Diffusive Panels In Stereophile Question

[mas]

1.) I think I get it. If you widen the wells enough to deal with the
longer waves, it still will have little to no effect if the depth of
the well is too shallow to be hit by the wave.

2.) The diffusor needs to protrude into the room to be hit by the wave.

2.) No. the diffusor does not have to protrude 'into' the room. The 'forward' limits of the diffusor can be flush with the wall and work great. But the pattern must become a characteristic of the incident reflective surface.

Typically, QRDs exhibit wells that are recessed. If you want something that is a 2D compliment and protrudes 'into' the room, then you might want to look at the 'skyline' style.

I must admit to having a bit of difficulty trying to get a handle on exactly where you are going with this... At first it seems like you might be trying to come up with a small diffusor that will be as effective as a large one - which misses the point, as it were; and now, well... I am not exactly sure where you are going!

And if it took you 12 posts to get the previous answer, you could have gotten it in one if you had simply asked upon what reasoning or method the QRDs were based. Hint: That's why they are called "quadratic residue diffusors".

1.) Huh? You are dealing with a complex mathematical relationship with many combinations and permutations. As far as trying to reduce this to a simple 1 size fits all English statement, I find the words a bit lacking. Especially as just when you think you have got it we can change the restraints placed upon another variable and change everything you think you have just said... Good luck with your attempt to distill the basis of the technology into a 25 word or less summary...but I am not sure what you are trying to gain that is not already available in a more precise form...

Is this what we are doing again? ...Another 12 questions to get what could be explained in a relatively straightforward manner if we can only figure out what information is actually desired? []

Note: The topic is a good deal more complex than has been (or will be) presented here. But as long as you stay near the middle of he road, the simple calculator can be a useful tool. But if you dare to travel too near to the edges of this proverbial road, things become a bit more squirrelly.

[Strabo]

What I was trying to do was get the lowest frequency diffusion out of a QRD without it protruding three feet into the room.

Given lower frequencies equals longer wavelengths I asked, only myself apparenty, what if it was wide instead of deeper? I was trying to make the wells wide enough to affect the lower frequencies. Somehow you understood this to mean I was trying to make it infinitely small. Apparently I didn't explain myself very well.

The question was, what if I built a QRD that was 20-some feet wide, six inches deep, and only 13 wells? Not eight sections of 3' square 13 well diffusers. Go big or go home. [] Would something like that work, and if so, how?

The answer seems to be that there is a relationship between the depth and width that needs to be followed. I can't change the width and expect it to work the way I want.

[PrestonTom]

You are on the right track about the need for depth (and that it will not "trade" with width).

A couple of comments. Most diffusers are effective until the depth is about 1 or 1/2 wavelength. With longer wavelengths (lower frequencies), their effectiveness is lost. This may not seem like a big constraint, but it is. The wavelength at 1000 Hz is about 13 inches, at 500 Hz it is 26 inches. The treatments would become huge if you try and extend their bandwidth lower.

The remarkable thing about the Schroeder diffusers is that they were now effective down another octave. That is, they were effective (with a comparable depth) down to 1/4 wavelength. This is big deal, since the treatments could remain a "manageable size" and still be effective over an extra octave. BTW, the recommended reading re: Schroeder, is a book on number theory and will be quite difficult (for most, including myself) to grasp.

Good Luck,

-Tom

[mas]

What I was trying to do was get the lowest frequency diffusion out of a QRD
without it protruding three feet into the room.

Given lower frequencies equals longer wavelengths I asked, only myself
apparenty, what if it was wide instead of deeper? I was trying to
make the wells wide enough to affect the lower frequencies. Somehow
you understood this to mean I was trying to make it infinitely small.
Apparently I didn't explain myself very well.

The question was, what if I built a QRD that was 20-some feet wide, six
inches deep, and only 13 wells? Not eight sections of 3' square 13 well
diffusers. Go big or go home. [] Would something like that
work, and if so, how?

The answer seems to be that there is a relationship between the depth and
width that needs to be followed. I can't change the width and expect
it to work the way I want.

I didn't misunderstand the question at all.

Nor did I ever think you wanted
to make it infinitely small. But you evidently misunderstood my extrapolated
exaggeration for exactly that point based upon your earlier proposed relationship.. The reasoning process is flawed. Well width and depth are not inversely
related in some simple relationship. And my extrapolation simply took your assumption to a logical extreme.

