John Sayers' Recording Studio Design Forum

sorry for the bad pic quality, camera phone.

traps are built to Ethan's spec including

2x low bass
1x mid/high bass
1x mid/high absorber (not panel trap)

the high bass trap has 1/8" hardboard instead of 1/8" plywood. This material is a bit heavier than the plywood, but we couldn't find 1/8" plywood.


The insulation is 703, and is spaced away from the wood surfaces as prescribed.

should be fun.

_________________
Technical director, Green Glue Company
(Audio Alloy)
All posted content copyright Brian Ravnaas or Green Glue Company except as noted and may not be reproduced without permission.

the first set of tests is completed.

a reference wall of

5/8" drywall
2x4 studs 24" OC with 3.5" R13 fiberglass insulation
5/8" drywall

had added to it the full set of Ethans traps, these being:

2x deep bass trap
1x high bass trap
1x mid/high absorber


Then, per discussions above or on Studiotips, the mid/high absorber was removed.


The results are presented below. I can't discuss in any length as we're still here at the lab working, but i thought i'd post them on-the-fly.


The overall weight added by all the traps was about in line with adding another layer of drywall (just a bit less), and the overall gains at low freq's were smaller than adding another layer of drywall.

However, and a bit surprisingly, at higher freq's the traps helped and (the surprising part) the mid/high absorber seemed to help at high freqs as well.

_________________
Technical director, Green Glue Company
(Audio Alloy)
All posted content copyright Brian Ravnaas or Green Glue Company except as noted and may not be reproduced without permission.

Brian:



I was expecting it to be more mass, so it would improve TL above 250hz.

From IRC 02-108


What's the 1DOF frequency estimate of the panels (depth, surface mass)?
Or more to the point, is it 65hz ?

1DOF

http://www.ethanwiner.com/BTPlans.gif

Assuming a 1/4" sheet of plywood usually weighs about 25 pounds, that's 0.78125 psf.

d = 28900 / (f ^2 * m)
where
d = depth in inches
f = resonance frequency in hz
m = surface density in pounds per square foot

becomes

f = sqrt( 28900 / (m * d) )


Deep Bass Trap: f = sqrt( 28900 / ( 0.78125 * 3.5 )) = 102 hz

Mid High Absorber: f = sqrt( 28900 / ( 0.78125 * 2.5 )) = 121 hz

High Bass Trap: f = sqrt( 28900 / ( 0.78125 * 1.5 )) = 157 hz


Clearly those are all wrong, at least on TL.

Bob,

> Deep Bass Trap: f = sqrt( 28900 / ( 0.78125 * 3.5 )) = 102 hz
> Mid High Absorber: f = sqrt( 28900 / ( 0.78125 * 2.5 )) = 121 hz
> High Bass Trap: f = sqrt( 28900 / ( 0.78125 * 1.5 )) = 157 hz

When I measured the high- and low-bass traps at IBM (third octave resolution unfortunately), the low-bass type peaked at 100 Hz and the high-bass at 180 Hz. So I'd say your calcs are pretty close. Note that the "mid high" absorber is one inch of 703 only with a 2-inch air gap. So it's not a resonant trap, as Brian correctly noted above.

> Clearly those are all wrong, at least on TL. <

Or put another way, this type of trap does not appear to reduce isolation at its resonant frequency, as has been suggested.

--Ethan

_________________
The acoustic treatment experts
Buy my DVD

Brian,

Outstanding. I owe you much more than a beer my friend.

--Ethan

_________________
The acoustic treatment experts
Buy my DVD

Ethan:

That's not the way I read what's happened.

The 1DOF calculations should correspond with a lab measurement if the walls of the lab were made out of 1' thick concrete or something (similar to infinite, relatively speaking).

In this case the test wall is fairly thin. The trap functions in combination with the wall, to create at least a 2DOF with a new lower MSM system, in this case around 63hz. That's where it resonanates, and probably also where the system absorbs the most from the room. This lower resonance is clearly shown in Brian's TL chart.
I was expecting it to also create a higher resonance, but that did not appear on the TL chart.

At 63hz, with the two resonant traps in place, the TL fell from 16.5dB down to 13.9 dB, which is a 2.6dB drop, or 16% (100*2.6/16.5) loss in TL. (I don't know what the margin of error is in this test. Presumably comparing one wall to another right after it, the relative error would be low). Hmm, I don't think % is a correct way to manipulate this (0dB isn't zero, etc), but 2.6dB is significant there.


mea culpa, I didn't notice
a) that the mid/high absorber had fabric instead of plywood
b) that the high bass trap is using 1/8" plywood instead of 1/4" (see recalculation below)

1/8" italian light plywood is .21 psf
1/8" mahogany plywood is .33 psf
1/8" birch 3 ply plywood is .44 psf
1/8" Marine Meranti (Phillipine Mahogany) plywood is .59 psf <- I'm cheating !
1/8" marine douglass fir plywood is .71 psf
( http://www.westwindhardwood.com/wood_marine.html )

Deep Bass Trap: f = sqrt( 28900 / ( 0.78125 * 3.5 )) = 102 hz
Mid High Absorber: see absorption coefficients
High Bass Trap: f = sqrt( 28900 / ( 0.59 * 1.5 )) = 180 hz

Well, the calculated performances of these walls are:


reference: STC=41, OITC=29, tennekes=24.2, flat noise = 38.1,
with all traps: STC=47, OITC=30, tennekes=24.9, flat = 41.3
without mid/high absorber: STC=45, OITC=30, tennekes=24.5 , flat=40.4


Some simple analysis of these results:

a total of 135lbs was added with all these traps, equivalent to about 1 more layer of drywall on one side, which should give a 2-3 point/dB improvementin all of the measurements.

