Air Gap Discussion

[DanDan]

I would like to start a little discussion of the Air Gap.
A suspended ceiling with say 1" Fibre tiles and a 16" Air Gap performs remarkably well, very high absorption down as far as 50Hz.
Note the whole boundary is treated.
But consider a typical Cloud, say four 4"x4'x2'.
In this case the boundary is only partially treated, even if the traps are butted against each other.
So, what is the optimum airgap, butted together, and with gaps between them?
I have seen recommendations for 'as large as possible'.
This cannot be valid for a single panel or for isolated single panels, or even a small group of butts.
We typically recommend use of airgaps at least equal to the depth of fibre. But if there are only a few panels, with gaps, is this actually valid at all?

DD

[Soundman2020]

Good question, Dan! I wish I had a good answer, with lab test reports and research papers and things. But all I have is assumptions, and empirical results.

What I do know is that the gap extends the low frequency response. It's the distance from the front face of the insulation, to the wall, that matters: that's what defines how low (in frequency) the device will work. I also know that thicker panels at the same "wall-to-face" distance don't go lower, but they are more effective, and "smoother". And I also know that angling the panel, so that one side of it is further away from the ceiling than that other, has positive effects, since it goes lower still (deeper gap on one side).

What I don't know (numerically), is how all that stacks up for small panels vs. large panels (in terms of surface area), or a bunch of panels. Large panels, or more panels, will obviously have a greater overall effect, since they cover a larger percentage of the ceiling area. I would suspect that separating a bunch of smaller panels while still covering a large area, and leaving gaps between the panels, could have an additional benefit, because of the edge effect: the sides of the absorption would then also be exposed to the room air, and thus alpha would probably appear greater than unit, (compared to frontal area). I don't think the gaps between panels would be detrimental: I suspect they would be beneficial. And I doubt that those gaps between panels would have any effect on the lowest absorption frequency: that would still be given by the "face-to-wall" distance, regardless of whether or not the panels are separated from each other, or butted tightly.

That's my take, at least! Maybe Andre or Bert might have a better understanding of how that would work.


- Stuart -

[DanDan]

Great answer. I have seen Lab Reports and the Bobgolds treasure chest of data. But those samples are all 10 Metres or Yards Square, and contiguous. But I think the devil is in the detail. At the edge of the envelope, let's consider just one regular MiniTrap on a reasonable sized regular domestic boundary. Will it benefit from a 1:1 air gap or not? How far from the boundary before it starts to lose any effect. Andre will remember some BBC doc which shows traps distant from the boundaries are only 30% as effect as close. Something like that.
I have been going on the common view, i.e. yes, but I am now questioning it. Small gaps probably fine, perhaps even better than butted together, but how big a gap before the effect starts to fail.
DD

[bert stoltenborg]

I think Stuart explains it good.

I add a document Eric Desart made that might give more feeling about this stuff.

Edit: Ok, that doesn't work.
Why can't I post an PDF attachment?

[bert stoltenborg]

I would like to start a little discussion of the Air Gap.
A suspended ceiling with say 1" Fibre tiles and a 16" Air Gap performs remarkably well, very high absorption down as far as 50Hz.
Note the whole boundary is treated.
But consider a typical Cloud, say four 4"x4'x2'.
In this case the boundary is only partially treated, even if the traps are butted against each other.
So, what is the optimum airgap, butted together, and with gaps between them?
I have seen recommendations for 'as large as possible'.
This cannot be valid for a single panel or for isolated single panels, or even a small group of butts.
We typically recommend use of airgaps at least equal to the depth of fibre. But if there are only a few panels, with gaps, is this actually valid at all?

