Best cooling solution for studio computer?

Xspringe · Post by **Xspringe** » Mon Feb 16, 2004 1:11 pm

Hello everyone! Nice little forum you got here

I'll be buying a computer in about 3-4 weeks which will be used in a studio enviroment. As a result, noise levels need to be as low as possible, preferrably inaudible. What would your recommendations be? I've been looking at the total fanless designs by Zalman and Silentmaxx but both are rather expensive. Are there any fanless water cooling products?

I will probably be using the Athlon 64+ 3000+ (I'm going to wait for the new, cooler stepping) along with the samsung 7200 rpm sata drive (reasonably fast, quiet and cold).

The only thing I've figured out so far is that I won't be buying any expensive HD enclosures, but instead am going to use a simple suspended HD construction. I thank Silentpcreview for saving me some euros with this

Curtis · Post by **Curtis** » Tue Feb 17, 2004 3:21 am

If fanless cases are too much, I'd probably recommend steering clear of watercooling solutions. Not becuase of the noise (I've never heard water cooling in action), but because of the potential electrical interference introduced into the studio environment by having a nice big pump motor in the room. You never know what kind of problems that might introduce into your audio setup.

Go with a low speed panaflo(s)/Zalman 7000A rig - at least that's my recommendation.

halcyon · Post by **halcyon** » Wed Feb 18, 2004 12:52 am

If you go water cooling route, I agree with Curtis that you need to electrically isolate the pump motor from your audio equipment electricity chain.

Otherwise you are in for a quite a lot of trouble (line noise, ground hum, etc).

Still, many studio systems use a safety transformer + separate ground path anyway to isolate the electricity problems, if they can't afford separate electricity lines for the audio equipment. My friends use this approach succesfully.

For silent operation I would recommend: a macintosh in a silenced computer "booth" or "enclosuse". But then again I believe macs are superior for studio work (yes, I'm biased).

regards,
Halcyon

Fury161 · Post by **Fury161** » Wed Feb 18, 2004 10:54 am

you could just stick the pc outside the studio

worked for me

Xspringe · Post by **Xspringe** » Wed Feb 18, 2004 11:47 am

Fury161 wrote:you could just stick the pc outside the studio

worked for me

Yeah I'm putting it inside a hatch behind a wall (triple wall construction, inner wall is merely for acoustical purposes and not isolation).

How did you take care of the cables?

silvervarg · Post by **silvervarg** » Tue Feb 24, 2004 1:31 am

If you want a really silent computer without paying too much you should start with evaluating how much computer power you really need.
Basicly the less power you need, the less heat the computer will dissipate.
If you end up with real high demands you might want to go with water cooling or lots of DIY projects. If you don't need tremendous power you will most likely be best of with some standard techniques for air cooling.

Try to get ridd of everything that you don't need in the computer to lower the generated heat. E.g. integrated graphics or very low end graphics card is preferable over high end cards.
Ofcourse you should not skip out on the audio-card....Better to spend a few extra watts on this one.

As for harddiscs you have already noticed that the Samsung P80 discs are a good choise if you need silence and rather good performance. Loads of cache, lots of RAM etc is also very good to limit the work needed by the harddrive.

Take extra care in looking at what monitor you choose and what input devices you use. The noise will be very close to the microphone, so it is important to pick really silent components here. There is lots of good threads on this forum about this. Do a few searches.

Try to specify what you acceptable noise level is. Is it ok if the computer makes 15dB(A) noise at operating position?
Does it need to be only 10dB(A)?
You will need to pay with either money or effort (or a mix) for every dB you want to go lower.

WayneSim · Post by **WayneSim** » Tue Feb 24, 2004 2:45 am

"inaudible" PC has been my goal for at least 2 years!!! I've tried many things. And let me tell you it is extremely hard!!!

IMO, Only 2 ways to make a truely silent PC. 1) Passive Water cooled rig (HEAPS of work!!!). 2) Silent Case i.e Zalman 500a (HEAPS of money!!!)

It's either a lot of work or you have to fork out the money. Have you seen "SilverSilence" My lastest watercooled rig?

I went down the watercooling path becasue it's the only way to get true silence. However the down side is there is some maintenance.

