Low tech diffusion and bass trapping. Did it work?

[magnetophon]

Hi everybody,

I built this studio:


I has a lot of foam behind the wood as bass trapping.
Here is a plot:


You can also download 6 measurements in REW format here.
They where done with a BM6A and a KSM141.

So, did it work? (the acoustic treatment I mean)
I have not seen enough of these plots to tell if this one is any good.

This is what the room sounds like. Chaupiques and PEC have a lot of room sound in them.

Greetings,

Bart.

[lord_bunny]

That's such a wild looking room! I like it a lot.

[jdg]

wild looking!
Look at Studio Magnetophon

hows it sound?

[takman]

well now thats different......

is it random..or based on some calculations?

its looks like you put a lot of work into it, and you are asking/telling us:

So, did it work? (the acoustic treatment I mean)
I have not seen enough of these plots to tell if this one is any good.

that just seems odd....

[mobius.media]

Whether it "worked" or not, that looks amazing!!! Your current approach should take care of your floor-ceiling diffusion, but not side-side wall diffusion. You seem handy enough, I'd add a few traditional QRDs to the side walls for sure.

How on earth did you do all that? Do you have build photos?

It's incredible, but I'd be afraid of pieces falling down!

[PaulP]

Originally Posted by mobius.media

It's incredible, but I'd be afraid of pieces falling down!

Yeah, I wouldn't feel safe under that pile

It looks like an earthquake, half way through, and we're standing in the basement.

It's a pretty cool atmosphere, though...

Paul P

[Guillermo]

So that its something that you dont see every day!! Awesome room!!!

Its seems that the peak frequency its 180 Hz but its only 5 dB higher that 1 kHz, thats really great.

Your current approach should take care of your floor-ceiling diffusion, but not side-side wall diffusion. You seem handy enough, I'd add a few traditional QRDs to the side walls for sure.

That whould really help, in my opinion the QRDs should be designed to work from 500 Hz - up.

Congrats!!

[DanDan]

The Waterfalls are the best I have seen. Congrats.
DD

[magnetophon]

Thanks for the compliments everybody!

I did some new measurements today, since the laptop I used for the last ones started aliassing above 5k or so. It seems to have messed up the measurements also in the lows, cause now i'm getting some resonances that where not there before.


I also figured out how to make a more traditional graph in REW.

Originally Posted by jdg

hows it sound?

Quite dry, but not lifeless. But actually I'm much more interested in the slutz opinion on how it sounds!

Originally Posted by takman

is it random..or based on some calculations?

It's random.

Originally Posted by takman

that just seems odd....

I did put a lot of work in it (getting the steel machines that where there before out was a lot more work than putting the wood in though...).
I have been very much into audio ever since I was 11 years old (20 years ago) So I do know a thing or two about this stuff. (I was being modest... ) But I have not seen to many of these plots, so it's hard for me to put the values in context.
I know I should have no room resonances, and no hefty cancellation valleys. And of course a somewhat even decay throughout the spectrum.
So let's see:
-32Hz and 52Hz: could these be resonances of the BM6A? They don't change frequency when i use a different speaker or mic placement. (not sure if they should move though.)
-The stuff at 200,800 and 20k is buzz and noise from the laptop (I don't want to put java on my lean mean music machine).
-You can not see it here, but on most measurements I get a cancellation dip moving between 70 and 100Hz (and another one at double that freq), depending on where I put the mic and speaker. They are between 10 and 35 dB deep. How bad is that?
Could it be that the foam behind the wood is too closed in structure? I think at least there is enough of that stuff to do some hefty basstrapping!

Originally Posted by mobius.media

Your current approach should take care of your floor-ceiling diffusion, but not side-side wall diffusion. You seem handy enough, I'd add a few traditional QRDs to the side walls for sure.

There are no bare walls facing each other. The diffusers are quite shallow in some places though.

How on earth did you do all that? Do you have build photos?

It's incredible, but I'd be afraid of pieces falling down!

There are beams attached to the walls and ceiling with metal corners. The rest is screwed on to that with at least 4 screws per piece of wood. I wanted to make sure I could record any bass sound without a rattling diffuser!

Originally Posted by PaulP

Yeah, I wouldn't feel safe under that pile

No worry's, I can hang my full weight from each and every piece of wood!

Originally Posted by Guillermo

Its seems that the peak frequency its 180 Hz but its only 5 dB higher that 1 kHz, thats really great.

