Case modding and parallel surfaces (Physics of Sound 101)

Green Shoes · Post by **Green Shoes** » Fri Apr 08, 2005 3:42 pm

A couple posts recently got me to thinking of my experiences in some recording studios around town....

....has anyone ever tried building a case with no parallel sides? Any recording studio worth its salt doesn't have a single wall/ceiling/floor parallel to any other surface. When sound exists in an enclosed space it bounces around from one surface to another, generally losing amplitude in the process due to atmosphere, etc. But when parallel surfaces exist, any wavelength that is a division of the distance between those two parallel surfaces can bounce back and forth for a much longer period of time; these are called "standing waves". They suck.

It seems to me that the problem is not only does a computer case have six very hard parallel sides, everything else inside of them is mounted parallel as well. The mobo, drive enclosures (4 more sides right there), PSUs, etc. Sound dampening is okay but a physicist would tell you that just not building the thing parallel to begin with will get you much farther. I suppose I understand why case manufacturers haven't picked up on this, as it can be much more expensive to build something composed entirely of rhomboids than nice, simple rectangles. But if I'm going to shell out ~$200 for something like the P180 I want them to have taken this into account.

So, has anyone ever attempted a custom case build without parallel? (Sorry, terrible pun, but I couldn't resist).

m0002a · Post by **m0002a** » Fri Apr 08, 2005 4:12 pm

I don't see much of a benefit. The difference between a recording studio or listening room is that the reflected sound must be attenuated (reduced) the first time it hits a wall surface because the microphone or listener is in the same enclosed area that the sound originates from.

For a computer case, we are outside the case where the noise originates from, so if the sound bounces around a few extra times inside the case it makes no real difference so long as it does not escape to the outside. Obviously, it does help somewhat to keep the sound off the metal walls of the case, to avoid the sound from be transmitted outside of the case, so a sound barrier inside the case does usually help some.

Also, standing waves are low frequency problems, and I am not sure how much of annoying computer noise normally falls in to that spectrum.

Just my opinion.

wim · Post by **wim** » Fri Apr 08, 2005 5:08 pm

if you're talking about sound being prolonged by bouncing between parallel walls, i think i know what you mean. it's like when you clap your hands loudly in a quiet room, you hear a kind of hum tone lingering for as long as a few seconds (some rooms are acoustically so poor that this effect makes lectures impossible) - but the standing waves you're talking about have wavelength of the order of meters. i'm not sure it could happen inside something as small as a computer case (probably the natural frequency of the walls, rather than the chamber, is dominant)

Thy Left Elbow · Post by **Thy Left Elbow** » Fri Apr 08, 2005 5:47 pm

The standing waves achieved in a normal mid-tower case would start at approximately 680 Hz given that either the height or the depth is 50 cm or so. This would be the lowest possible resonant frequency if there's an unbroken soundpath between two sides of the case. The chances for that are very slim, however, so the lowest possible frequency would probably be a lot higher, I'm guessing about 1700 Hz or so, which would be the result of a distance between the sides of the case of 20 cm. This is not a common range for disturbing noises in a PC. Most of them have lower frequencies. However, you might be able to catch the higher frequencies in fan noise and reduce that without the need for damping material but personally I don't think it would be worth it.
Still, it's an interesting concept for one obsessed with silencing and if I have nothing better to do I might give it a try sometime. I got a few ideas from reading this thread and writing so who knows, someday there might be such a case...

Green Shoes · Post by **Green Shoes** » Fri Apr 08, 2005 7:38 pm

m0002a wrote:I don't see much of a benefit. The difference between a recording studio or listening room is that the reflected sound must be attenuated (reduced) the first time it hits a wall surface because the microphone or listener is in the same enclosed area that the sound originates from.

I think you mean diffusers, which you usually see plastered all over the room; they have a bit of a different (but still very useful) purpose.

Thy Left Elbow, you're right, they would be pretty high frequencies. And granted, most of the discernible noise is probably more along the range of 200Hz, which would need a double plate-glass to stop. But what I wonder is how much higher frequency noise our brains are just tossing out. We still hear it as overall sound level, but just attribute it to what we can single out. It's like what most people think of a waterfall: the "rumble" that can be heard from far away is perceived as only low-frequency, but a glance at a spectrograph tells a very different story.

You're right, it's a project for someone with a lot of time (and maybe an anechoic chamber) on their hands. If anyone has the latter, let me know

.

~El~Jefe~ · Post by **~El~Jefe~** » Fri Apr 08, 2005 8:39 pm

If this is the case, in your, um, case, then the following would negate it.

A super sound proofing mat from soundproofing.org, liek the 1/4 inch thick stuff, would kill the sound if on 3 sides of the case.

However, anyone's case I know isnt really a parallel thing. There's a lot of random crap in these things. Cables, cards, frames, etc.