Having said all that my DAW is not watercooled atm. However all acoustic instuments are recorded in another room. But that doesn't help me... Let me explain what I've doen so far to help with noise. Currently my DAW case is covered in foam from the inside. Enermax PSU has been fitted with an 80mm and a 92mm Papst fans. Both run at 5 volts. Case fan is a Papst 80mm on 5 volts. Video card has been fitted with a Zalman heatpipe heatsink (no fan). Thermalright SP94 CPU heatsink will be the next thing added, and placed overhead will be Papst 80mm, most likely on 5 or 7 volts. Case will be tucked away under my desk. That setup will basically be very quiet. Far from "inaudible" but quiet. The only next step from that is watercooling. And that is a pain in the ass!

You, like me will learn "inaudible" is a pipe dream. I'm very sorry to tell you this but it's the truth. "Silversilence" is EXTREMELY close to "inaudible". However to do the same setup with a DAW will take a LOT of effort. It is quite possible, but heaps of effort! And NEVER trust a manufacturer who uses the word "silent" in the title of any product. I've been burned many times! And yes I read lots of reviews. Know that every product that comes with a fan. The fan will need to be replaced!

And dont skimp on the PSU! Get a well known brand!

That is my exprience.

Post by **MikeC** » Tue Feb 24, 2004 2:59 am

WayneSim -- Seems to me one of the reasons you've chosen WC is because of the hot weather that prevails in Aus. summers. Xspringe, you should tell us what your typical temp range is in your room. It will make a big difference. My PCs see 20~22C in winter and 23~29C in summer. Rarely past 30C. All are air-cooled, and one measures 16~17 dBA/1m. This one is ~20dBA/1m: MikeC's external PSU PC

silvervarg · Post by **silvervarg** » Tue Feb 24, 2004 6:29 am

I recognice most of the problems that WayneSim is experiencing, that is why one of my most important points was to point out what noise level is tolerable. This is close to defining what will be "inaudiable" too you.
If this level is set to 10 or 15 dBA we have something to work with.

The computer might not be inaudiable if you stick your ear to it. In a studio it might just needs to be inaudiable to the microphone.
As MikeC pointed out he got a computer 16-17dBA/1m. In this case we have the option to put the computer behind a noise isolating wall, and perhaps 1.5m away. Since the computer is out of sight we have the option to put sound absorbing materials on the outside of the computer box.

I think there is a good chance that we can give some help in how to make this computer "inaudiable" with just a few commonly used tricks and a little work.

sbabb · Post by **sbabb** » Tue Feb 24, 2004 10:01 am

You could always watercool and locate the noise generating components remotely. I've seen W/C setups with the pumps and radiators in different rooms than the PC's in. The only one that I could find quickly was BladeRunner's Zero Fan Zone web site where he goes to the extreme of using a buried water tank as a radiator. I've also seen one where the pump is in the garage and the radiator is a lot of copper pipe laid on the garage floor.

I'm still in the process of building my first W/C machine. I plan on bypassing the pump power leads to ground with capacitors to kill any electrical noise on the power leads. I'm using the Swiftech MCP-600 pump, which is powered off a standard 4-pin 12V DC line from the PSU, so I don't have any AC noise to worry about, just any electrical noise that might be generated by the pump.

It's an Athlon XP2500 (unlocked and already running at 3200 speeds with just the stock air cooling) that will be used for various multimedia projects including audio and video editing.

Scott

PretzelB · Post by **PretzelB** » Tue Feb 24, 2004 11:20 am

Xspringe wrote:
Fury161 wrote:you could just stick the pc outside the studio

worked for me
Yeah I'm putting it inside a hatch behind a wall (triple wall construction, inner wall is merely for acoustical purposes and not isolation).

How did you take care of the cables?

I think most of the cables for a pc can be extended with no problem. I seem to remember that usb has a limit to how far it can go but you can get hubs to extend that. Maybe someone can comment on the max length for video, keyboard, and mouse.

If you desk is close to the hatch/wall, you could easily put the pc in a totally separate room. Depends on the room layout.

IsaacKuo · Post by **IsaacKuo** » Tue Feb 24, 2004 12:30 pm

Speaking of water cooling, I wonder if people have ever tried simple evaporative cooling instead of a radiator. Instead of a completely closed system, the water resevoir is a big open top bucket placed at a higher altitude than the computer (i.e. the water level is above the computer).

WayneSim · Post by **WayneSim** » Tue Feb 24, 2004 6:55 pm

MikeC wrote:WayneSim -- Seems to me one of the reasons you've chosen WC is because of the hot weather that prevails in Aus. summers.[/url]

I guess that is true to some degree. However I must state that to "my" ear's, 95% of fans on 5 volts annoy me. Even at 1m! Yes, i will admit i'm a freak. However I do live in a quite area. And my ears are very sensitive.

