Diffuser: time & spatial scattering

Post and discuss acoustic topics, Studio design, construction, and soundproofing here

Diffuser: time & spatial scattering

Postby Yannick » Mon Sep 24, 2012 2:30 pm

Quote from Wikipedia: "It is also important that a diffuser spreads reflections in time as well as spatially. Diffusers can aid sound diffusion, but this is not why they are used in many cases; they are more often used to remove coloration and echoes."

Well, the above is not my opinion. However, the first proposed fact is a well-defended position on several acoustics forums (maybe less so here at studiotips).
Q: Why would any diffusing device spread reflections in time ? And even if this were possible, why would one need this ?
Q: a 34cm 1Khz wave hitting a PRD at 20 degrees will fall apart in how many waveforms ?
The same Question for a polycylindrical diffuser ...

I maybe ignorant, but IMO if a QRD/PRD were to return a 1Khz wave twice (at a different time point), this would only result in one new wavefront that is being attenuated, because they would be not exactly in phase.
Or is there a dark and secret theory that would suggest that our ear/brain can hear several superimposed waves of the same frequency, with the same direction ?
I suspect a flawed logic here, since no audio signal consists of just ONE period at a given frequency.

I have been trying to get my mind around this "time scattering" concept of a QRD/PRD - but I fail to see it. As I can see it, if it exists, it would be at high frequencies, corresponding to the width of the individual wells ?

The second phrase in the quote seems to contradict the first phrase. If diffusers are not used to diffuse, why would it be SO important that they would have temporal diffusion ?

?

I already read this thread
viewtopic.php?f=1&t=3194&start=20
But did not yet find the answer.

I get the argument of the widened impulse response, but wouldn't that be due to HF reflections only ? Or resonances ?
Anyway, again, why would one search for years for a loudspeaker with a near-perfect IR, and then smear that IR with diffusers that broaden the IR ?
Yannick
 
Posts: 130
Joined: Mon Nov 13, 2006 9:42 am
Location: Brussels

Re: Diffuser: time & spatial scattering

Postby bert stoltenborg » Mon Sep 24, 2012 3:11 pm

A good question doesn't need an answer....
:mrgreen:
If you view life with the knowledge that there are no problems, only opportunities, you are a marketing manager.......this is my personal philosophy
bert stoltenborg
 
Posts: 4572
Joined: Sun Apr 18, 2004 11:03 am
Location: Achterhood, Netherlands

Re: Diffuser: time & spatial scattering

Postby Yannick » Mon Sep 24, 2012 5:06 pm

I understand.
I could have sworn sound is a particle
:mrgreen: :mrgreen:

Anyway, maybe in a few years time, light will finally be a wave only.
Or all matter for that matter.

OT myself :mrgreen:
Yannick
 
Posts: 130
Joined: Mon Nov 13, 2006 9:42 am
Location: Brussels

Re: Diffuser: time & spatial scattering

Postby Bob » Tue Sep 25, 2012 12:09 am

Yannick wrote:Well, the above is not my opinion. However, the first proposed fact is a well-defended position on several acoustics forums (maybe less so here at studiotips).
Q1: Why would any diffusing device spread reflections in time ? And even if this were possible, why would one need this ?
Q2: a 34cm 1Khz wave hitting a PRD at 20 degrees will fall apart in how many waveforms ?
Q3: why would one search for years for a loudspeaker with a near-perfect IR, and then smear that IR with diffusers that broaden the IR


A1: Look for 'temporal' in Trevor Cox's book. "acoustic absorption and diffusers".

A2: Anyone remember my bathtub QRD photos?

A3: To overcome a room problem?
Last edited by Bob on Tue Sep 25, 2012 12:14 am, edited 1 time in total.
Regards
Bob Golds
"The only thing we regret in life is the love we failed to give."
"Be a rapturist -- the backward of a terrorist. Commit random acts of senseless kindness, whenever possible" - Jake Stonebender
Bob
 
Posts: 4360
Joined: Sun Feb 08, 2004 4:37 am
Location: Oakville, Ontario, Canada

Re: Diffuser: time & spatial scattering

Postby Bob » Tue Sep 25, 2012 12:10 am

Yannick wrote:I could have sworn sound is a particle

Wouldn't it have to move 880,991 times faster at sea level before sound becomes a particle?
Bob
 
Posts: 4360
Joined: Sun Feb 08, 2004 4:37 am
Location: Oakville, Ontario, Canada

Re: Diffuser: time & spatial scattering

Postby Yannick » Tue Sep 25, 2012 8:08 am

Bob wrote:
Yannick wrote:Well, the above is not my opinion. However, the first proposed fact is a well-defended position on several acoustics forums (maybe less so here at studiotips).
Q1: Why would any diffusing device spread reflections in time ? And even if this were possible, why would one need this ?
Q2: a 34cm 1Khz wave hitting a PRD at 20 degrees will fall apart in how many waveforms ?
Q3: why would one search for years for a loudspeaker with a near-perfect IR, and then smear that IR with diffusers that broaden the IR


A1: Look for 'temporal' in Trevor Cox's book. "acoustic absorption and diffusers".

