The No.1 Website for Pro Audio
 Search This Thread  Search This Forum  Search Reviews  Search Gear Database  Search Gear for sale  Search Gearslutz Go Advanced
Calculating slat / slot width for a binary diffuser Diffusion Products
Old 8th September 2011
  #31
Gear Addict
 

I think both types have their strength/weakness (depending on setting), qrd`s have much better temporal diffusion than a normal poly array with each poly having a width of +- 2 feet/60cm, on the other hand qrd`s are bound to absorb quite abit of sound due to edge diffraction and resonant behaviour.
Old 8th September 2011
  #32
Lives for gear
 
Jens Eklund's Avatar
Quote:
Originally Posted by avare View Post
You are disagreeing with Boggy then, as I am agreeing with him.

Andre
In which post does he refer to RPG?
Old 8th September 2011
  #33
Lives for gear
 
avare's Avatar
 

Quote:
Originally Posted by hsal View Post
I think both types have their strength/weakness (depending on setting), qrd`s have much better temporal diffusion than a normal poly array with each poly having a width of +- 2 feet/60cm, on the other hand qrd`s are bound to absorb quite abit of sound due to edge diffraction.
Well put. What is also ignored is that such diffusers are multiple 1/4 ? absorbers!

Andre
Old 8th September 2011
  #34
Lives for gear
 
boggy's Avatar
Quote:
Originally Posted by hsal View Post
...... on the other hand qrd`s are bound to absorb quite abit of sound due to edge diffraction and resonant behaviour.
Nevertheless, it's a good behavior, if we talking about control room treatment.

For example, wall/ceiling treatment need to be a fairly absorptive (average alpha~0.5) if we like to have AES/EBU/ITU recommendations about RT60 in the control room satisfied, especially with reflective floor.
Old 8th September 2011
  #35
Gear Addict
 

Yes,I agree in a control room setting
Old 8th September 2011
  #36
Lives for gear
 

Quote:
Originally Posted by Chris_Nicks View Post
sorry for the noob question but.. how do you measure ISD?
envelope time curve (ETC) response.

for hunting down early reflections, take this measurement one speaker at a time.
Old 8th September 2011
  #37
Gear Maniac
 

Quote:
Originally Posted by boggy View Post
If I can, I always place slats perpendicular to direction of sound radiation from loudspeakers. This means that orientation of slats at side walls need to be vertical, for example. But, I find that acousticians usually have horizontally oriented slats at side walls, and I suppose that his reasons for this is only aesthetic. In this case we have some liveliness in room but not diffusion.
I am trying to figure out why slat orientation makes a difference. But since I am a total zero in physics I don't find the answer. Can anyone explain?

BTW, most or all of the slat diffusers / absorbers I found thorough a Google search had vertically oriented slats, so I guess there must be a benefit over horizontal orientation.
Old 8th September 2011
  #38
Gear Addict
 

Vertical oriented slatts on sidewalls/front and backwalls=scattering on the horizontal plane (example:sound reflected from slatts behind you and source is in front, scattering is oriented left/right)
horizontal oriented slatts on front and backwalls/sidewalls=scattering towards ceiling/floor
Old 9th September 2011
  #39
Gear Maniac
 

thanks guys.. So there is an Ideal ISD. now how do you custom make a diffuser to increase or decrease ISD? :D
Old 9th September 2011
  #40
Lives for gear
 
Jens Eklund's Avatar
Quote:
Originally Posted by Chris_Nicks View Post
thanks guys.. So there is an Ideal ISD. now how do you custom make a diffuser to increase or decrease ISD? :D
The "ideal" ISD-gap of the control room is about 5ms longer than the ISD of the recording space.

EDIT:
If there’s no dedicated recording room connected to the control room, one can choose freely depending on that kind of work one does and what the room has to offer, but I would not go below about 10 ms.