The best way to ensure that the sounds do not amplify is to make a case smaller, wouldnt it be? currently, everything in a case is square, the cd room has to be mounted somewhere sensible, the pci slot and interface pluggy things are at 90 degrees from the motherboard.... all components are square... if you put things on angles, it would only increase the amount of empty volume in the case as it would have to accomodate very boxy pieces of equipment inside.

Thy Left Elbow · Post by **Thy Left Elbow** » Sat Apr 09, 2005 4:51 am

Green Shoes wrote: You're right, it's a project for someone with a lot of time (and maybe an anechoic chamber) on their hands. If anyone has the latter, let me know .

I might have both. I'm an acoustics major and I've been thinking of what I should do this fall. I've had a few projects in mind but this one might be worth a little attention. Then, as luck would have it, we happen to have an anechoic chamber at the university which I'm pretty much free to use whenever I want.
As El Jefe says, there is definately a problem with incorporating current designs as most components are square. However, I think you might be able to reduce the effective volume of a case by only using rhomboid shapes by applying something along the lines of stealth technology. Granted, you will probably get a case with a pretty large footprint but if you make the sides go inwards you'll reduce the volume by quite a bit (It's pretty hard to explain in writing). The only area you really need a 90 degree angle is the back of the motherboard unless you design some kind of riser for all the expansion slots. Drive bays can be designed in different ways with a little thought. I'll try to make a quick sketch sometime this weekend to show you how I mean.

~El~Jefe~ · Post by **~El~Jefe~** » Sat Apr 09, 2005 4:59 am

Bowing the window side of a case and making the the front on an an angle, with a slope downwards from the front to the back would be a way to kill this sort of amplifying of sound. the front of the case would have to have verticle mountings of cd roms and suspended hd's hung perpendicular to the ground.

This allows the rear of the case to be 90 degrees to the motherboard tray. it is possible that only like 15% increase in volume would occur in such a case.

Also, the bowed side could be really neat looking

Thy Left Elbow · Post by **Thy Left Elbow** » Sat Apr 09, 2005 5:24 am

Yeah, that's pretty much what I'm considering. If you then decide to bow the sides inwards you get a decrease in volume.

burcakb · Post by **burcakb** » Sat Apr 09, 2005 5:48 am

So that Feng Shui motherboard isn't a bad idea after all

Reachable · Post by **Reachable** » Sat Apr 09, 2005 7:56 am

I think you could be on to something, Green Shoes.

I'd like to see a case in the shape of a tetrahedron, the form that natural structures take for the most strength and economy.

It would involve modifying drive cages necessary to hold drives at the proper attitude. The case would be very stable, and have a huge footprint for it's volume.

I'm a great tetrahedron fan and would like to see more artifacts in that shape. For something as complex as a computer, you could compare all sorts of properties against our familiar hexahedrons, and maybe find out something illuminating.

dano · Post by **dano** » Sat Apr 09, 2005 8:06 am

This is going to be horrible for airflow, if the case gets smaller towards the back that means the air is going to have to be compressed.

Gorsnak · Post by **Gorsnak** » Sat Apr 09, 2005 9:06 am

I don't think that this whole parallel surfaces analysis thing here holds any water. I mean, I understand the acoustic principle under discussion, and it's perfectly valid. Ran into a very nice illustration of it around Christmas, when we hung a bunch of absorbing panels in the dining room of a nursing home. It was a largish room with a tile floor and an 8' ceiling. It would have been fine if it had been suspended tile ceiling or the like, but the ceiling was just painted drywall. Two big hard parallel surfaces, and it was almost painful spending a lot of time in the room - which was why we'd been called in, no doubt. We covered somewhere around 1/3 of the ceiling with absorbing panels, and the place was much improved.

But enough of my digression. My point is that there aren't really any parallel surfaces of the required sort inside a computer case, and even if there were it would only be a real problem if you commonly sat with your head inside the case. The clutter inside a case - mobo, drives, etc - is going to act as a diffusor and break things up. The only really large, flat, hard surface (usually) is the inside of the left side, and it's faced by a lot of irregular surfaces. Even if the inside of the case was conducive to setting up this sort of ringing reverberance, the whole issue is one of sound being trapped between the two surfaces and only being damped by the air between them. But if the sound is trapped inside the case, it's not getting out, and hence not an issue. I suppose if the frequency of the standing wave between the surfaces was the same as the resonant frequency of the sides of the case then you might get an annoying hum going on, but you're far more likely to set off case resonance from fan or drive rotation. In short, I wouldn't worry about cases with parallel surfaces. Any acoustic disadvantages are vastly outweighed by practical considerations. If you are still concerned, put some absorbing on the inside of any largish flat areas (which is a good idea anyways).

halcyon · Post by **halcyon** » Sat Apr 09, 2005 10:13 am

Some basic acoustics 101:

Standing waves are formed in every room, regardless of shape. In a room with non-parallel surfaces, they are just more difficult to calculate, but they still form (both parallel and axial modes).

