It’s Not My Fault…(Part 2)

Note this won’t make much sense to you if you haven’t seen part 1…

https://www.theburningplatform.com/2017/02/12/its-not-my-fault-part-1/

First of all let me make it plain that I’ve never done this before.  I’ve built various things like a storage shed, framed a wall or two, made some book cases, trimmed in with various moldings, and generally know my way around a compound miter saw.  Most of this has been learned by trial and error, looking over the shoulder of other trades on our project job-sites, the occasional youtube video, and a small library of books I keep about on innumerable topics.  Suffice to say I’m pretty handy with tools, but very inexperienced with this particular brand of work.  There are certainly a host of ways I could have done this easier, but I’ll describe what I did.  If there are things I discovered along the way to make an improvement I’ll note that as I go.

While I may be something of a novice when it comes to framing and traditional construction I’m definitely not a novice when it comes to acoustics, physics in general, and engineering.  I live that in my every day work.  Because of this expertise, I was fairly confident that I knew how to build a high STC, high NRC, and low mechanical noise enclosure.  The great thing about understanding basic principles of physics is that fundamental knowledge generally seems to apply very well to the real world.  This opportunity to use that largely theoretical knowledge to do something practical and useful was too compelling to ignore.  I had to see if what I thought was true had any significant relationship to physical reality.

[A quick note on sound and noise.  For the purpose of this post think of noise as sound we don’t want, and sound as all sound including the sound we want along with the sound we don’t.]

Below is a brief outline of the basic principles I designed around.  This is not so brief for a blog post, but considering you can get a PhD in this stuff it is definitely a 30,000′ type overview.

STC is a measure of the amount of sound energy that transmits through a barrier like a wall or floor.  For recording, keeping noises like the sound of passing trucks from getting into the recording is critical to having a high quality finished project.  Moreover, when I recorded the original demo for the author, my recording engineer noted engine noise from a souped up 4×4 truck that parks at a house across the street from my building.  All of this energy was low in frequency which can be especially difficult to deal with when using low mass materials (like dimensional lumber).

To create a high STC wall, ceiling, and floor the key word is separation.  In a nutshell you have to decouple one side of the barrier from the other.  Sound is just changing pressure over time.  The changes in pressure are translated by our ears into electrical signals that are then decoded by our brains and filtered for the important bits.  Recording devices aren’t that smart.  They can only read the pressure changes over time which means that any and all changes go into the recording.  There are only limited ways of eliminating those unwanted sounds during and after the recording is made.  Because of those limits, the burden is on the recording space to eliminate as much of the unwanted sound as it is practical (in terms of budget, location, and geometry) to block.

The first basic mechanism for sound to get from one side of a wall to the other is for the pressure of a sound to push on the surface of the wall.  This push causes the wall surface to flex and push on the supporting structure inside the wall as well as the air inside the wall.  The air and structure then push on the other side of the wall which in turn pushes on the air on the other side of the wall.  That push on the air then becomes the sound you can hear on the other side of the wall.  That mechanism is why banging on a wall is louder on the far side than shouting.  The banging gives the wall a bigger physical push.

The second basic mechanism is resonance.  I expect everyone has experienced resonance is even if they don’t know what the word means.  Resonance is when something rings.  To ring is to vibrate at a particular frequency with more enthusiasm than at other frequencies.  (Note: You can think of  frequencies as the notes on a piano.  The low notes on the left side of the keyboard are low frequencies; the high notes on the right side of the keyboard are high frequencies.  You don’t really need to know more than that to follow along.)  Every material will vibrate longer with the same energy input at one frequency better than it will at others, and, as you might expect, that particular frequency will transmit far more easily through a given material to the other side than others will.  That is because the material likes to wiggle at that frequency, and if it is wiggling it is pushing on anything it touches.

So to make a space quiet in terms of noise from outside getting inside and noise from inside getting outside we need to keep the outside from pushing on the inside, and we need to make sure that whenever possible each element of the structure has a different natural resonance.  That way we can limit energy transmission from one thing on the outside pushing on something on the inside, and we can limit energy that gets in because of sympathetic resonances in materials on the outside and the inside.  Easy right?

One last thing: no matter how well we do building our little isolated, absorptive, and low noise room it won’t do much good if we leave any holes for sound to get in.  Think of a perfectly isolated space with a window left wide open.  It might be incredibly isolated, and perfectly constructed otherwise, but the maximum performance of the room will be equal to the amount of noise coming in through the window.  Any time you drill a hole for a wire or an air duct you are creating a window for sound.  That means we have to close as many of the little “windows” as we can find by blocking them up with caulk, making sure they are as small as possible, and making sure that there are not “windows” in the same spot on both sides of the wall.  By now should you may be thinking this may be a bit hard to do; you would be right.

The second type of noise we have to worry about is that of sound generated inside of the space hanging around longer than we like.  In this case we’d like for any sound originating inside our recording space to go away as close to instantaneously as we can manage.  In recording, if you want to sound like you are singing in a cathedral that is easy to add later as an effect.  If you are actually singing in a cathedral you will have a very difficult time pulling the cathedral out of the recording later.  It isn’t completely impossible, but it is neither easy to do, nor is it completely possible.