EDIT 2:
It is the placement (and orientation if 1D) of diffusers in combination with other treatments (absorption or reflection) that controls your ISD-gap.
Old 9th September 2011
  #41
Gear Maniac
 

Quote:
The "ideal" ISD-gap of the control room is about 5ms longer than the ISD of the recording space.
ok jens. now question is how do I make a diffuser create 18-20ms of ISD? Is there a particular design, distance... position?
Old 9th September 2011
  #42
Lives for gear
 
Jens Eklund's Avatar
Quote:
Originally Posted by Chris_Nicks View Post
ok jens. now question is how do I make a diffuser create 18-20ms of ISD? Is there a particular design, distance... position?
See my second edit above.
Old 9th September 2011
  #43
Gear Maniac
 

ohh. thanks.. so more reflections=higher ISD, more absorbers=lower ISD?
Old 9th September 2011
  #44
Lives for gear
 

Quote:
Originally Posted by Chris_Nicks View Post
ohh. thanks.. so more reflections=higher ISD, more absorbers=lower ISD?
ISD = initial signal delay gap, which is the gap in time (milliseconds) from the original (direct) signal to the first loud specular (or diffused) energy that 'terminates' the ISD. remember that speed of sound is a constant in your room, so the reflection path distance traveled directly corresponds to the time it takes for that energy (path) to reach the listening position. the longer the path, the later it will arrive (in time) at the listening position

in the Energy Time Curve (ETC) below, the top is an untreated room (you see the specular reflections from the boundaries impeding the listening position)... and in the bottom graph, you see the 'early reflections' have been attenuated (creating the RFZ/ISD-gap) - and diffusers provide a decaying termination.




in the natural world, the ISD gap is determined by the distance of the boundaries. so in a small room (bedroom), the physical reflection path (total distance traveled) for the first-order-reflections are not very long - and as such, arrive very close in time to the original signal. now, take a much larger room (theater), and now the first-order-reflections take a much longer path to reflect and arrive back at the listening position - increasing the delay (in time) between the direct signal and those reflections. this is one reason why a large room 'sounds' large - is because there is a longer time-gap between the direct signal and reflections from boundaries. if i blind-folded you and took you into a bathroom and made you 'speak' - you can 'hear' how small the room you are in (the reflections arrive very quickly right after you speak). now if i blindfolded you and took you into a very large space and you spoke, it would take much longer for those reflections to reflect and arrive back at your ears - and your brain can perceive how 'large' of a space you are in based on this time delay. however, too long of a time delay and the reflection will be perceived as an 'echo' (not good).

since specular energy is modeled like light (ray-traced), we can manage the energy with absorption or redirection, to mold the room's specular response based on one's design requirements.

by attenuating the first-order reflections, you are essentially tricking your brain into perceiving it is in a larger space --- since the time-gap between the original signal and the next loud reflection is increased (reflections are delayed in time).



Quote:
Originally Posted by Chris_Nicks View Post
now question is how do I make a diffuser create 18-20ms of ISD? Is there a particular design, distance... position?
you are essentially creating an anechoic space for that duration of time (ISD-gap) - where you only want to hear the direct signal and nothing else. once this time-gap (ISD) is over, you then have the termination where the energy then impedes the listening position like normal. however, the way this termination (gain, decay, etc) impedes the listening position can trigger many pyscho-acoustic effects. diffusers are for the termination and decay, NOT for creating the ISD-gap. the ISD-gap is created by absorbing or redirecting specular reflections that would normally impede the listening position. you choose the length of your ISD-gap based on other factors as Jens has listed above.
Old 9th September 2011
  #45
Lives for gear
 
boggy's Avatar
Quote:
Originally Posted by Chris_Nicks View Post
thanks guys.. So there is an Ideal ISD. now how do you custom make a diffuser to increase or decrease ISD? :D
You can't control an "ISD gap" and make it "ideal" only with diffuser.

Also, you really don't need it, if you don't like to build your studio acoustics following LEDE or RFZ design principles, because you must use one of this design principles in entire room to obtain an ISD gap recommendation.

