields", posted on April 15, 2007 at 13:05:38
cics
Audiophile

Posts: 1320
Joined: November 9, 2006
 Interesting that inside metal plate of Zalman PSU is not aluminium. This closes onto computer internals and PSU sits outside.

Plate is most likely steel. Preventing stray magnetic fields into case is a big help as well.

 

Re: Disable system devices, posted on April 16, 2007 at 18:35:48
texastea006@yahoo.com
Audiophile

Posts: 58
Location: gulf coast
Joined: March 22, 2006
 Cics, I just disabled integrated audio in my bios and that seemed to improve the sound further. I'm going to investigate some of the other bios options. I was looking for the spread spectrum option, but I don't see it on my dell.

I forgot to mention disabled terminal services, as one that I had already disabled, and that improvement was noticable as well. Also,I use an IDE dvd player to put movies on my hard drive quiteoften, so I'd like to keep my ide dvd drive plugged in for now.

Also, in the folder options, I've changed my mind and prefer the "show both parts and manage them individually" option what I said in my previous post.

 

Re: Disable system devices, posted on April 16, 2007 at 18:55:41
texastea006@yahoo.com
Audiophile

Posts: 58
Location: gulf coast
Joined: March 22, 2006
 I did like to see file details in my windows, but it seems that changing the view to List also helps a lot. All you do is right-click in any directory window and going under the View menu, change the view to List.
Them , as always, you need to go into the folder options--> View and hit the "apply to all folders": button to get the toal benefit.

 

Dell (spread spectrum) & Folder Options, posted on April 17, 2007 at 10:31:29
cics
Audiophile

Posts: 1320
Joined: November 9, 2006
 Our home PC is a Dell and it also doesn't offer option to disable spread spectrum. Thats why I went with Biostar mobo as you want this disabled. Also, there are other PCI devices/features (modem, LAN, EHCI & emulation) in BIOS that you should disable - it all helps free up PCI bus traffic.

On Folders/Explorer, I found very interesting stuff in Windows that you want to avoid/disable - I'll post the solution (which is a bit radical) once my test is 100%.

Yes, disabling 'Terminal Server Device Redirector' gives a good improvement. Likewise, the new stuff on Folders/Explorer gives yet more improvements.

 

Changes to Version 0.1, posted on April 18, 2007 at 10:20:23
cics
Audiophile

Posts: 1320
Joined: November 9, 2006
 Paper is at version 0.1.

Additional useful suggestions and insights have been gained from Audio Asylum. I wanted to summarize these by separate reply posts under this post (that way no need to review posts elsewhere). Note, AA links are given where optimization was already explained.

These additional enhancements (changes, more information and new optimizations) yield good improvements. At this refined level of optimization, I don't need shielding on my soundcard (i.e. removed Quantum Physics Noise Disruptor).

 

1. [Change] Permanent Underclocking, posted on April 18, 2007 at 10:22:01
cics
Audiophile

Posts: 1320
Joined: November 9, 2006
 .

 

2. [Change] No Video hardware acceleration, posted on April 18, 2007 at 10:23:44
cics
Audiophile

Posts: 1320
Joined: November 9, 2006
 .

 

3. [For Info] LCD / Plasma screen setting, posted on April 18, 2007 at 10:25:09
cics
Audiophile

Posts: 1320
Joined: November 9, 2006
 .

 

4. [New] Disable System Devices, posted on April 18, 2007 at 10:26:42
cics
Audiophile

Posts: 1320
Joined: November 9, 2006
 .

 

5. [New] Change to System Cache mode, posted on April 18, 2007 at 10:28:42
cics
Audiophile

Posts: 1320
Joined: November 9, 2006
 Control Panel > System > Advanced TAB > Settings BUTTON under Performance > Advanced TAB > Select System Cache under 'Memory usage'

Above helps with storing entire .wav files into RAM and reduces need to have Foobar perform full file buffering (which doesn't work for .cue files).