Your example proposed making the diffusor wider (widening the wells) allowing for the decrease in well depth. As you continue to widen the well widths and reduce the well depths, you will eventually reach the limit being an infinitely surface of wider and shallower wells until you reach an infinitely large surface well with 0 depth wells. And as was noted, aside from infinite surfaces being rather rare, this might be considered as an infinitely large flat wall![]

If we could make wide but shallow diffusors, almost all walls would qualify,
albeit tuned at a very low single frequency! ...Think the end of a room that is
14 feet wide and bordered by the adjacent walls! Instead what we have is a
standing wave in a tuned pipe exhibiting little wave decomposition!

I appreciate your interest, but you continue to ask if you can or how you
can make a diffuser with wider and shallower wells, assuming some inverse
relationship between width and depth.

There are some iterative options, but the short answer to your question is
no. That is not the relationship! And simply asking the same question over and
over won't change this. The nature of the system is
more complex than this.

From diffraction theory, we can determine a low frequency limit and depending
upon differing well depths that have the same magnitude, the differing depths
exhibit a different phase (time) and thus result in a differing propagation
time for the sound wave to travel down and up each well. So the polar
distribution of the scattering is determined by the choice of well
depths.

It is possible to construct Schroeder surfaces to maximize absorption or diffusion. And although fundamental relations exist, critical design differences result in different absorptive properties.

Without going into their use as absorbers or addressing quadratic residue number theory (as well as Kirchoff and Fresnel models), the summary of diffusive differences optimized for minimal absorption are:

• Well width is usually >2.5 cm to minimize boundary layer losses.
• Covering should not be utilized. If this is unavoidable, the use of a low flow resistivity material away from well entrances is advised.
• Number of different well depths, N... A larger number of wells usually results in a better diffusor.
• Depth sequence is chosen to maximize dispersion, which is best done using numerical optimization, with narrow period widths being avoided.
• Deepest well depth determines the low frequency limit of diffusion, except where period witdth is small.
• Construction mass elements should be well sealed with no slits and constructed of smooth surfaces.
  

Generally

If the wells of a diffuser are very shallow compared to the wavelength, then
the diffusers surface profile will not be seen by the soundwave. Likewise, if
the diffuser is very narrow it will also not be seen.

Diffusion is a result of basic diffraction theory. And the periodicity is a
function of the quadratic residue sequence based on a prime number, N. The nth
term in the sequence is given by n^2 modulo N
-> n^2 mod N.

The simple procedure for determining the well depths works by examining the resonant frequencies of the wells. To a first approximation, neglecting viscous boundary layer losses of the wells, each well is a quarter wave resonator with resonant frequencies, f, given by:

f = [ {(2m -1)c} / {4dn} ] where m= 1,2,3... and where dn is the depth of the nth well, and c is the speed of sound.

There is much more! The underpinnings of the theory is a bit more involved than may be apparent from the useful and simple calculator that does not feature any error checking. And it is not valid to make simple assumptions from a simplified calculator.

I can appreciate your wanting a shallow diffusor. But the fundamental problem is pursuing a presumed basic relationship that does not follow from your assumptions.

[]

[mas]

While I have issues with the notion that we can make quadratic diffusors that are effective at low frequencies by simply trying to make them wider and shallower, I can certainly understand the issue with a diffusor being large and deep.

And while I will admit that having a full range diffusor is indeed attractive and desireable, I think we need to stop for a second and acknowledge a few more practical aspects that can reduce the effectiveness for such a device in the small acoustical space.

Below 300 Hz, and also extending to the region between 300-500 Hz, we really aren't generally going to appreciate the benefits of diffusion as the wavelengths are simply too long. The effectiveness of diffusors also are impacted by the distance from the diffusor itself. At these low frequencies with excessively long wavelengths, the effective distance from the diffusor would be large - too large - in many smaller rooms.

