Tuning a Tracking Booth (video/audio files included)

[O.G. Killa]

the density of the product effects HOW WELL it absorbs at a specific frequency...

When you look at the NRC charts for these products, you'll see they rate them usually in octaves, so 125Hz, 250Hz, 500Hz, 1KHz, 2KHz, and 4KHz. The number associated with each frequency is the PERCENTAGE absorbed by the product.

So for example, 2" auralex wedge foam at 250Hz has an NRC of 0.31.

http://www.foambymail.com/pdf/Aurale...oam-Wedges.pdf

That means that WHEN 2" Auralex foam CAN absorb at 250Hz (based on placement) it will ONLY absorb 30% of the the energy at 250Hz. If you look at that same test chart, you'll see that at 1.5KHz, Auralex 2" foam only absorbs 0.62. So it doesn't completely stop the sound from bouncing off at the frequency.

Here is the NRC ratings for Roxul

http://www.roxul.com/residential/products/roxul+safe%27n%27sound™

You can see at 250Hz it absorbs 0.96, or 96% of the sound that hits it and at 1.5K it absorbs around 1.04 or 1.05, or 104% or 105% of the sound. I know it sounds weird that it could absorb over 100%, but because of the way the test is conducted in an anechoic chamber when a panel absorbs 100% it can sometimes come out as more than 100%. but for these charts, anything that is 1.0 or higher is 100% absorption.

Moving the panel off the wall so there is a air gap is a double edged sword. Here's what happens when you do that...

By moving the panel away from the wall, you are in essence making the panel seem thicker to the sound coming at it. If you take a 2" panel and hang it 2" off the wall, the panel's lowest effective frequency drops and is the same as though the panel was 4" thick.

BUT!!! it doesn't absorb 100% to what the NRC numbers are. You are basically "tuning" the absorption and creating a little bit of a bell curve since you now have a resonant space behind the panel. If you fill behind the panel with more insulation, then it absorbs everything. But leaving it empty air means that frequencies that have their 1/4wavelength WITHIN that 2" air gap AND also have their 0º or 180º angles within the 3"~4" distance of the wall, will not get absorbed by the panel. The reason being, when they reach the panel they are in a state of pressure, not movement, and pass through the absorber as pressure without any hinderance.

An example of this... walk around your control room with some hiphop or other bass heavy music playing... notice how there are spots in the room where the bass goes away significantly and other spots in the room where the bass is very loud? The spots where the bass goes away are the spots where the bass sound waves are at a null 180º point in the room. The spots where the bass is loudest are the spots in the room where the bass is at 90º to its original self. Putting a bass trap in the spot where the bass went away will have little to no effect on the overall bass in the room. Putting a bass trap where the bass is loudest will have the most effect on the overall bass in the room...

The same applies to high frequencies at wall boundaries. There are spots right up against the wall where high frequencies will be null and sound non-existant (it's just hard to hear them that close to the wall), and the panels at those spots will be ineffective at stopping them, unless they are filled all the way to the wall.

So let's pick a frequency to illustrate what I'm talking about...

Let's try to find a frequency whose 1/4 wavelength might be between 1" and 2" and whose 1/2 wavelength is between 3" and 4".

Frequency = Speed of Sound / wavelength

1.75" = 1/4 Wavelength
3.5" = 1/2 Wavelength
7" = Wavelength

Frequency = 13560 / 7

Frequency = 1937.14 Hz

So if you take a 2" panel and hang it 2" off the wall will absorb down to 847Hz BUT at around 1937Hz the panel will not absorb ANYTHING even though it normally would/could. As you go up in frequency the panel will start absorbing again. And if you used a 4" panel flush to the wall or two 2" panels it would absorb everything and not have little "holes" of non-absorption.

Make sense? Am I explaining this well enough?

Anyway... On to your room... Are you doing sine wave sweeps to see that 1.5KHz is still resonating or are you just doing it by ear using a guitar recording?

Is the room COMPLETELY covered in 3" roxul? Meaning ceiling, and all four walls (including the door) with no gaps? Obviously the floor won't be, but I can't imagine the floor is creating a peak at 1.5K... but you never know.

Try this, see if the distance you are from the mic changes the 1.5KHz resonance. Also put a speaker in the room and try making pink noise and see if 1.5K on an RTA really is peaking. You can also do burst tests (short 10ms or 20ms bursts of noise) to see if there are any reflections coming back at the mic and then you can determine what they might be based on the timing. Also try putting an extra piece of roxul on the floor between you and the mic when recording guitar to rule out any reflections off the floor.

[Guitar Guy]

Thanks again!