Unfortunately my new measurements look a lot more real world...
(and a lot less like Real World Studios)

That whould really help, in my opinion the QRDs should be designed to work from 500 Hz - up.

You mean 500Hz as a good compromise between size and usefulness?

Originally Posted by DanDan

The Waterfalls are the best I have seen.

What do you think of the new one?

Thanks again for the feedback everybody!

Greetings, Bart.

PS: I have more measurements, if anybody's interested.

[jdg]

those resonances are pretty high up, esp the 800hz.

is something in there vibrating? drum kit? something?

[magnetophon]

Originally Posted by jdg

those resonances are pretty high up, esp the 800hz.

is something in there vibrating? drum kit? something?

The stuff at 200,800 and 20k is buzz and noise from the laptop (I don't want to put java on my lean mean music machine).

There is also a kit in the room, but I don't think it's responsible for those. When you zoom out the waterfall in REW, you can see there is a steady state signal at those freqs.
You could also hear it trough the speakers when not measuring.

Bart.

[Weasel9992]

Originally Posted by magnetophon

Could it be that the foam behind the wood is too closed in structure? I think at least there is enough of that stuff to do some hefty basstrapping!

Hmmm...foam? What kind? How much? Foam isn't typically much good for bass trapping, even given the structure you've built.

Frank

[magnetophon]

Originally Posted by Weasel9992

Hmmm...foam? What kind? How much? Foam isn't typically much good for bass trapping, even given the structure you've built.

Frank

It's a mixture of all kinds. I got it from a mattress factory.
I did the "blow trough it test" and at least 50 % has an open structure, but there is also rubbery feeling foam, and sheets of foam with a plastic backing.
There is at least 20 cubic metre of foam and a couple of cubic meters of textile.
I made sure there is an uneven distribution, so there are a couple of corners where the material has a good thickness. (1 to 1.5m)

Bart.

[mobius.media]

Originally Posted by magnetophon

There are no bare walls facing each other. The diffusers are quite shallow in some places though.

Maybe but there are clearly bare stretches of wall (brick/stone) that you could easily add some 9"-12" deep QRDs to if you felt like it. It could only help further I should think.

Anyway, you seem to know what you're going for. Again, great job. thumbsup

[Weasel9992]

Originally Posted by magnetophon

It's a mixture of all kinds. I got it from a mattress factory.
I did the "blow trough it test" and at least 50 % has an open structure, but there is also rubbery feeling foam, and sheets of foam with a plastic backing.
There is at least 20 cubic metre of foam and a couple of cubic meters of textile.
I made sure there is an uneven distribution, so there are a couple of corners where the material has a good thickness. (1 to 1.5m)

Bart.

I mean, the room plots look pretty good, but that's also a fairly large space so I wouldn't expect to see small-room problems in it. The only real issue with the foam is that you really can't say *what* it is doing exactly. Without test data on the material all you can do is kinda guess.

Frank

[magnetophon]

Originally Posted by Weasel9992

I mean, the room plots look pretty good, but that's also a fairly large space so I wouldn't expect to see small-room problems in it.

The room is about 6 by 10 by 2.5 meters. (I'm guessing here!)
It's not very visible on this plot, but there are some 10 to 35dB nulls at 70 to 100Hz (and at double that), depending on where I put the speaker and mic. Time to score some 703 I guess.

Is it correct that any null or resonance caused by the room will change in frequency and severity when moving the mic or speaker, and that if they don't they are caused not by the room but by the speaker (or mic, theoretically). Or is that only true with nulls, and not resonances? I'm asking because of the 32 and 54 Hz resonances. I know the BM6A does some hefty electronic eq to extend the lows, and it's a bassrefex, so maybe in the plot you see the speaker decaying...
Did anyone else sweep a BM6A ?

Thanks,

Bart.

[SAC]

Is it correct that any null or resonance caused by the room will change in frequency and severity when moving the mic or speaker, and that if they don't they are caused not by the room but by the speaker (or mic, theoretically).

Not exactly.

The room dimensions will determine the distribution of the pressure wave and hence the modes and nulls. (as evidenced by what Floyd Toole's research with subwoofer placement tried to accomplish)

The actual response is a combination of the room's pressure wave and the direct signal of the speakers. Thus, when a speaker (subwoofer) is placed 'in' a null its direct signal will tend to mitigate the null a bit.