 

6. [New] Bypass Windows Explorer (User Interface), posted on April 18, 2007 at 10:39:36
cics
Audiophile

Posts: 1320
Joined: November 9, 2006
 Thanks to texastea006 (Mike) for sharing experience on Windows Explorer settings. If you change settings in Folder options or Task bar properties, this affects sound quality. Windows Explorer (desktop user interface) must not be confused with Internet Explorer as these are separate things. Windows Explorer is a nasty piece of work which consumes memory (10MB-60MB+) and randomly performs tasks. It interferes with running tasks / progams and files being accessed. Hence it affects sound quality.

Solution is to bypass Explorer through stopping Explorer and only running Foobar. After exiting Foobar, Explorer is restarted. Steps:

1. Separate Foobar from Explorer. Foobar > Preferences > Display > Columns UI > Status bar/Systray TAB > Uncheck all items in 'System notification area'

2. Detune Windows Explorer UI.
Control Panel > Folder Options > View TAB > Uncheck as many items (leave 'Use simple file sharing' checked).
Right-click on Start > Properties > Start Menu TAB > Customize > Advanced TAB > Uncheck and Disable as many 'Start menu items' you can (key things needed are Control Panel, Run and My Computer).

3. Download 'Process' utility from link below (freeware). Zip file contains a small (25k) program called process.exe which you must extract to c:\ (your root directory). This program allows access to windows tasks where you can stop, suspend… processes as seen in Task Manager's Processes TAB.

4. Create a ?.bat file and save to your Desktop using Notepad. I call mine 'cicsStart.bat'. Copy following commands into this file:

@echo off
rem Kill Windows UI
start /wait c:\process.exe -k explorer.exe

rem Start Foobar2000 in realtime (%1 allows playing .wav files from File Explorer)
start /wait /REALTIME c:\progra~1\foobar2000\foobar2000.exe %1

rem Start Windows UI after exiting Foobar2000
start /wait /B c:\windows\explorer.exe

5. Start Task Manager (under Options, make sure 'Always on top' is checked). Goto Processes TAB and locate explorer.exe. Take note of memory consumed and CPU activity.

6. Run cicsStart.bat (or other name you have chosen) by double-clicking file on your Desktop. Process explorer.exe in Task Manager should disappear and foobar2000.exe & cmd.exe should appear. Also note all memory consumed by explorer.exe is now available (under Performance TAB). You can copy this .bat file in your Startup directory instead of the Foobar shortcut. This way, computer boots up directly with Foobar without any interference from Explorer.

7. Test sound quality. There should be a nice improvement.

Note you could choose to replace default Windows Explorer UI with other 3rd party versions but I prefer not having any UI present when playing music (most optimal option).

 

7. [New] Computer Case/Chassis, posted on April 18, 2007 at 10:41:45
cics
Audiophile

Posts: 1320
Joined: November 9, 2006
 .

 

Re: Disable system devices, posted on April 18, 2007 at 10:48:03
cics
Audiophile

Posts: 1320
Joined: November 9, 2006
 I've posted a solution to get rid of Explorer effects. Please test and let me know.

 

Have you tried changing affinity?, posted on April 18, 2007 at 12:07:27
edward


 
cics - This thread is very coincidental for me, as I've been in the process of building a new computer over the past couple of weeks, and even before I read this thread I was going through my computer making a lot of these same suggested tweaks.

Now, forgive me if you mentioned it already and I just missed it, but I think I have one more to add to the experimentation table. It might fall into the same category as making the Windows Explorer tweaks. I have a Core 2 Duo CPU, so I decided to try something. After I open foobar and begin to play a song, I stop it right away and then go into the task manager (CTRL-ALT-DEL). Then I click on the Processes tab. At this point, since I started foobar already, it shows up in the processes as well as the ASIO (exe) dll I am using. So, I right click on those two processes and change the affinity to have only CPU (0) checked. Then I start changing all the other processes to have the affinity on only CPU (1) checked. Some of the system processes you cannot change, but you can definitely change explorer.exe.

