DIY Diffusors to the Max

[Rolff]

[lucey]

Very nice thread. Beautiful work...

[G. E.]

Those interested in happy experimenting might be interested in some comments on "Quick Sound Field" application of diffusors based on the mobile diffusors shown in this thread.

[Shaman]

Great information - thanks for sharing !

[G. E.]

Originally Posted by DeadPoet

... When assembling I first put all pieces on a piece of wood in the right sequence ...

I took your suggestion and put the 714 pieces for the remaining three diffusors in three boxes in the right order. Worked out fine, now I can be sure that I had orderd the right quantities and the glueing will be faster than the first time.

[DeadPoet]

thumbsup

[SeniorityFedup]

wow just saw the thread nice mobile diffusors im gonna attempt to make some!

[G. E.]

Originally Posted by G. E.

Not as deep as the first design but adding symmetry to the back of the room:

Update for the very deep absorber in the back of the room: today I came up with a different absorber construction based on what can be read about (acoustic) impedance matching for maximum absorption. This kind of absorber is built so to have low impedance facing the air and high(er) impedance closer to the wall.

Since the impedance goes along with with flow resistivity one could basically achieve this goal with multi-layer porous absorbers of different flow resistivity as described by Trevor J. Cox and Peter D'Antonio*) or an appropriate geometry (less absorbing material on the "outside", more on the inner side).

In my example I use seven units of one or two strips (each strip 60cm x 6cm [~2' x 2.4"] profile Thermohanf -- with already low resistivity) simply hanging down from the ceiling to the floor. Total depth to the corner: 0,9m/1,5m [~3'/5']

*) T. J. Cox, P. D'Antonio, "Acoustic Absorbers and Diffusors: Theory, Design and Application" on Amazon

[awagner08]

Originally Posted by Balz

G.E. you're the best. Thanks so much for taking the time to detail how you made those diffusers. I plan to make a couple based on your design this summer.

Hey Balz, just wondering if you built your diffusers based on the good work of G.E. ?

[G. E.]

Originally Posted by G. E.

Update for the very deep absorber in the back of the room: today I came up with a different absorber construction based on what can be read about (acoustic) impedance matching for maximum absorption. This kind of absorber is built so to have low impedance facing the air and high(er) impedance closer to the wall. ...

Well... last week I actually built this kind of trap with 6 units of two strips (each strip 58cm x 5cm [~2' x 2"] profile Thermohanf) but it didn't measure any better than just putting those three rolls of thermohanf in the corner. Simply didn't work out the way I had expected...

[mikahanau]

Hi G.E
I think you missunderstood the principal aof these basstrap.The idea is to have impedance jumps, which means that you have to sandwitch layers with different flow resistance. The way these absorbers are working are NOT based on the porous principal, it is based on phaseshifts at the diffrent impedance surfaces.
The way acoustic hangers work is the same by the way.
In your case there is NO impedancejump expect the air/insulation jump, which is the same all the time. So you will have only one frequency which is more damped, beside the porous absorption

Sorry, i don`t know how to explain in english, but maybee you will find some links when you type "impendanzsprung" in google.

Or you can ask me at a place where i can speak german.

cheers, and good start into 2010
Mika

[Northward]

Mikahanau is right about the problems in your idea. You didn't implement it right.

A change in the transmission medium properties, in this case density will cause an impedance jump (so for ex air to another medium to yet another medium etc: so when you have an anistropic conditions)

Every time you have this impedance jump, energy will be dissipated / released. This particular phenomenon will happen only when the 'jump' happens.

These are very special traps and probably the trickiest ones to design properly. Finding the right sequence is very difficult. You need to know the materials you use very well.

[G. E.]

Originally Posted by mikahanau

... The idea is to have impedance jumps, which means that you have to sandwitch layers with different flow resistance. The way these absorbers are working are NOT based on the porous principal, it is based on phaseshifts at the diffrent impedance surfaces. ...

I totally understand that I didn't build a multi-layer absorber with impedance jumps -- I was after (continuous) impedance matching described by "... corrugating the front face of the porous absorber..." similar to wedges as an alternative approach.

To find out how good/bad it works in the back of the room I tested three absorber configurations in an otherwise pretty naked room:

• A: room with absorbant cloud
• B: A plus 3 rolls of Thermohanf with a total volume of 0.49 m^3 [~17'^3]
• C: A plus 6 units of two strips (each strip 58cm x 5cm [~2' x 2"] profile Thermohanf) with a total volume of 0.77 m^3 [~27'^3]
• D: A plus 1 + 2 + 3 strips (each strip 58cm x 5cm [~2' x 2"] profile Thermohanf) with a total volume of 0.38 m^3 [~13'^3]

[G. E.]