Just out of curiosity, can you post the unsmoothed frequency response? And do you have an ETC response?

Also, how are you making the measurement? And where is the mic positioned for this measurement (pardon me if this has been stated) Is this with one speaker driven, or more? Are you using time averaged pink noise? . I ask as above about 150Hz, this might have a significant impact and with the waterfall, it is difficult to tell as it obscures as much as it illuminates.

Also, to follow up on what Frank is asking, do you 'conveneint' access behind the facade to the actual absorption?

[Jeffrey Hedback]

What a combination of effort and imagination! I bet is sounds really good.

If you are truly trying to objectively qualify this space & given it's a tracking room:
- Get a dodecahedron speaker, place in center of room and take multiple captures of impulse response, moving test mic around in room. This will allow you to study energy (essentially reflections) over time. You can then fine tune treatments or find areas in space more suitable for certain applications. For example: an area with dense yet even decay of energy could be ideal for drums...an area with singularly identifiable strong reflections would not be great for vocals...but could be cool for some guitar sounds.
- for LF response: get quality sub, place in corner to excite all modes and again test in multiple locations (waterfall plots). You should move sub to more than one corner. You would then have more objective data that relates to how you'll use the room. Seems you have some vertical corner areas still available for trapping if needed.

I do have one lingering question: is the mattress foam fire rated? Anything less than Class B/self-extinguishing (or equivalent) would not be comforting and should petrify your local fire inspector. I hope that's a non-issue.

[magnetophon]

Originally Posted by SAC

Not exactly.

The room dimensions will determine the distribution of the pressure wave and hence the modes and nulls. (as evidenced by what Floyd Toole's research with subwoofer placement tried to accomplish)

The actual response is a combination of the room's pressure wave and the direct signal of the speakers. Thus, when a speaker (subwoofer) is placed 'in' a null its direct signal will tend to mitigate the null a bit.

I'm not sure I'm getting this.
Are you saying the nulls only change in severity and not in frequency when using a different place for the speaker?
Cause my nulls definitely change frequency, and that also lines up with what I think I know about them.

What about the resonances of the room, would they change freq when moving the speaker? I don't really know, but I think they would...

Originally Posted by SAC

Just out of curiosity, can you post the unsmoothed frequency response? And do you have an ETC response?

Here is a freq response.
What is an ETC?

Originally Posted by SAC

Also, how are you making the measurement? And where is the mic positioned for this measurement (pardon me if this has been stated) Is this with one speaker driven, or more? Are you using time averaged pink noise? .

The measurements where all done with one BM6A and a KSM141. The speaker and mic where in various positions, but all of them within a meter of the normal listening position. I was using sweeps.

Originally Posted by SAC

I ask as above about 150Hz, this might have a significant impact and with the waterfall, it is difficult to tell as it obscures as much as it illuminates.

Do you mean the resonances in the graph obscure the nulls by being in the way?

Originally Posted by SAC

Also, to follow up on what Frank is asking, do you 'conveneint' access behind the facade to the actual absorption?

What do you mean?
The only things Frank asked where:
"foam? What kind? How much?"
But also the question itself, without the part about Frank, is totally unclear to me.


Best,
Bart.

[magnetophon]

Originally Posted by Jeffrey Hedback

What a combination of effort and imagination! I bet is sounds really good.

Thanks a lot! Feel free to listen for yourself and tell me how it sounds! That way I can learn from it.

Originally Posted by Jeffrey Hedback

If you are truly trying to objectively qualify this space & given it's a tracking room:
- Get a dodecahedron speaker, place in center of room and take multiple captures of impulse response, moving test mic around in room. This will allow you to study energy (essentially reflections) over time. You can then fine tune treatments or find areas in space more suitable for certain applications. For example: an area with dense yet even decay of energy could be ideal for drums...an area with singularly identifiable strong reflections would not be great for vocals...but could be cool for some guitar sounds.

Thanks for the tip, seems like the way to go, but the only time I've even seen pictures of one of those is in a paper by Professor Angelo Farina. Is there a way to obtain one of those without building one?

Originally Posted by Jeffrey Hedback

- for LF response: get quality sub, place in corner to excite all modes and again test in multiple locations (waterfall plots). You should move sub to more than one corner. You would then have more objective data that relates to how you'll use the room. Seems you have some vertical corner areas still available for trapping if needed.

This is something I plan on doing some time. Would you help me interpret the results?