Anyway, as I said this is one for those who like to experiment. I think I can hear a difference, but haven't come to a definite conclusion yet. And, of course, what might make a difference on my system may not make one on someone else's.

So what do you think? Is this valid or just whack?

 

8. [New] Remove Foobar component: foo_ui_std.dll, posted on April 18, 2007 at 23:18:05
cics
Audiophile

Posts: 1320
Joined: November 9, 2006
 With Columns UI, above default UI (foo_ui_std.dll) is not needed. Delete this file from C:\Program Files\foobar2000\components directory. When you start Foobar, answer 'No' to save settings for this component.

This way, there is no chance of any Foobar / Windows Explorer interaction.

 

Re: Have you tried changing affinity?, posted on April 19, 2007 at 05:58:42
cics
Audiophile

Posts: 1320
Joined: November 9, 2006
 Hello Edward,

Nothing whacky when it comes to Windows (unrelated stuff can affect sound). I’ve tried using affinity setting in Task Manager but can’t say there’s an improvement. You can automate this (without going through Task Manager) using process utility (link below) as used for bypassing Windows Explorer. This way, you can set affinities in .bat file (e.g. start /wait c:\process.exe –a foobar2000.exe 01). Beware of increasing CPU’s core voltage (more RF radiation & power consumption) with this – I’ve seen this increase to 1.28V (from 1.13V low).

On your setup, add Zalman TNN-300 case (~$700), Biostar mobo ($55) and do SATA RAID 0 with 2 laptop SATA drives. You’ll save on case and PSU (~$350). On version 0.1 setup and these additional changes, my system (computer transport > AA Prestige SE > Levinson 33H > JMLab Nova Utopia Be) is getting closer to benchmark reference system (based on turntable front-end with nothing left to chance – this setup costs ~$400k)!

 

Re: Have you tried changing affinity?, posted on April 19, 2007 at 09:45:26
texastea006@yahoo.com
Audiophile

Posts: 58
Location: gulf coast
Joined: March 22, 2006
 Here is what I have been using for the past year in my dell dual core:
Sysinternals.com Process Explorer (find it with google..free dl)
It allows you to change priority and affinity for all processes, including those the windows task manager will not let you modify...winlogon.exe smss.exe, services.exe, etc...

It takes a long time, so adding these to any batch fiile (which I don't have yet) would be a big time saver. I select everything other than foobar and change the priority to IDLE and gave affinity for only one processor. I'm not sure yet if one sounds better over the other, but I was using 1. Then, I set foobar to priority=realtime and use both processors.
Now that we have all these improvements to make by avoiding the windows explorer, it may be time for me to revisit which settings sound best..but that's what I was doing in the past.
I am looing forward to removing windows explorer from the chain to see how that sounds.

 

Quantum Physics Noise Disruptor, posted on April 19, 2007 at 12:12:21
texastea006@yahoo.com
Audiophile

Posts: 58
Location: gulf coast
Joined: March 22, 2006
 Off subject, did you compare the qp noise disruptor and shakti stones at all? It wasn't until I installed four stones in my audio rig before I was able to really discern which explorer changes were improvements. I have one stone SNUGLY between my RME sound card and pci-e video card. I also have audio magic power conditioners without noise disrupters in them, although the manufacturer is selling an upgrade to install them in his conditioners. I haven't done that, but am using two stones around my conditioners. He says the qp devices are better than the stones, so have you compared them and can you quantify the difference?
Any experience with "Quantum symphony" brand devices either?
Mike

 

Re: 6. [New] Bypass Windows Explorer (User Interface), posted on April 20, 2007 at 15:24:35
texastea006@yahoo.com
Audiophile

Posts: 58
Location: gulf coast
Joined: March 22, 2006
 Yes, that does sound noticably cleaner.
Now I have one more secret I've been saving to share, but I don't know if it can be safely used. See the following link:
http://blogs.technet.com/markrussinovich/archive/2005/07/24/running-windows-with-no-services.aspx
I saved this after reading it long ago, hoping I could use it in the future, but someone else has to figure out how it can be done safely because I'm new to batch files:

Basicaly, you can kill Smss.exe first and then winlogon.exe (which includes all your services) using that Process Explorer program (or batch file, I assume).
I used to do this frequently, and heard a nice improvement, but the catch is that you are forced to turn off your computer with the power button since windows can't function properly anymore.
If there was a way to use a batch file to restart those processes or services in the way you just showed us how to restart explorer, it would be real cool. That way when running foobar you could temporarily disable almost everything else that uses memory.
Trying this should not damage your pc, but some are not comfortable with doing a hard reboot, so proceed at your own risk. It's fun to at least and see how deep the rabbit hole goes.
Let me know if this is feasible.

 

It gets even better ..., posted on April 21, 2007 at 09:58:53
cics
Audiophile

Posts: 1320
Joined: November 9, 2006
 Very interesting – more air & blacker backgrounds!

My Audigy 2 ZS Platinum (with kX driver) yields more improvements than my RME HDSP 9652. RME still edges out Audigy by a small margin. RME card has very latest software (firmware and driver @ version 3.02 March 2007). This RME software update gives better playback results (latency via ASIO/RME down to 0.32ms vs kX @ 0.25ms using Kernal Streaming/Audigy).

I’m posting 2 solutions (reply post to this) on how to automate this. Summary:

1. Can’t terminate all services as you get no playback from Foobar. Some Windows services are critical
2. Option 2 offers more aggressive culling of Windows
3. Must shutdown computer by pressing power button 4-5 seconds – force shutdown
4. Both options attempt to recover Windows UI after exiting Foobar – depending on soundcard, option 2 may not give Windows UI
5. Preference is to select option that gives Windows UI after exiting Foobar – in my case, only option 1 works for both cards
6. Not much sound difference between options even with option 2 being more aggressive
7. Don’t run .bat file at startup – rather double-click from Desktop until right option is chosen
8. Any changes made on desktop is not saved after exiting Foobar (as proper Shutdown doesn’t happen). On startup and before running .bat file, do necessary changes and reboot to save Desktop UI changes
9. Both options start Windows Task Manager (at lower priority). Very useful to see Windows processes being eliminated & memory freed. For information only – exit Task Manager at any time.

 

Option 1: Kill Windows – Medium, posted on April 21, 2007 at 10:03:21
cics
Audiophile

Posts: 1320
Joined: November 9, 2006
 Only need Process Utility (process.exe) as per change item 6. Instructions copied into .bat file has ‘[YOUR SOUNDCARD PROCESS]’ – put your soundcard’s process here (e.g. kX = kxmixer.exe and RME = hdsp32.exe). Also remove ‘rem ’ prefix. Next instruction kills RME’s mixer process (if ‘rem ’ prefix removed).

1. Edit .bat file and clear all contents

2. Copy following instructions into .bat file

@echo off
rem # Kill Windows UI
start /wait c:\process.exe -k explorer.exe

rem # Start Task Manager
start C:\WINDOWS\system32\taskmgr.exe

rem # Kill Windows & VTTimer.exe (for Biostar mobo using VIA chipset)
start /wait c:\process.exe -k VTTimer.exe
start /wait c:\process.exe -k smss.exe
start /wait c:\process.exe -k winlogon.exe

rem # Adjust priorities, eg. High, RealTime, BelowNormal, Low
start /wait c:\process.exe -p taskmgr.exe Normal
rem start /wait c:\process.exe -p [YOUR SOUNDCARD PROCESS] AboveNormal
rem start /wait c:\process.exe -k hdspmix.exe

rem # Start Foobar
start /wait /REALTIME c:\progra~1\foobar2000\foobar2000.exe

rem # Start UI after Foobar
start /wait /B c:\windows\explorer.exe


3. Test by double-clicking .bat file

 

Option 2: Kill Windows – Aggressive, posted on April 21, 2007 at 10:06:35
cics
Audiophile

Posts: 1320
Joined: November 9, 2006
 Another process utility (pskill.exe) is needed to kill process and its dependencies. Instructions copied into .bat file has ‘[YOUR SOUNDCARD PROCESS]’ – put your soundcard’s process here (e.g. kX = kxmixer.exe and RME = hdsp32.exe). Also remove ‘rem ’ prefix. Next instruction kills RME’s mixer process (if ‘rem ’ prefix removed).