A2: Anyone remember my bathtub QRD photos?

A3: To overcome a room problem?


Q&A1: I do not have the book. 160 dollar seems too much to look for an answer that maybe is not really there. I can build 10 big polys for 160 dollars :mrgreen:
I could be wrong, but my guess would be Mr. Cox is talking about temporal diffusion, not as a concept or end goal, but as one solution to a problem: diffusion.
If someone could show me a quote where Cox litterally talks about temporal diffusion as another concept than spatial diffusion, I would be much obliged.

Q&A2: no, I was not around then, can you post a link ? I am guessing, is this an experiment with just one period (one wave in the tub) instead of a continuous oscillation ?

Q&A3: how can you overcome a room problem by introducing time shifts in the order of 5-15 cm - that corresponds to 0.15-0.45 ms, our ear will never detect that as discrete reflections, even if they were not to recombine in just one attenuated wave reflection. If one could appreciate delays in that magnitude, one would be able to hear reflections off a small loudspeaker baffle. Or like large diaphragm microphones. Oh wait :bang Now I understand what someone could like about QRDs

Still, somebody should explain to me to my how several 1KHz waves, travelling in the same direction, can coexist. Sound is not about single events.

Unless it is becoming a particle :mrgreen:
Yannick
 
Posts: 130
Joined: Mon Nov 13, 2006 9:42 am
Location: Brussels

Re: Diffuser: time & spatial scattering

Postby bert stoltenborg » Tue Sep 25, 2012 9:52 am

Don't say it's a phase problem. The proper name is: Time Smear, and it's the most important property of a QRD!
If you view life with the knowledge that there are no problems, only opportunities, you are a marketing manager.......this is my personal philosophy
bert stoltenborg
 
Posts: 4572
Joined: Sun Apr 18, 2004 11:03 am
Location: Achterhood, Netherlands

Re: Diffuser: time & spatial scattering

Postby Yannick » Tue Sep 25, 2012 1:04 pm

OK, but even when you call it time smear, it sounds dirty to me.
I would not want that, not in a microphone, not in a loudspeaker, not in a Steinway D.

But am I correct in assuming that the proponents of Time Smear blatantly assume that these extra spikes in the IR are NOT coming from the same direction ?
Because IMESHO (inmyeversohumbleopinion) the only way to see this broader group of spikes as a good thing, is to think about them as scattered the hell around the place.
Yannick
 
Posts: 130
Joined: Mon Nov 13, 2006 9:42 am
Location: Brussels

Re: Diffuser: time & spatial scattering

Postby bert stoltenborg » Tue Sep 25, 2012 2:41 pm

As the IR you measure from a freestanding QRD or something similar is coming from that device, alll the spikes are indeed coming from virtual 1 direction.
It can be compared to a series of loudspeakers distributed in space like the bottoms of the wells of a QRD. And of course this gives time trouble.
A QRD is theoretically based on infinite large prime numbers.
And such a QRD, based on a very large prime, resembles, guess what, a poly :D
Cutting it down to a prime 7 design, it will not nearly have a lineair Fourier power response.

Did you see this? viewtopic.php?f=8&t=3831
If you view life with the knowledge that there are no problems, only opportunities, you are a marketing manager.......this is my personal philosophy
bert stoltenborg
 
Posts: 4572
Joined: Sun Apr 18, 2004 11:03 am
Location: Achterhood, Netherlands

Re: Diffuser: time & spatial scattering

Postby Yannick » Tue Sep 25, 2012 4:06 pm

Yes I did.
Nice that you (someone) did this.
That is of course how one should measure.

But of course, some will say this test is not valid, because you completely ignore temporal diffusion.
:twisted: :twisted: :twisted:

I have some theories about why these "some" actually can say in one phrase that a poly measures like a perfect diffuser but does not sound like one. MAybe they are mixing up residual comb filtering with non-diffusion. As a QRD absorbs quite a lot, comb filtering would obviously be much less than a poly, when listening quite close ? Or maybe some actually prefer components with a signature sound. This also goes for mic preamps, ad converters, dare I say it : cables 8O

I am currently building some extra polys in my studio, at the first side reflection points. I will be able to listen tomorrow.
And measure. Broadband absorbtion works, but it sounds strange to me.
Yannick
 
Posts: 130
Joined: Mon Nov 13, 2006 9:42 am
Location: Brussels

Re: Diffuser: time & spatial scattering

Postby Bob » Tue Sep 25, 2012 5:20 pm

Yannick wrote:Q&A1: I do not have the book. 160 dollar seems too much to look for an answer that maybe is not really there. I can build 10 big polys for 160 dollars :mrgreen:
I could be wrong, but my guess would be Mr. Cox is talking about temporal diffusion, not as a concept or end goal, but as one solution to a problem: diffusion.
If someone could show me a quote where Cox litterally talks about temporal diffusion as another concept than spatial diffusion, I would be much obliged.