DD

I'm a simple loudspeaker guy.
If I want to know how a loudspeaker behaves I measure it.
In a room of dimensions under 10 x 10 x10 meter these measurements encounter a low frequency limit where the room boundaries introduce reflections that screw up your measurments.
A solution is to use the ground plane measurement method by Doug Button where the speaker and microphone are located flat on a reflective floor so that the direct sound emmited by the speaker and the floor refelction are in phase, as counts for the microphone, thus enlarging the room.
If you use this technique in a large room like a sports venue or a parking lot you can, using digital time windows, measure to the lowest frequencies we can perceive without reflections to bother about.
If you want to measure the first reflection of for example a ceiling couldn't you use a similar technique where you use a rigid wall of f.e. the sports venue to measure the response of it by projecting sound on it using a speaker and then add the treatment, measure again and subtract both measurements using above mentioned windowing postprocessing by removing the direct sound and concentrating on the impulse response of the reflection..
Eric Desart, Scott Foster and me did something similar when measuring diffusing devices, and so did Angelo Farina.
https://www.researchgate.net/publicatio ... _of_panels

It gives you solid time- and frequency domain information as you measure an impulse response that contains all the acoustical information and you don't have to grasp to artificial, non intuitive and course data like scattering- and absorption coefficients, and you don't need a reverberation room that won't give you usable data in the low frequency domain anyways.
You could also measure through an absorber to investigate how much of which frequencies is absorped as a variation on a Kundt tube measurement. As a single reflection can only add 3 dB to the direct sound a 3 dB transmission loss would be comparable to a absorption coefficient of 1.

But maybe this is all to easy .

[Soundman2020]

I think Stuart explains it good.

I add a document Eric Desart made that might give more feeling about this stuff.

Edit: Ok, that doesn't work.
Why can't I post an PDF attachment?

Not sure, Bert! You SHOULD be able to post PDF! I have PDF posting enabled for all users... In fact, the forum allows pretty much any document, even SKP and MDAT... unless the file is too big. What size is it (in kB) I can increase the allowed size, if necessary. Also, you could e-mail it to me, and I'll post it direct on the server.

- Stuart -

[bert stoltenborg]

I think Stuart explains it good.

I add a document Eric Desart made that might give more feeling about this stuff.

Edit: Ok, that doesn't work.
Why can't I post an PDF attachment?

Not sure, Bert! You SHOULD be able to post PDF! I have PDF posting enabled for all users... In fact, the forum allows pretty much any document, even SKP and MDAT... unless the file is too big. What size is it (in kB) I can increase the allowed size, if necessary. Also, you could e-mail it to me, and I'll post it direct on the server.

- Stuart -

Let's try it again.

[bert stoltenborg]

A random pic works, but the pdf (500k) doesn't.

[bert stoltenborg]

On the laptop it works.

[Avare]

I am at a loss to add anything usefull. Eric's Playing With Baffles is fantastic. Thanks Bert for the reminder!

[DanDan]

Interesting random pic bert! On another forum deedeeyah posted spectra of presumably large gigs. Pronounced HF roll off. I have looked at Erics Baffles and Ethans fibre tests. The reason I started this ball rolling here is that I am not at all convinced that isolated panels do benefit from airgaps. More precisely, I think there is a gradual failure as area is decreased. Ethan sent me some Lab tests years ago. Like bobgolds.com, the 1:1 gap greatly helped LF. But again, the ANSI sampling is 10 sq contiguous yards afaik? I will look thro Erics Baffles again. Andre did you sot simething specific in that? Ta all

[DanDan]

I’m on a phone here. Eric’s Baffles seem to be 65 mm with 35mm air. I am not seeing data with no airgap. Am I missing something? Ta all

[endorka]

I've never seen the improvement given by angling the baffles at 45 degrees in the room corners so clearly and accurately measured and depicted. That graph should be a sticky!

Cheers,
Jennifer

[DanDan]

Corner straddling has a unique effect. In a way the fibre is as both as close to both boundaries, and as far from them without springing a leak. Even a 2” gap collapses the boost. http://forum.studiotips.com/viewtopic.php?f=8&t=534

[Avare]

I have been thinking about this since it was posted. I see no difference. The sound wave path is through the absorbent material, the intentional gap, reflected off the wall/ceiling, back through the intentional gap and finally through the absorbent material again. The is no tuned cavity action to be contained by closing the gap.

Now if there is significant mass to the absorbent material and membrane absorption occurring...