I commend companies like Zalman that make pc products without fans. I'm personally waiting for the Zalman case to get released in Australia. So I can way up the cost and performace. With the possibity of buying one for my DAW. WHOLE SETUP WITHOUT FANS!!! And without the fuss and mantainace of a watercooled rig (perfect for the studio)! Although it will most likely cost double that of a high end passive watercooled rig. I'll just have to wait and see.

Charlie Escher · Post by **Charlie Escher** » Tue Feb 24, 2004 10:27 pm

I think most of the cables for a pc can be extended with no problem. I seem to remember that usb has a limit to how far it can go but you can get hubs to extend that. Maybe someone can comment on the max length for video, keyboard, and mouse.

If you desk is close to the hatch/wall, you could easily put the pc in a totally separate room. Depends on the room layout.

I know folks who use 30 or 50 foot KVM cable sets and locate in another room, but allow me to suggest a potentially more satisfying solution. Locate your fast, powerful machine remotely. Patch it balanced audio via standard cables for recording and playback. Get a silent, very low powered machine to use in the studio. Others here can comment on how to do that, I'm actually making my first post here.

Install Windows XP Pro on both machines. Network them by ethernet (plenty of length available), Firewire networking (only good to 15 feet without a repeater, but an external drive can be the repeater, I believe), or wi-fi. Run the fast computer remotely, using Remote Desktop in XP. I've run very noisy video processing machines at work in this manner, with the vid machine in another office, with the door closed. I still use only an old Athlon 750 at work, and remotely run 2-3 gig computers with big RAID arrays from the confines of my relatively quiet office. Remote Desktop is a pretty cool feature of XP, and seems possibly well suited to the original poster's needs. PC Anywhere and similar software will do the job too, of course.

50 foot KVM sets are less than $100, BTW. If you are going to use a dual monitor setup, which is nice for audio editing apps, you'll need two long video cables if you go the single computer route.

My own little living room studio necessitates moving my computer behind the pickup pattern of the room mikes. Thanks to this site, that may soon change. One of those truly silent, rack mounted computers is looking like a good excuse for gear acquisition syndrome. In any case, I've reduced the noise levels in my audio computer very substantially, with more to come.

chylld · Post by **chylld** » Tue Feb 24, 2004 10:50 pm

IsaacKuo wrote:Speaking of water cooling, I wonder if people have ever tried simple evaporative cooling instead of a radiator. Instead of a completely closed system, the water resevoir is a big open top bucket placed at a higher altitude than the computer (i.e. the water level is above the computer).

problem with that is that all sorts of dust and insects and crap get stuck in the bucked and end up in the watercooling loop. not only will it clog up the system quickfast, but it'll also hasten the process of corrosion since the coolant isn't 'pure' anymore.

Post by **MikeC** » Tue Feb 24, 2004 11:17 pm

There's a lot of loose talk in this thread about 15 and 10 dBA target sound levels. Have any of you who mentioned these numbers actually heard the noise what 10 and 15 dBA really sounds like? I am guessing not. I am fairly certain that the ambient noise level in typical pro studios is NOT below 20 dBA.

My main office, in a sleepy neightborhood many blocks from any main roads, has a background noise of 23-25 dBA with the windows closed. When a car gloes by on the lane in front ~60 ft away, the noise jumps to well over 30 dBA. Someone walking in the house -- almost anywhere in the house, pushes the noise up 4-5 dBA. And no, it is not an old creaky house, it's fairly solid.

I have to wait till past 1~2 AM in my downstairs lab to get ambient noise down to 16~17 dBA. The quietest room in the house, my wife's office in the back downstairs, measured ~14 dBA at 2AM after a snowstorm that dumped over a foot last winter. You know the quieting effect snow has. It was so quiet I heard whining in my ears. Maybe some of you have also experienced that in a super quiet room. It's a common phenomenon.

The point is that under 15 dBA is not at all a realistic target for a Athlon 64+ 3000+ and any 3.5" IDE drive. Only way you'll get tghat is with no moving parts -- certainly no fans -- or maybe just with a single super quiet notebook drive encased in an insulated chamber inside an insulated case. You would do very well to reach 20 dBA.