Originally Posted by Jeffrey Hedback

I do have one lingering question: is the mattress foam fire rated? Anything less than Class B/self-extinguishing (or equivalent) would not be comforting and should petrify your local fire inspector. I hope that's a non-issue.

It's a non smoking studio...










I do have decent fire extinguishers on the spot.


Best,
Bart.

[SAC]

First, I should say that I am amazed at the aesthetic treatment! Extremely creative!!!!

Regarding your questions...


A fully pressurized room will have a stable pressure structure that will vary with frequency. This is addressed right on the mark by Jeffrey’s suggestion regarding the corner loaded sub and the dodec… a delightfully abysmal sounding speaker complete with all of the comb filtering one could ever desire… But terrific for stimulating a space! (As far as a source, for example, B&K sells their Pyrite version for an absolutely obscene $5048USD!!! Hence the reason most build theirs. If you need info regarding this, PM me. Or, for a one time use, it would probably be cheaper and easier to rent the services of someone who owns one for an hour or two.)


Various positions and energy sources can stimulate the pressure response differently, as they do not completely pressurize the room. Thus the localization of the energy source within the room and relative to mic placement will have a varying effect on the perceived room response.


This is specifically referenced in the referral to Floyd Toole’s research regarding the placement of subwoofers to maximally mitigate the LF room response.


Time averaged pink noise can be more effective in measuring low frequency modal behavior as it more thoroughly pressurizes the room (and averages out any time limited variances).


The question regarding how many speakers will tell us allot – as using more than one speaker will cause additional comb filtering aside from what the room and single speaker will cause. Hence my question.


For instance, the apparent multiple nulls at ~200Hz, ~400 Hz & ~600 Hz and ~800Hz. The causes of which might very well be further indicated by examining the ETC response. And something else is happening blow ~30-40Hz...be it a modal or HVAC 'infrasonics' or something else... And above 1kHz, there is significant comb filtering. But I suspect this is caused by the superposition of the direct and virtual sourced reflected signals. But here the frequency response tells us only that it is occurring, it provides no detailed insight into the causes. That is what the ETC is for.


Yes, the CSD/waterfall displayed at an angle obscures much of the detail of the frequency response viewed alone, as well as being able to determine interval spacing of some the nulls. They both have strengths and weaknesses regarding what is displayed and how clearly and easily information may be gathered from them.

Additionally, my question referencing Frank’s query was in regards to the LF absorptive material and also to the nulls at 40Hz and 100 Hz and the peaks at 120Hz and 180Hz. More porous material might be more effective at mitigating some of these anomalies.

I am also curious regarding the ~900 Hz spike.

In order to understand what is happening ‘above’ the room modes (higher in frequency), the ETC response is the tool of choice.

The ETC response is a powerful tool for analyzing individual reflections within a room. While the frequency response/CSD is very useful for LF modal behavior, the ETC (envelope time curve), which is derived from the impulse response providing a total view of both the real and imaginary (potential and kinetic energy) of the system, directly shows the gain and arrival times of individual reflections within the room. These specular reflections are responsible for intelligibility, sound stage/ imaging, comb filtering and polar lobing, and additional spatial cues. It does not matter if the ETC is a log-squared or envelope ETC response.

Thus it provides an important complement to the CSD/waterfall and frequency response for behavior above the critical frequency, where the signal wavelengths become shorter than the room dimensions and cease to behave as waves and behave as rays/specular reflections.

In other words, what the frequency response/CSD-waterfall provides below the critical frequency for room modes, the ETC provides for reflections above that frequency.

Below is a sample ETC derived from an impulse response of a large acoustical space.








Oh, and fire ratings have little to do with non-smoking. As electrical shorts and insurance companies seldom care about the personal habits of the occupants in the event of an electrical fire... Nor do the shorts cooperate and limit themselves to occurring when someone is there to use the fire extinguisher. ;-) Check your local building codes and insurance carrier.

[DanDan]

Quote:
Originally Posted by DanDan
The Waterfalls are the best I have seen.

What do you think of the new one?

Not too bad but I can't really see into it properly. There are a few puzzling aspects.

Do those speakers do 32Hz? I suggest that you try a Room Mode Calculator to see if there are any credible modes around 32 Hz. www.bobgolds.com has one.
That should confirm the 53 Hz one as well plus indicate where to trap it.
Perhaps HVAC or traffic Rumble? Apart from that and the Laptop anomalies, it looks decent.
DD