I was wrong on the Roxul, it is is 3.5 inches, not 3. We left a one inch air gap between the floating walls and drywall. Yesterday we spent most of the day installing the 2" foam over the burlap-covered Roxul. It turns out I'm 6 panels short so I ordered the material needed to finish it out.
However, I did do another test of the guitar and the sound is MUCH tighter with less resonance! It isn't where it needs to be by a long shot, but I'm very encouraged by the direction we are headed in. I did want to ask you if you thought my 18" pc monitor in the room is affecting sound at all? I'm having to run my DAW from inside the room so the monitor is necessary, but I'm sure that I'm getting reflections off the screen. For the guitar's sake, I'm going to mount it higher to avoid any initial reflections back into the microphone.
The resonance does change with mic placement, but not in such a way that I can find enough isolation from the honkiness. I'm very green with technology, and am not sure of how to generate pink noise or connect a RTA software to my speakers. I will try to research that more today. All I've been doing is recording the guitar and using an eq to sweep the spectrum, taking note of the severely audible spikes in those mid range areas.

[Emcha_audio]

Originally Posted by Guitar Guy

Thanks again!

I was wrong on the Roxul, it is is 3.5 inches, not 3. We left a one inch air gap between the floating walls and drywall. Yesterday we spent most of the day installing the 2" foam over the burlap-covered Roxul. It turns out I'm 6 panels short so I ordered the material needed to finish it out.
However, I did do another test of the guitar and the sound is MUCH tighter with less resonance! It isn't where it needs to be by a long shot, but I'm very encouraged by the direction we are headed in. I did want to ask you if you thought my 18" pc monitor in the room is affecting sound at all? I'm having to run my DAW from inside the room so the monitor is necessary, but I'm sure that I'm getting reflections off the screen. For the guitar's sake, I'm going to mount it higher to avoid any initial reflections back into the microphone.
The resonance does change with mic placement, but not in such a way that I can find enough isolation from the honkiness. I'm very green with technology, and am not sure of how to generate pink noise or connect a RTA software to my speakers. I will try to research that more today. All I've been doing is recording the guitar and using an eq to sweep the spectrum, taking note of the severely audible spikes in those mid range areas.

This link will provide you with a sample of white & pink noise, from what I remember, pink noise is more useful for audio application. I think you can download the samples to use within your daw.

And here's an rta software and an other one.

You'll need a measurement mic that you will plug into your interface to your computer. Run the rta software, while the pink or white noise is playing in your daw.
Tests from realtraps shows that the nady cm 100 is pretty good, the berhinger ecm8000 is pretty good also. Comparing the graphs shown by realtraps, it seems they are better (flatter) than the dpa's or other high end mics, but in any case, for the price it will help you. You're biggest problem will remain having a good system (as flat as possible monitors) to play back the pink noise, if you have the possibility to test the monitors before hand, some shops have calibrated listening rooms, you will be able to see what's the respons of the monitors and take that into account when you test your room.

As for your screen, if you can tilt it up (away) from the mic, that should help.

[Guitar Guy]

Originally Posted by Emcha_audio

This link will provide you with a sample of white & pink noise, from what I remember, pink noise is more useful for audio application. I think you can download the samples to use within your daw.

And here's an rta software and an other one.

You'll need a measurement mic that you will plug into your interface to your computer. Run the rta software, while the pink or white noise is playing in your daw.
Tests from realtraps shows that the nady cm 100 is pretty good, the berhinger ecm8000 is pretty good also. Comparing the graphs shown by realtraps, it seems they are better (flatter) than the dpa's or other high end mics, but in any case, for the price it will help you. You're biggest problem will remain having a good system (as flat as possible monitors) to play back the pink noise, if you have the possibility to test the monitors before hand, some shops have calibrated listening rooms, you will be able to see what's the respons of the monitors and take that into account when you test your room.

As for your screen, if you can tilt it up (away) from the mic, that should help.

Awesome! Thank you for this very helpful information and advice!

[O.G. Killa]

How far away is the monitor to you? Yes it can cause some problems but so long as it is behind the mic it shouldn't be too bad. If I remember correctly the TLM103 gets somewhat omni directional around 4K and then becomes hypercardioid by 10KHz (or is it 16KHz)?

Anyway... sweeping an EQ might not be the best test. Moving one of your monitors into the booth, connecting it to a line output (could use the headphone send, just connect it straight to your daw instead of going through a headphone amp) and use one of the measurement mics that Emcha mentioned.

Pink noise is equal noise per octave. White noise is equal noise per frequency. Because of that, white noise actually gets 3dB louder with each octave and so doesn't really sound "flat". Pink noise attenuates each octave by 3dB to compensate and so pink noise has equal amounts of bass, mid and treble frequencies per octave and more closely simulates the sounds we hear from most instruments.