Charlie Escher · Post by **Charlie Escher** » Wed Feb 25, 2004 12:43 am

There's a lot of loose talk in this thread about 15 and 10 dBA target sound levels. Have any of you who mentioned these numbers actually heard the noise what 10 and 15 dBA really sounds like? I am guessing not. I am fairly certain that the ambient noise level in typical pro studios is NOT below 20 dBA.

You're right, somewhere between 20-30dBA is considered normal for a decent studio. And that's before you put humans in there. However, there are exeptional studios that surpass those numbers handily. In regard to this studio: http://www.galaxystudios.com , one of the designers made this claim in rec.audio.pro :

Even during the day the noise levels are around 0dB(A) (measured!). With ventilation on they still remain below the noise floor of the recording
microphones (level ca 13 dB(A)).
Practically you can't hear a thing, and the noise floors of the recording mics is the limiting factor, not the rooms.

The designer's name is Eric Desart. The Google link is reduced to: http://tinyurl.com/2yrhd . It's a long, contentious thread, but I found parts of it interesting. There are links from Mr. Desart regarding measurements techniques and gear applicable to these very quiet levels. I realize that this information is not new to you, Mike, but perhaps it is useful to those of us still coming to grips with this discipline. I've only read a few hundred articles and forum posts on the SPCR site thus far, apologies if I'm merely repeating old material.

sbabb · Post by **sbabb** » Wed Feb 25, 2004 11:29 am

Ten posts further along in that Google thread is one from a guy named Jay Kadis who says that his main studio noise floor is 21.3 dBA in a mechanically isolated studio and it was "quite an expensive proposition to get this low a noise floor." In his control room it's 23.2 dBA with everything turned off and 33.3 dBA with his equipment turned on.

Scott

Post by **MikeC** » Wed Feb 25, 2004 4:25 pm

sbabb wrote:Ten posts further along in that Google thread is one from a guy named Jay Kadis who says that his main studio noise floor is 21.3 dBA in a mechanically isolated studio and it was "quite an expensive proposition to get this low a noise floor." In his control room it's 23.2 dBA with everything turned off and 33.3 dBA with his equipment turned on.

Scott

That to me sounds more believable than 0 and 13 dBA, tho I should read the thread before commenting.

Last night at 1AM after the post above, I checked the ambient in my office, which is ~10x11' and has a cheap carpet w/ underlay on the exposed wood floor area (and bout 1/2 the floor). It measured 15~16 dBA, which is 1~2 dBA lower than the lab downstairs (which I mentioned was 16~17 dBA); surprising because the lab is better insulated from outside noise. The only real difference is the carpet & underlay -- it absorbs more sound, so everything sounds a bit quieter.

This leads me to thoughts about anechoic chambers...

They are quiet because they block outside noise AND have a ton of damping to cut internal reflections to zero (in theory). So in reality, everything measures lower (SPL) and sounds quieter in there. Does this make anechoic readings better? Well, less interference from other noise sources is a good thing, and so is repeatability: the measurements should be repeatable from one anechoic chamber to another. Repeatability is a cornerstone of scientific method...

But a 20 dBA noise source in an anechoic might actually be 23 dBA in a more typical room.

Which brings me to Sound Power (in bels) instead of dBA SPL. Sound power is an expression of the total acoustic power radiated by a noise source, rather than a single point measurement at a specific distance. It is the acoustic measure least affetced by environment. I'd like to replace SPL reading with Sound power, but just don't have the resources to do this efficiently.

------------------

On another tangent, the World Health Organiztion published a paper which recommends an ambient level of 35 dBA for classrooms, especially with learning disabled students. Rooms in commercial/public buildings are quieted the same way anechoic chambers: Blocking outside sound and absorbing / dffusing internal ones. So if there is a lot of acoustic absorption in the classroom, the teacher's voice would also be softer -- and maybe harder to hear. You might have your 35 dBA, but would you have better learning conditions?

Post by **MikeC** » Wed Feb 25, 2004 4:25 pm

sbabb wrote:Ten posts further along in that Google thread is one from a guy named Jay Kadis who says that his main studio noise floor is 21.3 dBA in a mechanically isolated studio and it was "quite an expensive proposition to get this low a noise floor." In his control room it's 23.2 dBA with everything turned off and 33.3 dBA with his equipment turned on.

Scott

That to me sounds more believable than 0 and 13 dBA, tho I should read the thread before commenting.