To be really effective the sound damping material needs to be circa 1/4 of the wavelength of the sound to be damped. Calculate yourself how easy it is to damp lower frequencies :)

So, I wouldn't so much worry about parallel surfaces. I'd worry about self-resonating frequencies of various components, structural vibration, isolating vibration sources structurally and trying to achieve air tight seal in the case towards the listener direction (if you can).

A good source is: Handbook of Sound Engineers (and to some extent: Master Handbook of Acoustics).

rgds,
halcyon

m0002a · Post by **m0002a** » Sat Apr 09, 2005 10:38 am

Standing waves are a problem when trying to obtain flat frequency repsonse (such as in audio listening room), but I don't believe they actually increase to the total power output of a signal. If I am wrong about this, please correct me.

Gorsnak · Post by **Gorsnak** » Sat Apr 09, 2005 10:44 am

m0002a wrote:Standing waves are a problem when trying to obtain flat frequency repsonse (such as in audio listening room), but I don't believe they actually increase to the total power output of a signal. If I am wrong about this, please correct me.

If they did, they'd be violating the conservation of energy. That'd make it pretty long odds that you're wrong.

m0002a · Post by **m0002a** » Sat Apr 09, 2005 10:53 am

Gorsnak wrote:If they did, they'd be violating the conservation of energy. That'd make it pretty long odds that you're wrong.

Yes, but in this thread, anything is possible.

Green Shoes · Post by **Green Shoes** » Sat Apr 09, 2005 3:27 pm

m0002a wrote:Standing waves are a problem when trying to obtain flat frequency repsonse (such as in audio listening room), but I don't believe they actually increase to the total power output of a signal. If I am wrong about this, please correct me.

Nah, you're right, they just tend to dissipate a little less quickly. Gorsnak's right, although an Indiana-Jones style strike-the-bell-listen-to-it-gradually-get-louder style design would be cool if it didn't defy the laws of physics

.

halycon wrote:So, I wouldn't so much worry about parallel surfaces. I'd worry about self-resonating frequencies of various components, structural vibration, isolating vibration sources structurally and trying to achieve air tight seal in the case towards the listener direction (if you can).

I guess I am answered. Oh well, it was just a theory, thanks all for the thought you put into it. I'd still like to mess around with it someday, but it can wait.

halfpower · Post by **halfpower** » Mon Apr 11, 2005 10:15 pm

Sound reflecting inside the case may not be the only issue here at hand. Acoustic waves will travel within the walls of the case. In this instance having rectangular case walls is probably a good condition for the development of standing waves.

ronrem · Post by **ronrem** » Fri Jul 22, 2005 5:10 pm

I once helped build an acoustic echo proof room for a marine lab studying the sounds of seals and sealions.. Walls were a stainless + foam sandwich. All interior surfaces had these sawtooth triangles mounted,they were stainless perf stuffed with fiberglass. Not cheap.

Anyhow the pattern of interior angles and the absorbing fiberglass meant NO echo-none-as well as no outside sound coming in.

For our purposes-goo one of those corrugated type backpacker pads to the inner of the door and the floor and the echoes are killed----where that matters is than SOUND WAVES do not bounce out-mainly come out in a direct path...so no front fan,or they resonate through the walls...so you carpet the ottside? You build a case of Particle board?

teapot · Post by **teapot** » Thu Jul 28, 2005 9:33 am

Gorsnak wrote:
m0002a wrote:Standing waves are a problem when trying to obtain flat frequency repsonse (such as in audio listening room), but I don't believe they actually increase to the total power output of a signal. If I am wrong about this, please correct me.
If they did, they'd be violating the conservation of energy. That'd make it pretty long odds that you're wrong.

That's not necessarily true; it is conservation of energy not conservation of sonic power. If two signals both of equal power are cancelling each other then there is no noise heard. If two signals of the same power are reinforcing each other then a louder noise is heard. Energy is conserved in both cases, but in the first some heat is generated in the medium.

ronrem · Post by **ronrem** » Fri Jul 29, 2005 9:37 am

Standing waves in a speaker cab,or a recording room are an issue because there is a mutation,or coloration of the pure source sound. Here,the issue is amplitude and echo bounce,we don't want a sound wave to bounce full strength out an opening. the box can be a cube if internal walls are absorbant and with a textured or ribbed surface,if the airflow in/out paths are not a straight line. think of a silencer on a pistol or the muffler on a car, Sound waves get bounced around and degraded

justblair · Post by **justblair** » Mon Aug 08, 2005 10:32 am

I used to be a hi fi tester, many moons ago, and I can tell you your as we say in these parts "On to plums" on this idea. Listening rooms, concert venues etc are designed off the parrallel to prevent noise cancelling as many more clever people than me have already explained.

A good example of this is the Sidney opera house, of which in glasgow we have a similar building the "Armadillo". I can tell you that the principle works. I have been going to Jools Holland concerts for years, and the non parrallel venues sound way better than the older square edged venues when your sat inside.

When you nip out to the bar to beat the half time ques though, the sound through the walls is just as loud as a square hall built from similar materials...

JB