To me,
spatial diffusion is a reflection that is a scattering of sound waves in many directions (as opposed to specular reflection where the incidence angle equals reflection angle). Generally, the less the absorption (zero) and the greater the scattering (even distribution in all directions), the more diffusive the diffuser.

temporal diffusion is where the sound energy is distributed in time. I think of it as hitting two 'walls' that are a different distance from the sound source, thus one reflection arrives measurably earlier than the other. The 'walls' can be quite small (inches?), as long as they are a few times larger than the wavelength. Big QRDs have temporal diffusion as well. Or the 'walls' can be actual walls, and the sound bouncing around the room boundaries and furniture, with parts of the original impulse arriving at different times due to 'the speed of sound vs difference in distance' (aka time) that the various reflections took to reach your ear.
A side effect of temporal diffusion is that, because of the necessity of smaller than room size objects doing the diffusion, and the relativity of effect to wavelength, that it's also a frequency domain with different frequencies reflecting differently (lower more specular, higher more time dependant, and close to wavelength more spatially), possibly altering the harmonic tonal qualities of the sound.

In this image, notice how the reflection extends through time in the last graph, as opposed to the specular -- a temporal diffusion.
Image

Yannick wrote:Q&A2: no, I was not around then, can you post a link ? I am guessing, is this an experiment with just one period (one wave in the tub) instead of a continuous oscillation ?

My silly bathtub experiment
http://www.bobgolds.com/Diffusion/batht ... _waves.avi
http://www.bobgolds.com/Diffusion/flat_diagonal.avi
http://www.bobgolds.com/Diffusion/flat_ ... icular.avi
http://www.bobgolds.com/Diffusion/pot_front.avi
http://www.bobgolds.com/Diffusion/pot_front2.avi
http://www.bobgolds.com/Diffusion/pot_side.avi
http://www.bobgolds.com/Diffusion/pot_side2.avi
http://www.bobgolds.com/Diffusion/QRD_double.avi
http://www.bobgolds.com/Diffusion/QRD_p ... ular_1.avi
http://www.bobgolds.com/Diffusion/QRD_p ... ular_2.avi
http://www.bobgolds.com/Diffusion/QRD_p ... ular_3.avi
I noticed with the QRDs, before the bathtub wall reflection became dominant, that the QRDs reflections didn't look like the original wave, and were more numerous in direction and time as well as lower in amplitude.

Yannick wrote:Q&A3: how can you overcome a room problem by introducing time shifts in the order of 5-15 cm - that corresponds to 0.15-0.45 ms, our ear will never detect that as discrete reflections, even if they were not to recombine in just one attenuated wave reflection.

The idea is that if you still want the energy in the room, but want the reflection to be significantly less loud than the direct sound to avoid an echo or perception of a wrong direction, or an unwanted comb filter effect, then a diffuser might be the way to go. I've only seen one person put a commercial diffuser (as opposed to an object you had to put in the room for some other purpose such as a chair or a lamp) in a first reflection point, and that was a world renouned home theatre designer.
That's the theory, expressed by the untrained, unaccredited, and unexperienced -- as to if this makes any practical sense, I have no idea.
Regards
Bob Golds
"The only thing we regret in life is the love we failed to give."
"Be a rapturist -- the backward of a terrorist. Commit random acts of senseless kindness, whenever possible" - Jake Stonebender
Bob
 
Posts: 4360
Joined: Sun Feb 08, 2004 4:37 am
Location: Oakville, Ontario, Canada

Re: Diffuser: time & spatial scattering

Postby bert stoltenborg » Tue Sep 25, 2012 6:42 pm

It almost makes me sentimental, to see these old experiments, Bob :D :D .

In fact you should have a bigger tub so the wave front gets almost flat when hitting the DUT's.
:D
If you view life with the knowledge that there are no problems, only opportunities, you are a marketing manager.......this is my personal philosophy
bert stoltenborg
 
Posts: 4572
Joined: Sun Apr 18, 2004 11:03 am
Location: Achterhood, Netherlands

Re: Diffuser: time & spatial scattering

Postby Yannick » Tue Sep 25, 2012 7:02 pm

smearing.jpg
smearing.jpg (54.12 KiB) Viewed 10860 times


Can anyone explain how you can get an additional 10 ms in the diffuse reflection peak, with a well depth of less than 1.7m ?
Would that not be a result of a very large DUT plus spatial diffusion INSTEAD of temporal diffusion ?
Or worse, ringing ?
What is the bandwidth of those diffuse reflections ?