When talking about reducing the accumulation of sound inside a space you are generally talking about NRC.  NRC is a measure of the amount of sound absorbed by materials inside a space and not allowed to reflect back into it.  If you’ve ever been in a large gymnasium with hard floors, a hard ceiling, and cinder block walls you probably noticed that any sound you make will hang around for at least several seconds.  If you keep making a constant noise you’ll notice that the sound will build past the level you are generating into a wash of new and old sound blending together into a jumbled mush.  That build up is why it can be difficult to understand someone talking from a few feet away even though you can clearly hear them speaking.  The wash of sound you get back from the room reflections overpowers and obscures the sound that travels from their mouth straight to your ears.  You can easily hear the first word or two, but then the rest is washed away by the sound building up in the room.

In a recording space or in a gymnasium the idea is basically the same.  We want to absorb the sound so that reflections are diminished to the point of our desired use.  In a gymnasium we want to reduce it to a level where it isn’t dangerously loud, oppressive to experience, or impossible to understand critical speech like announcements.  It isn’t generally a concern about the exact balance of frequencies in the reflections we absorb unless there is a strong musical component to the activities that are hosted in there.  On the other hand in a “dry” recording space we want to come as close to obliterating all reflections across as wide a range of frequencies as we can afford.  We try hard to ensure that we absorb evenly across the widest range of sound we can.  This is easier said than done since low frequencies are notoriously more difficult to deal with than higher frequencies are.

Another aspect to noise within a space is how coherent it is.  Let’s go back to our imagined gymnasium.  What happens if you stand in the middle of the room and clap?  Well you’ll hear the sound bounce around the room for a while and then die out, but you’ll also hear a stuttering sound that is the sound of the clap bouncing back and forth between the parallel walls and between the floor that is parallel to the ceiling.  This stutter is actually called a flutter echo, and it is very detrimental to your ability to hear clearly because it shows up as discrete pulses of sound in rapid fire succession as it bounces back and forth between the walls and gets picked up by your ears along the way.  You’ll notice these pulses more than you would notice a more diffuse sound at the same level.  That means for our purposes they are bad, and we want to eliminate them.

Parallel reflections can also be resonances within a room.  Since resonant sound tends to hang around a lot longer than non-resonant sound, and we want the sound to go away as quickly as possible we want to eliminate as many parallel surfaces as we can (and by extension resonances).  This means we have to try and make sure no wall, ceiling, or floor is parallel to any other wherever we can.

So to make our space quiet in terms of the sound generated inside the space we have to make sure that wherever possible we are absorbing sound energy across as wide a range as we can, and that any reflections that do happen don’t happen in a way that are parallel.  This is easier in some ways than dealing with the STC problem, but as you’ll see it does make it a good bit harder to build since we can’t build a nice easy rectangle sided box (parallelepiped).  Well we can build a rectangular box, but since our theory says it would suck, why would we bother?

The third type of noise we have to worry about is the noise we generate in using the room.  This shouldn’t be confused with the sound we make when singing, talking, and/or playing an instrument.  Those sounds we generally want and are what we are trying to record.  The noise we’re concerned with are the noises of computer fans, squeaky chairs, foot falls, shifting papers, typing keys, and similar things.

Some of these we can deal with by not recording while we create them.  For example adjusting a microphone, coughing, drinking water, etc are all things you can hit “pause” do, and hit “un pause” and go on with your recording.  Those just require the self-awareness necessary to hit the pause and un-pause buttons.

Other noises like computer fans can be eliminated by moving the computers out of the space and using remote control (wireless keyboard/mouse/etc) to operate them.  This has the side benefit of taking away the heat which is pretty important as well.  The big deal here is just to plan for the necessary connections (and how to get them into the space while making sure that the holes we create are created in a way to make it difficult for sound to get in).

The last, and hardest, is how to manage the air conditioning.  Air flowing through a duct is inherently noisy.  The noise is caused by the turbulence in the air as it flows through and around obstructions, and by the noise of the fan that makes it move in the first place.  The faster it flows the higher the noise and generally the higher the pitch of the noise.  Worse, sound loves to travel down ducts.  Just look at a french horn.  The sound is created at the mouthpiece of the horn, travels down 12′ or so of brass ducting (pipe) and then exits the bell.  You’ll note it doesn’t seem to get quieter as it comes out the other end.  That means it is important to make sure the ducts we use absorb the sound, bend a lot to keep from acting like a window to the sound, a that the flow of air is as slow as possible.  BUT at the same time we need to be sure we have adequate air flow to keep from overheating inside what will become a *very* well insulated little room.

So here is our checklist of design criteria:

1. Outside surfaces and their supporting structure can’t touch the inside wherever possible.
2. Make use of dissimilar materials wherever possible. (ie one side of the wall needs to be covered with a different material or thickness than the other side of the wall.
3. Be sure that no surface within the space is parallel to another surface inside the space.
4. Use broad spectrum sound absorption to minimize absorbing one range of frequencies more than another.
5. Use absorbers over as much square footage as possible inside the space.
6. Seal up any holes or gaps  and make sure that any holes that must remain unsealed (ie airducts) follow an absorptive path that makes it hard for sound to follow
7. Ensure there is adequate airflow at low speed so the person inside the recording room doesn’t overheat while at the same time ensuring the airflow is super quiet.
8. Eliminate any noisemakers like computer fans by moving them outside of the recording space, but plan for the necessary connections to allow them to be used while inside of the recording space

I think that is it…easy peasy.

(If you like this stuff enough to read this far stay tuned for part 3)

Reposted from http://thesonicsingularity.com/the-gist-of-shhh/