So, there are designs where you don't have an (ideal or not) ISD gap, like Blackbird studio c:



Source:
https://www.gearslutz.com/board/6834790-post49.html

As you can see, reflections starts from about 30dB lower level than direct sound, but there isn't any gap... or there isn't any louder reflection which is delayed....
Also, this is very similar to results of an anechoic room measurements, but this room isn't anechoic, or better, it don't sound anechoic at all.

With MyRoom Design principle we obtain similar results, but our initial reflection attenuation is 20 or 25dB, probably because much smaller rooms and closer diffusers...

Here are a results:




Also someone can get similar measurement results without any diffuser over absorption (similar to anechoic room), so, ISD measurements may not be that much important.... if you don't know exactly what you measure and what you look.

From my experience, especially in small rooms, much more important is symmetry and LF treatment, because no one can hear clearly effects of an ISD gap, effects of diffusion, or any other room behavior from 300Hz and above..... if room still has a "wild" LF response, which mask all other details.
Old 10th September 2011
  #46
Gear Maniac
 

Wow! Local host and boggy, cheers! thanks so much for the explanation guys. It was really informative :D
Old 10th September 2011
  #47
Lives for gear
hi boggy, the myroom outcome is inspiring!

in the white papaer it clearly outlines the reasons for the 1khz diffusors, but i was wondering what was used to define the bass absorption frequency cutoff?

cheers.
Old 10th September 2011
  #48
Lives for gear
 
boggy's Avatar
Hi gouge,

Quote:
Originally Posted by gouge View Post
hi boggy, the myroom outcome is inspiring!
Thanks! heh
Quote:
Originally Posted by gouge View Post
in the white papaer it clearly outlines the reasons for the 1khz diffusors, but i was wondering what was used to define the bass absorption frequency cutoff?

cheers.
Good question!

250Hz is mostly from my experience in small rooms. I define this cutoff before I begin to design acoustic treatment for two particular rooms that is described/analyzed in paper.

This frequency can be "attached" to Schroeder frequency (f=2000*(T60/V)^0.5) or to LEDE cutoff frequency (f= 3*c/a) where is the start of low frequency absorption for back wall for LEDE design principle (GS link)
But whatever we use for definition of this cutoff frequency, from my experience, we need ALL room modes fully dampened in small rooms from this particular frequency to the lowest frequencies... no matter if this is MyRoom, LEDE, RFZ, N-E etc...

cheers.
Old 12th September 2011
  #49
Gear Guru
Truth

+1 Well said Boggy.
DD
Old 12th September 2011
  #50
Gear Maniac
 

Let me quickly return to the initial question of slat width in binary "diffusers" as Boggy builds them.

I've just started reading the Master Handbook of Acoustics and it says: "Sound will be reflected from a rectangular panel if each of its two dimensions is five times the wavelength of sound".

But that would mean that the slats in the picture below that border with absorption on either side are completely useless, because they are much too thin to reflect sound waves in the audible range. Only when 4 or 5 of them are placed next to each other there will be an audible reflection.

It would also mean that if I want my binary diffuser to reflect frequencies from 1 KHz upwards I would need to use panels about 150cm / 5 ft wide.

I am confused...

Old 13th September 2011
  #51
Lives for gear
 
boggy's Avatar
Quote:
Originally Posted by bermudaben View Post
Let me quickly return to the initial question of slat width in binary "diffusers" as Boggy builds them.

I've just started reading the Master Handbook of Acoustics and it says: "Sound will be reflected from a rectangular panel if each of its two dimensions is five times the wavelength of sound".

But that would mean that the slats in the picture below that border with absorption on either side are completely useless, because they are much too thin to reflect sound waves in the audible range. Only when 4 or 5 of them are placed next to each other there will be an audible reflection.
In AES paper:
"Two Decades of Sound Diffusor Design and Development, Part 1: Applications and Design" by Peter D'Antonio and Trevor Cox, JAES Volume 46 Issue 11 pp. 955-976; November 1998
you can find information that strips (slats) for 1D amplitude grating diffuser needs to have a maximum width of half of wavelength of maximum working diffuser frequency.