Last night at 1AM after the post above, I checked the ambient in my office, which is ~10x11' and has a cheap carpet w/ underlay on the exposed wood floor area (and bout 1/2 the floor). It measured 15~16 dBA, which is 1~2 dBA lower than the lab downstairs (which I mentioned was 16~17 dBA); surprising because the lab is better insulated from outside noise. The only real difference is the carpet & underlay -- it absorbs more sound, so everything sounds a bit quieter.

This leads me to thoughts about anechoic chambers...

They are quiet because they block outside noise AND have a ton of damping to cut internal reflections to zero (in theory). So in reality, everything measures lower (SPL) and sounds quieter in there. Does this make anechoic readings better? Well, less interference from other noise sources is a good thing, and so is repeatability: the measurements should be repeatable from one anechoic chamber to another. Repeatability is a cornerstone of scientific method...

But a 20 dBA noise source in an anechoic might actually be 23 dBA in a more typical room.

Which brings me to Sound Power (in bels) instead of dBA SPL. Sound power is an expression of the total acoustic power radiated by a noise source, rather than a single point measurement at a specific distance. It is the acoustic measure least affetced by environment. I'd like to replace SPL reading with Sound power, but just don't have the resources to do this efficiently.

------------------

On another tangent, the World Health Organiztion published a paper which recommends an ambient level of 35 dBA for classrooms, especially with learning disabled students. Rooms in commercial/public buildings are quieted the same way anechoic chambers: Blocking outside sound and absorbing / dffusing internal ones. So if there is a lot of acoustic absorption in the classroom, the teacher's voice would also be softer -- and maybe harder to hear. You might have your 35 dBA, but would you have better learning conditions?

Trip · Post by **Trip** » Wed Feb 25, 2004 5:05 pm

(had my hearing tested in a room with foam everywhere several years ago, up there with my dad and the doc offered to test me)[/OT]

Whatever you do be sure the PC has enough air.

If you go watercooling, you'll always have the pump and HDD

fanless radiator.

If you go heatpipe case (ie. Zalman's) you will still have the HDD.

If you air cool and buy melamine foam, acoustifoam, solex, etc., it will be expensive, like everything else, but perhaps cheaper. It seems that even the highest powered system can be easily air cooled at <30dB and I wonder ... maybe your hatch and some foam could handle 30dB

Oh, and some good audiophile threads like this one: "Seeking best sound card w/S/PDIF" are scattered around.

Charlie Escher · Post by **Charlie Escher** » Wed Feb 25, 2004 8:42 pm

That to me sounds more believable than 0 and 13 dBA, tho I should read the thread before commenting.

Most of the guys in that very long thread were spec'ing 20-30DBA, as I mentioned. The Galaxy studio is definitely not your typical pro studio, as you'll see from their website. I'm not making any judgements, as I've never even been to Belgium, much less to that studio. Eric Desart moderates an acoustics forum here: http://forum.studiotips.com/

I would be happy to post a question or two over there, or invite him into this thread if it is germane. I guess I'll dive into that forum a bit as well, since music production fuels much of my interest in quiet computers.

I believe the Skywalker scoring stage has similar claims to Galaxy's, but I only found one graph on the web, which is ambiguous to me given my lack of proper training in this field. Look at figure 5, well down the page: http://www.zainea.com/Dynamic%20range.htm

Which brings me to Sound Power (in bels) instead of dBA SPL. Sound power is an expression of the total acoustic power radiated by a noise source, rather than a single point measurement at a specific distance. It is the acoustic measure least affected by environment. I'd like to replace SPL reading with Sound power, but just don't have the resources to do this efficiently.

Here's another approach I found while surfing around last night: http://thunder1.cudenver.edu/aes/tech/intnoise.html

Thanks for hosting such a useful forum Mike. I still have a lot of lurking and catching up to do before I'm likely to be able to offer much here. The links I've presented reinforce my suggestion to network a quiet computer in the studio with something more powerful in a machine room. This is certainly not practical for everyone though, and I'd still prefer that any computer I own be as quiet as makes sense.

Post by **MikeC** » Wed Feb 25, 2004 9:36 pm

http://forum.studiotips.com/

Hey I've chatted with those fellas before -- both the moderators Eric Desart and Scott Foster! They were active in a Yahoo audio/recording group, and I asked them advice about various sound recording / measurement techniques for SPCR -- going back some 2 years ago. I was not aware of this site, will have to check it out!

Best cooling solution for studio computer?

Best cooling solution for studio computer?

Remote Desktop Solution