Originally Posted by G. E.

... I tested three absorber configurations ...

I think it's safe to say that configuration D is the most efficient. The strong ringing at ~87 Hz is the second room mode of the room width. Obviously it needs additional treatment.

[G. E.]

Originally Posted by Northward

Mikahanau is right about the problems in your idea. You didn't implement it right. ...

Thanks a lot to both of you for clarification -- I've never "noticed" this impedance jump principle in the literature before. I guess I'm back on track with the well-known corner bass-traps.

[G. E.]

Originally Posted by G. E.

... glueing will be faster than the first time.

Today the last of the four 17x14 diffusors finally got glued. It's fun to play with light and the acrylic glas...

[PaulP]

Originally Posted by G. E.

Today the last of the four 17x14 diffusors finally got glued. It's fun to play with light and the acrylic glas...

Impressive thumbsupthumbsup

You're going to have to coin a new term for your hybrid configuration.

How much did they end up weighing each ?

Paul P

[SAC]

Originally Posted by G. E.

I've never "noticed" this impedance jump principle in the literature before. I guess I'm back on track with the well-known corner bass-traps.

GE, regarding the tangent, the entire concept of absorption and diffusion is based upon the concept of complex impedance, just as it is more commonly encountered in electronics.

Each material has a characteristic acoustical-mechanical impedance. And each transition involves the acoustical analog of the maximum transfer theorem.

And using the acoustical analogy of the the maximum transfer function, a perfect impedance match between signals transported via mediums of different complex acoustical impedance results in all of he acoustical energy being absorbed.

Likewise, any differences in the impedance match between signals results in a dependent mix of partial absorption and partial reflection with respect to both time and frequency, the degree and nature of each being dependent upon the degree and manner of mismatch.

This perspective provides a much more comprehensive view of such behavior. But it one wants to delve into it in significant detail, one must be open to dealing with the language defining the precise relationships - math.

A pretty good and accessible description of the concept is presented in chapter one of
D'Antonio & Cox.

[G. E.]

Originally Posted by SAC

... one must be open to dealing with the language defining the precise relationships - math.

A pretty good and accessible description of the concept is presented in chapter one of D'Antonio & Cox.

I 've been making my way through this very book for some time now but my math has truly declined since my thesis on non-linear stability of tethered subsatellite systems for the European Space Agency in 1995. Hard to get back to this level -- I'll give my best...

[G. E.]

Originally Posted by G. E.

... The only "measurment" I can think of right now is to capture the temporal and frequency responses of the diffusers versus a flat higly reflective surface and visually compare them to some typical results published in Figure 2.7*) and Figure 2.8*).

*) T. J. Cox, P. D'Antonio, "Acoustic Absorbers and Diffusors: Theory, Design and Application" on Amazon

Following the typical set-up for in situ measurement (after Nocke) I prepared seven different 120cm x 60cm [~4' x 2'] sized surfaces for measurement:

• A: 8cm [~3"] of Pinta Acoustic Aborber Plano
• B: 10cm [~4"] of Thermohanf
• C: Vicoustic Wave Wood Absorber Panels
• D: PVC flooring
• E: standard BBC diffusors
• F: 17x14 2D-PRDs (speaker-mic axis perpendicular to the acrylic glas wells)
• G: 17x14 2D-PRDs (speaker-mic axis parallel to the acrylic glas wells)

[G. E.]

Originally Posted by G. E.

... I prepared seven different 120cm x 60cm [~4' x 2'] sized surfaces for measurement: ...

For the tests I used ARTAs Periodic noise generator (128k sequence length with Pink noise spectrum and 200 Hz cutoff), windowed ~7 ms (= 234 cm [~7']) to exclude unwanted late reflections from walls etc., and let the Dual-gate smoothed FR render the output.

{D}shows the full impact of a strong first reflection after ~1.2 ms resulting in massive comb filtering. {A} (best absorber, no reflection) and {G} (best diffusor, lot of reflections) show the smoothest response with different approaches. No surprises so far!?

Please remember that this is a very basic setup for qualitative measurements and the diffuse field at these small distances between speaker, diffusors, and mic is not fully developed.