Quote:
Originally Posted by bermudaben View Post
It would also mean that if I want my binary diffuser to reflect frequencies from 1 KHz upwards I would need to use panels about 150cm / 5 ft wide.

I am confused...

Old 13th September 2011
  #52
Gear Maniac
 

Quote:
strips (slats) for 1D amplitude grating diffuser needs to have a maximum width of half of wavelength of maximum working diffuser frequency
Sorry, but this formulation is a bit, well, diffuse.

What is the maximum working frequency? The highest frequency at which diffusion is happening?

So if I want my slats to act as diffusors up to 10 kHz the slats' width needs to be anything between zero and 1.65cm (half of 3.4cm which is the wavelength of 10 kHz)??

Can you give me an example so I understand your formulation? What is the diffusion working range of a 4cm slat? And what happens below and above this range?

Unfortunately, I can't download the linked publication without paying for it...
Old 14th September 2011
  #53
Lives for gear
 
boggy's Avatar
Quote:
Originally Posted by bermudaben View Post
Sorry, but this formulation is a bit, well, diffuse.
I understand.
Quote:
Originally Posted by bermudaben View Post
What is the maximum working frequency? The highest frequency at which diffusion is happening?
Yes.
Quote:
Originally Posted by bermudaben View Post
So if I want my slats to act as diffusors up to 10 kHz the slats' width needs to be anything between zero and 1.65cm (half of 3.4cm which is the wavelength of 10 kHz)??
Slats need to be about 1.7cm (slats with zero width doesn't have sense)
Also keep in mind that you have 10kHz only near axis of your tweeter... you don't really need diffusion above 4-6kHz in all walls, in control room.
10kHz may be useful in recording room... but this is also questionable...
Quote:
Originally Posted by bermudaben View Post
Can you give me an example so I understand your formulation?
Yes.
Quote:
Originally Posted by bermudaben View Post
What is the diffusion working range of a 4cm slat?
Up to about 4kHz
Quote:
Originally Posted by bermudaben View Post
And what happens below and above this range?
Diffusion works below this range. Above is mainly balanced absorption and reflection... for example, absorption coefficient is about 0.5 in areas where binary diffusers exists because binary diffusers has about 50% absorbing surface covered with relatively highly reflective slats.

As I already wrote somewhere, binary diffusers are a balance between highly reflective (possibly resonant!) and totally absorptive surface. Binary diffusers aren't a best diffusers that is invented until today. They are needed when we make room (too) dead because room modes, to bring back some liveliness in room together with some level of diffusion, nothing "spectacular".... There are much better diffusers than binary types, but they are more expensive.

Also 2D binary diffusers are better than 1D diffusers if you like to build something better: RPG Bad Panel -- Detailed Plans
Old 14th September 2011
  #54
jrp
Lives for gear
 
jrp's Avatar
 

I am afraid this is a bit diffuse for me as well...
I´ll tell you what i think. I expect some errors in my idea of this, so any correction is very welcome.

Isn´t it possible to see every slat like a little reflector?
Then anything smaller hitting a slat will be reflected, anything larger will bend around and be absorbed.
Anything smaller hitting the space between the slats will be absorbed as well.
Anything smaller hitting the corner of a slat will be thrown in some direction.
Right?

1000hz is 34cm.

Cover everything with 34cm slats and leave 1cm gap
- 97% above 1000hz will be reflected, 3% will be absorbed.
Right?