It strikes me it is always an either/or situation when showing measurements of QRDs and polys (except for the document Bert linked to :) )
Also, Cox always compares to semicylinders. Why ?
Yannick
 
Posts: 130
Joined: Mon Nov 13, 2006 9:42 am
Location: Brussels

Re: Diffuser: time & spatial scattering

Postby Yannick » Tue Sep 25, 2012 7:06 pm

Bob, with all due respect for the use of your bathtub, this experiment shows some very non-flat waves almost inside the DUT. Also the ripples you create could more be compared to a 50 KHz wave hitting a QRD, instead of a 1-2KHz continuous tone.
I cannot see anything usefull in these clips.

But that could be just me :oops:
Yannick
 
Posts: 130
Joined: Mon Nov 13, 2006 9:42 am
Location: Brussels

Re: Diffuser: time & spatial scattering

Postby Scott R. Foster » Wed Sep 26, 2012 11:17 am

There... I fixed it.

Pic was wrong because QRD's don't reflect sound diffusely - you want a poly to do that job. As to time smear, a great way to spread the time of arrival of the reflection back to the listening position is to bounce the wave in a manner non-spectral [diffusely] so that while some small amount of energy comes back directly the vast majority of the energy gets sent bounding off the side walls / ceiling / furniture along other longer paths. 8O

Image
SRF
Scott R. Foster
 
Posts: 3854
Joined: Mon Feb 02, 2004 12:41 pm
Location: Jacksonville, FL USA

Re: Diffuser: time & spatial scattering

Postby bert stoltenborg » Thu Sep 27, 2012 9:10 am

Do Cox and Antonio know this, Scott? :twisted:
If you view life with the knowledge that there are no problems, only opportunities, you are a marketing manager.......this is my personal philosophy
bert stoltenborg
 
Posts: 4572
Joined: Sun Apr 18, 2004 11:03 am
Location: Achterhood, Netherlands

Re: Diffuser: time & spatial scattering

Postby Scott R. Foster » Thu Sep 27, 2012 1:21 pm

Apparently not...

:roll:

Maybe they should get in the bathtub together?
SRF
Scott R. Foster
 
Posts: 3854
Joined: Mon Feb 02, 2004 12:41 pm
Location: Jacksonville, FL USA

Re: Diffuser: time & spatial scattering

Postby Ido » Fri Sep 28, 2012 11:16 am

Bob,
never saw these tubby things,
very nice. way to go,
the idea of putting a diffuser in water and filming is yours?
I've seen many diffraction vids, never thought of putting actual DUT inside.
I think the idea itself is brilliant.
for same test I woulda taken the QRD apart, fixed it chaotically, and filmed again.

I still say before anything else, one needs to be able to define what is "ideal" indoor sound (on it's various types), and try to stretch that to what would be "ideal" diffusion.
signature
Ido
 
Posts: 2201
Joined: Mon Feb 09, 2004 7:47 pm
Location: Israel

Re: Diffuser: time & spatial scattering

Postby Scott R. Foster » Fri Sep 28, 2012 11:27 am

IMO defining ideal diffusion is trivial. Defining ideal sound is too subjective to be a useful concept. Kind of like an ideal breakfast.
SRF
Scott R. Foster
 
Posts: 3854
Joined: Mon Feb 02, 2004 12:41 pm
Location: Jacksonville, FL USA

Re: Diffuser: time & spatial scattering

Postby Ido » Fri Sep 28, 2012 12:42 pm

first off cousin Foster, I'm guessing most us ain't got no favorite breakfast,
cause who eats breakfast anyways?

what I mean in "ideal" is not that there is such a thing, but rather that the quest should be based on it, rather than on the "given" patchwork (as in, "hi, these are acoustics, there are 1-3 type diffusers, go jump a lake")
f.e., I do think "Ideal" characteristics should be searched in regard to the interaction that goes on in between all the surfaces of the envelope.
usually development goes on analyzing given situations (Beranek on halls? Microsoft on computers?). IMO that can be very limiting.

you really think ideal diffusion is trivial? well, I don't :mrgreen:,
even though I do eat 1.5 wholegrain poly every friday
signature
Ido
 
Posts: 2201
Joined: Mon Feb 09, 2004 7:47 pm
Location: Israel

Next

Return to Discussion

Who is online

Users browsing this forum: Majestic-12 [Bot] and 1 guest

cron