Before the testing I expected configuration {F} to outperform {G} -- now I think {G} benefits from better "sidewards" diffusion that lets more reflections from the side reach the mic (which smoothes the FR). Yet I'm not sure if this translates to a real world application...

[SAC]

G.E.

Since you are busy with the process of measurements, you might want to see the 'latest' on the topic.

I only wish I could have been quicker to provide them in the hopes they might help you.

BTW. ARTA is the 'best buy' in measurement platforms. Heck, its worth the price for the manual alone! And many are anticipating a substantial price jump of perhaps $1k up to the competition when folks realize the competitive value of the product.

Oh, and ARTA 1.6.1 was just released for download...

All are pertinent to diffusion and deal with the higher than expected absorption coefficients in diffusors.


A Suggested Method to be Used to Measure the Scattering Coefficient of Full Scale Samples

Absorption Coefficents Pt 1: Is Area Enough?

Absorption Coefficents Pt 2: Is "Edge Effect" More Important than Expected?

[G. E.]

Originally Posted by SAC

... Since you are busy with the process of measurements, you might want to see the 'latest' on the topic. ...

Great links!

For those who want to see a totally unsmoothed view (same impulse responses as above) I used ARTAs DFT frequency response for the expected diffusion range of 500 Hz to 5 kHz.

[Millo 3.1]

Originally Posted by G. E.

The portable diffusor/absorber elements in the back of my room already got some use as advanced gobos in a lively recording-room (RT60 ~450ms) some time ago -- hm, very good results at mixdown time, my "sluttiest snare ever"...

These look so awesome! Great idea for my garage in the not-so-near future!

[807Recordings]

I don't want to sound to ******** here but I am trying to get a fuller understanding on all this.

What or would it work if on a diffusors instead of using square objects you use something a bit more pointed. I have seen in certain test chambers things a bit more pointed and I was thinking would this not kill off energy a lot faster. Or is that to miss the point?

[PaulP]

Originally Posted by 807Recordings

What or would it work if on a diffusors instead of using square objects you use something a bit more pointed. I have seen in certain test chambers things a bit more pointed and I was thinking would this not kill off energy a lot faster. Or is that to miss the point?

You don't want to kill off the energy, you just want to chop it up into little
pieces and spread them around. Instead of one big echo you get a whole
bunch of little echos arriving at different times. Like in a well-mannered
concert hall.

Paul P

[807Recordings]

Originally Posted by PaulP

You don't want to kill off the energy, you just want to chop it up into little
pieces and spread them around. Instead of one big echo you get a whole
bunch of little echos arriving at different times. Like in a well-mannered
concert hall.

Paul P

Gotcha!
But does that not take away from absolute neutrality in say like a studio because the broken up wave would still modulate the original source?
I suppose the analogy I often sort of looked at is water. Or at least for me to understand. If for example you have sharp 90 degree corners and the water waves come together there you get the various standing waves, etc. However if you have smooth rounded edges it can dissipate to almost nothing.

[PaulP]

Originally Posted by 807Recordings

But does that not take away from absolute neutrality in say like a studio because the broken up wave would still modulate the original source?

It depends what you want to do in your room. If what you care about is
the original signal and nothing but the original signal, then eliminate all
reflections. The room will be really dead but some people like that. If you
want a room with some life you have to have some sound bouncing around
the room and the idea then is to diffuse it enough so its effect is evenly
spread out.

Paul P

[G. E.]

Originally Posted by G. E.

... The strong ringing at ~87 Hz is the second room mode of the room width. Obviously it needs additional treatment.

After a little experimenting I found a treatment that really helped a lot with the 87Hz ringing. I've put up 3 rather massive 120cm x 60cm x 20cm [~4' x 2' x 8"] absorbers on both sides of the listening position about 8" away from the walls.

Downside: now the environment is rather dry with a broadband RT60 of 275ms (minimum of 220ms at 2kHz).

To get back some "room" I temporarily placed all of my BBC-diffusors in the back of the room right in front of the big bass trap. Configuration {A} = without BBC-diffusors, {B} = 4 BBCs, {C} = 8 BBCs.

[G. E.]

Originally Posted by G. E.

... I temporarily placed all of my BBC-diffusors in the back of the room right in front of the big bass trap. Configuration {A} = without BBC-diffusors, {B} = 4 BBCs, {C} = 8 BBCs.

Putting up the massive BBC-diffusors changed the frequency response in the bass-/modal-region. It looks like they act pretty much as a wall from ~50-200 Hz. Graphs show completely unsmoothed response.