Cover everything with 34cm slats using a random sequence and no gap between neighboring slats
- you get a random sequence of reflectors with different bandwidth

Same as above but also a small gap between slats
- Random array of 1000hz reflectors, about 50% absobtion

Then why do the slats on those pictures look so slim and sexy?
Old 14th September 2011
  #55
jrp
Lives for gear
 
jrp's Avatar
 

You posted while i was writing...
Seems clear.
But how do you determine the lowest freq then?
BTW, it looks like you use little gaps between slats that are neighboring, is that correct? Sorry if this has already been talked through...
Old 14th September 2011
  #56
Lives for gear
 
boggy's Avatar
Quote:
Originally Posted by jrp View Post
You posted while i was writing...
Seems clear.
But how do you determine the lowest freq then?
Width of diffuser can be limit, but this is very questionable... there is Schroeder frequency below which we can't make diffusion because room volume is too small etc.... etc... low frequency diffusion isn't a best that binary diffuser can do.
As I wrote... they are better than only reflective or only absorptive surface, we can call it "balance" or "compromise" ... whatever... nothing spectacular, only better...
Quote:
Originally Posted by jrp View Post
BTW, it looks like you use little gaps between slats that are neighboring, is that correct? Sorry if this has already been talked through...

Yes, it's needed (in some situations) to make absorption coefficient 0.6, not 0.5 for example.
Old 14th September 2011
  #57
jrp
Lives for gear
 
jrp's Avatar
 

well, this sounds interesting. But i wasn´t thinking about bass diffusion. It´s the bass we want to absorb after all...
I would have thought the wider the slats the deeper will be the transistion to complete bass absorbtion.
Old 14th September 2011
  #58
Lives for gear
 
Jens Eklund's Avatar
I’m also having trouble believing that the ”diffuser” in the ceiling seen in the pic in post 50/51 is a “diffuser” since it will absorb a lot, if not most of the energy hitting it except the very high range but still a lot even in the highest range. Is there some measurement data on this kind of device one can have a look at?
Old 14th September 2011
  #59
Gear Maniac
 

Quote:
Quote:
Originally Posted by bermudaben
What is the diffusion working range of a 4cm slat?
Up to about 4kHz
Quote:
Originally Posted by bermudaben
And what happens below and above this range?
Diffusion works below this range. Above is mainly balanced absorption and reflection...
I am having trouble bringing in line your statement with what I read in F. Alton Everest's book.

You are saying a 4cm slat reflects sound waves above 4300 Hz. According to the Master Handbook of Acoustics, each dimension of a rectangular surface needs to be 5 times the wavelength of the sound wave in order to reflect it. In this case this would be 40cm!! I can hardly believe this.

In the example jrp is giving above, he is assuming that slat width needs to be equal or greater than wavelength in order to reflect the sound wave.
Old 14th September 2011
  #60
Lives for gear
 
Jens Eklund's Avatar
Quote:
Originally Posted by bermudaben View Post
I am having trouble bringing in line your statement with what I read in F. Alton Everest's book.

You are saying a 4cm slat reflects sound waves above 4300 Hz. According to the Master Handbook of Acoustics, each dimension of a rectangular surface needs to be 5 times the wavelength of the sound wave in order to reflect it. In this case this would be 40cm!! I can hardly believe this.

In the example jrp is giving above, he is assuming that slat width needs to be equal or greater than wavelength in order to reflect the sound wave.
There’s no definite answer here. Even if the with of a plate was infinite in length (and width), it would still not reflect completely (assuming plane wave) since there will always be some losses due to diffraction at the edges.
Post Reply

Welcome to the Gearslutz Pro Audio Community!

Registration benefits include:
  • The ability to reply to and create new discussions
  • Access to members-only giveaways & competitions
  • Interact with VIP industry experts in our guest Q&As
  • Access to members-only sub forum discussions
  • Access to members-only Chat Room
  • Get INSTANT ACCESS to the world's best private pro audio Classifieds for only USD $20/year
  • Promote your eBay auctions and Reverb.com listings for free
  • Remove this message!
You need an account to post a reply. Create a username and password below and an account will be created and your post entered.


 
 
Slide to join now Processing…
Thread Tools
Search this Thread
Search this Thread:

Advanced Search
Forum Jump
Forum Jump