So for the mass addition, the triple leaf performs worse than keeping the double leaf.



The improvements at higher frequencies relate to another air cavity with very flexible panels and so forth, and aren't entirely unexpected.

Most surprising to me is the high freq effect of the mid/high absorber.


The low frequency effect of adding the traps does not seem to fall in line with the predicted resonance points of the traps. An effect is discernible, but it is not extreme.

The bulk of the weight of the traps is in the frames (which had 2x4's for the deep bass traps instead of 1x4s for installation practicality)



The reference wall was installed and tested. With no changed at all to that wall the traps were added on top of it and sealed. Weather stripping + additional caulk was used to seal them.

For the test without the mid/high absorber, it was simply removed and the test was re-run.


attached picture shows the middle of the installation process w/the lab guys doing the work so i can drink coffee.

_________________
Technical director, Green Glue Company
(Audio Alloy)
All posted content copyright Brian Ravnaas or Green Glue Company except as noted and may not be reproduced without permission.

ehtan and Z, you're very welcome & i'll try to get back to answer questions as best i can when i can.


For the oher folks, i hope if you have time to peek at the data you can comment.


In short, from an isolation standpoint this is (like all triple leaves) essentially a waste of materials. But this data supports the basic hypothesis that adding to create a triple leaf isn't necessarily bad. It is feasible that it could be bad in some particular situation, but in this case the results were not worse.

The same traps will be installed on hopefully at least one more wall before weeks end. I wouldn't expect to see any dramatic problems.


hope its interesting/helpful

_________________
Technical director, Green Glue Company
(Audio Alloy)
All posted content copyright Brian Ravnaas or Green Glue Company except as noted and may not be reproduced without permission.

Brian:

That is neat and unexpected.

(Dennis Erskine's design of of covering the 65% of the walls in absorption is to increase HF TL? , or did it just make your source room quieter )

the source room is factored in, as TL is shown as source minus reciever with various compensations and adjustments.

in this case, Ethans mid/high absorber has an air gap, and as USG demonstrated 20-some years ago by testing TL of their mineral fiber in various thicknesses and other absorbing materials, a material like this has some TL at higher freqs.

I don't know if the effect would be the same w/o the air gap or not (at high freqs). The effect may have occured because it does form another leaf, with cavity.

The results weren't what i'd expect, necessarily, but i wouldn't necessarily expect anything. A single wood stud wall, as this one, is a different beast than other walls as the resonant behavior relates to mechanical things and interactions between the sides and so forth, and instead of resonnce superimposed on each other, you probably simply create a new system with a different behavior.

as eric mentioned, the test needs to be repeated on a higher performance wall - a wall with a different basic set of behviors. a different construction.


However, as a purely-for-thought thought... if absorption on walls in the normal theater or studio manner did affect high freq TL (not just affecting source room) , then that would be an important thing to incorporate into testing of these type of assemblies, i guess.

_________________
Technical director, Green Glue Company
(Audio Alloy)
All posted content copyright Brian Ravnaas or Green Glue Company except as noted and may not be reproduced without permission.

Bob,

> That's where it resonanates, and probably also where the system absorbs the most from the room. <

I'd need to see some evidence of that! I'm quite certain those traps still absorb much more at their design frequency. Again, I've personally used this type of trap in four different rooms having standard sheet rock walls, and the improvement was far more significant than a teensy blip over a narrow range around 63 Hz!

> At 63hz, with the two resonant traps in place, the TL fell from 16.5dB down to 13.9 dB, which is a 2.6dB drop ... I don't know what the margin of error is <

Yes, the margin of error is key. But there is some evidence anyway given that two of the "with traps" tests show the same slightly reduced isolation.

--Ethan

_________________
The acoustic treatment experts
Buy my DVD

Ethan

Yep, absorption measurements would be nice. For a couple of reasons.

They're resonant traps, so whatever frequency they're actually resonanting relative to the room air is where they'll absorb most.

Below I have question marks about my reasoning on almost every sentence.

Assuming the absorption is exactly equal to the energy lost to the TL, then 2.6dB is like 0.45 absorption coefficient -- which is very good for a membrane trap, especially one with insulation in it.

Also, only 1/3 of the wall had the bass trap on it. Covering 100% of the wall with the bass trap would probably lower the TL at those frequencies. If the entire wall were covered it might be as much as 6dB TL drop, or a .75 absorption coefficient.

Bob,

> Below I have question marks about my reasoning on almost every sentence. <

Those are both excellent points.

--Ethan

_________________
The acoustic treatment experts
Buy my DVD

Bob,

> only 1/3 of the wall had the bass trap on it. Covering 100% of the wall with the bass trap would probably lower the TL at those frequencies. <

Thinking about this further, it seems just as likely to improve isolation even more. Certainly isolation will be even better at the higher range, and possibly better at lower frequencies too due simply to the additional mass. If the trap side of the wall were covered completely it would be heavier, and the slight reduction in loss at 63 Hz could be pushed down in frequency to where it may not even matter anymore.

--Ethan

_________________
The acoustic treatment experts
Buy my DVD