Sine wave sweeps can be helpful too. Because as the sine wave increases in pitch, it's volume is supposed to be even. If you hear it increase or decrease in volume, it is most likely from some sort of acoustic problem in the room.

bursts are just very short bursts of pink noise. You can duplicate the pink noise file and cut it into a couple of short bursts. The shorter the better, but you'll find that as you get extremely short, your speaker probably won't even make any sound because the speaker itself is too sluggish to reproduce it. The better your speaker is, the better its "transient response" is and the smaller you can make the burst.

Anyway, if you can make a 1ms or smaller burst and have the speaker play it back that is best. If not, keep increasing the size until you can hear it from your speakers. You might need to make a 10ms (or some other number) burst in PT in order to get a 1ms burst out of the speaker. It'll sound like a "tick" or "click". You record the output of the speaker with the measurement mic into protools. You can set the speaker where your guitar normally is and the mic where your tlm103 normally is. Set the volume of the speaker kind of loud. Loud enough that the ticks would interfere with you talking.

When you record, the first recorded tick will be the direct signal from the mic. You can calculate the time it took to go from the speaker to the mic based on the distance between the speaker and the mic. (sound travels at 13.56 inches per millisecond or 1.13 ft per millisecond). If the mic is closer than 13.56 inches, then the direct sound should hit the mic under 1 ms. Then in the recording you will see any other reflections of the sound bouncing around the room and going past the mic afterward.

Measuring how many milliseconds after the initial click can tell you the distance the click has traveled before it passed by the mic.

So for example, if you are in a 5'x5'x8' room. The speaker is 2' out from the wall behind it (3' from the wall in front of it). The mic is 1' in front of the speaker (2' from the wall behind the mic). When the speaker makes a burst, you'll see the initial spike in the recording (that happened 1ms after the sound was produced). you will then see another spike in the recording about 4 or 5ms later. That 4.5ms is the click bouncing off the wall behind the mic and passing back over the mic.

If the speaker is 3' from the wall in front of it and the mic is 2' away from the same wall, the sound will travel 3', hit the wall then travel another 2' to the mic, for a total of 5'. 5' x 1.1ft/sec = 5.5 ms THEN, you have to subtract the 1ms it took the direct sound to hit the mic, creating the first spike in the recording. So when measuring from the first recorded spike to the next one (assuming the reflection off the wall behind the mic is the second spike), the time difference will be 4.5ms give or take. This will also happen off the ceiling, floor, wall behind the speaker, walls to the sides and so on. It'll also have multiple surfaces included, for example from the speaker to the front wall to the ceiling and back to the mic. Or from the back wall to the side wall to the mic.

So why is this important? Because it can help you figure out what is possibly reflecting back and causing phase interference (those peaks and dips in the sine sweeps).

It's a little like detective work. You have to do a lot of tests and study the results and then use those as clues to help you find the culprits that are causing problems in the room.

That is also why I suggested moving the mic around. If the resonance is because of the room and not because of the instrument, moving the mic around would change the resonance (just like walking around the room you hear the bass get louder and softer).

[O.G. Killa]

also, i'll throw out one other little tidbit of info now that you are starting to get used to all this...

The calculation I have you for figuring out the lowest absorbed frequency is a ballpark number, not an exact one. Why is that?

the reason being it is an exact number but only for sound that hits the panel square. If sound is coming at the panel from an angle, the panel will appear to be a little thicker. So a 3.5" thick panel, could potentially appear to be 4" or even 4.5" thick in some parts of the room depending on the angle of the panel to the guitar (or vocalist).

What this creates is more of a falling off in absorption instead of a strict start/stop point.

Also I know you mentioned that there is a 1" gap between the "floating wall" and the drywall. I'm not sure what you mean by that. Is the floating wall the pegboard? Or did you do a wall within a wall construction with two separate layers of drywall? The drywall, once sealed acts as a boundary that most sound doesn't pass through easily and will reflect off it. Depending on what you did for your floating wall, it could either be replacing the drywall as the boundary or doing nothing at all.

Is the floating wall on the inside and drywall on the outside of the booth's recording space or is the drywall the inside face of the booth and the floating wall outside it?

[Guitar Guy]

You're right, floating wall was a bad choice of words. The 2x4 metal studs are filled with the 3.5" Roxul, and behind that layer is one inch of dead space before the next layer which is the original drywall used in the house construction. You really know your stuff and I want you to know that this has inspired me to become more educated with calculations and what it all means to the grand picture. I will follow up with how it's going soon.

[tonejunkee]

Great thread as I am currently wrestling with Rockwool for a drum room