1. Download PSTools.zip from link below

2. Open zip file and copy pskill.exe (183KB) to c:\ (root directory)

3. Run pskill.exe (double-click on it) to do away with license agreement confirmation

4. Edit .bat file and clear all contents

5. Copy following instructions into .bat file

@echo off
rem # Kill Windows UI
start /wait c:\process.exe -k explorer.exe

rem # Start Task Manager
start C:\WINDOWS\system32\taskmgr.exe

rem # Kill Windows & VTTimer.exe (for Biostar mobo using VIA chipset)
start /wait c:\process.exe -k VTTimer.exe
start /wait c:\process.exe -k smss.exe
start /wait c:\process.exe -k winlogon.exe
start /wait c:\pskill.exe -t lsass.exe
start /wait c:\pskill.exe -t svchost.exe

rem # Adjust priorities, eg. High, RealTime, BelowNormal, Low
start /wait c:\process.exe -p taskmgr.exe Normal
rem start /wait c:\process.exe -p [YOUR SOUNDCARD PROCESS] AboveNormal
rem start /wait c:\process.exe -k hdspmix.exe

rem # Start Foobar
start /wait /REALTIME c:\progra~1\foobar2000\foobar2000.exe

rem # Start UI after Foobar
start /wait /B c:\windows\explorer.exe

6. Test by double-clicking .bat file

 

Option 2: Kill Windows – Aggressive (Signature Edition), posted on April 22, 2007 at 02:24:52
cics
Audiophile

Posts: 1320
Joined: November 9, 2006
 I wanted more of Windows culled without loosing Windows Explorer UI (after exiting Foobar) and came up with this solution. Explorer is suspended (that way all its inference stops) and resumed after Foobar is exited. This will work with all soundcard drivers. By suspending Explorer, no CPU resources are consumed but unfortunately, memory remains allocated to Explorer.

I get best results using this option.

Please don’t minimize Foobar – you MUST exit to get back to Explorer. Once application is minimized, it cannot be restored (as explorer is suspended).

Under step 5 (of option 2), use these instructions:

@echo off
rem # Suspend Windows UI
start /b /wait c:\process.exe -s explorer.exe

rem # Start Task Manager
start /b C:\WINDOWS\system32\taskmgr.exe

rem # Kill Windows & VTTimer.exe (for Biostar mobo using VIA chipset)
start /b /wait c:\process.exe -k VTTimer.exe
start /b /wait c:\process.exe -k smss.exe
start /b /wait c:\process.exe -k winlogon.exe
start /b /wait c:\pskill.exe -t lsass.exe
start /b /wait c:\pskill.exe -t svchost.exe

rem # Adjust priorities, eg. High, RealTime, BelowNormal, Low
start /b /wait c:\process.exe -p taskmgr.exe Normal
rem start /b /wait c:\process.exe -p [YOUR SOUNDCARD PROCESS] AboveNormal
rem start /b /wait c:\process.exe -k hdspmix.exe

rem # Start Foobar
start /wait /REALTIME c:\progra~1\foobar2000\foobar2000.exe

rem # Resume Windows UI
start /b /wait c:\process.exe -r explorer.exe

 

Page: [ 1 ] [ 2 ] [ 3 ]

Page processed in 0.073 seconds.

[ Top Top of Page ] [ Contact Us ] [ Support/Wish List ] [ Copyright © 2025, Audio Asylum, all rights reserved ]
[ General ] [ Speakers ] [ Tubes ] [ Vinyl ] [ Digital ] [ Hi-Rez ] [ Video Asylum ] [ Cables ] [ Tweaks/DIY ] [ Music ] [ Films ]