So, just how far from a diffusor should a listener sit.from a diffusor? The distance from a listener to a diffusor can be determined by considering the scattered and total field. First consider the scattered field - just the reflections from the diffusor. A diffusor requires a certain time or distance to develop a wavefront. There is an analogy to loudspeakers that can be made here. A listener would not consider sitting 15 inches from a multi-way loudspeaker, because the listener would be in the near field of one of the drivers. At some distance from the loudspeaker, the individual drivers' wavefronts combine to form a coherent wavefront. The same holds true for scattering surfaces. These surfaces can also be thought of in terms of near and far field, although the situation is more complex than that of loudspeaker sources.

It is common to describe a scattered field by its spatial response; similar to the far field polar response of a speaker. But the spatial response of a diffusor is a much more difficult characteristic to measure. In the far field the response of the diffusor is invariant to the angle of incidence, the observer and the frequency within its operational bandwidth.

Unfortunately, in most critical listening rooms, it is usual for sources and listening positions to be in the near field rather than in the far field. Consequently, listeners should be positioned as far from the scattering surfaces as possible. Precedence indicates that it is best if the listener is at least three wavelengths away from diffusors. Since diffusors generally have a lower frequency limit of ~300 -500Hz, this means a minimum distance of ~10 feet. In some situations, this distance will have to be compromised.

A listener positioned near a multi-way speaker within the near field will hear anomalies associated with the individual drivers; and similar problems are encountered when a listener gets too close to a diffusor. Many phasing errors encountered in a room are due to the fact that listeners are positioned too close to a diffusor and they are hearing near field comb filtering effects. Furthermore, being positioned too close to the diffusive surface means that the temporal response is overly dominated by the surface closest to the ear, which means that the temporal dispersion generated by the diffusor is not heard. In other words, you are within the range where reflections from the surface of the diffusor remain specular (focused) in nature, and as a result, the direct and reflected sounds are more similar and comb filtering increases.

(Also, in the case of the QRD, fewer and wider wells increase the anomalies within the near field. Hence, increases in the prime seed and number of wells and narrower wells assists in minimizing, but does not eliminate this issue.)

This naturally leads us to a consideration of the total sound field...which is a topic for another discussion...

But returning to address the use of diffusion in a small acoustical space where the wavelengths are large compared to the room dimensions, for frequencies below 300 Hz (and even up to ~500 Hz), the room modes would best be addressed with bass traps, reducing the need for deep diffusors and further contributing to listening positions being located withing the near field of the diffusor.

[TheSoundBroker]

Back to the original question...

Those look like Auralex AudioTiles, a system that Russ Berger designed for Auralex. Interesting concept. I have some of the SpaceArray diffusors on order to give them a try.

[pauln]

An inexpensive way to make your own diffusors for experimenting is to buy some 55 gal plastic garbage containers. The side are already curved, just cut the sides vertically and remove the sections, use string to set the radius of curvature - don't punch holes, just put the strings around the whole thing to hold the curve under tension, and use the same strings to hang them. When you have all the sizes and locations just right, measure the base widths of the things while still strung and curved, then remove the strings and paint the plastic curves, and put them back up between thin board slats distanced the same as the base measurements. Total cost of a couple of big plactic waste containers, paint, and board slats is probably about $40?

[mas]

Russ' products are the space array (a hemifusor variant - a phase grating with binary 'filler') and the space array, (a simple phase grating).

Also, please be aware that the classic hemispherical scatterers, as they do not diffuse but merely scatter the reflection in the form of additional specular (focused) reflections, were effectively outdated with the advent of the Schroeder quadratic diffusors.

In fact, there is an interesting story regarding the introduction and testing of the quadratic diffusors back in (I want to say) 1989 in Nashville in the Oak Ridge Boys then newly designed studio with SynAudCon, Russ, Peter (D'Antonio), et. al. Russ had requested Don fly Peter and his new toys in for the seminar. And after measuring the effects of the two systems - the hemispherical back wall and the quadratics, the hemis were promptly removed and the quadratics installed. That was the effective end of hemi use in small acoustical spaces.

So if you chose to go with the hemis, please understand their fundamental limitations.

The most easily designed and user built diffusors are the 'simple' QRDs and the 'skyline' style of diffusor.

Also, if one gets a chance and is in the Nashville or Middle Tennessee area, stop by the MTSU campus in Murfreesboro and check out the Russ Berger designed RIM (Recording Industry Management) building studios. Its hard to believe that that campus is now larger than UTKnoxville's!

[gaspr]

Russ' products are the space array (a hemifusor

variant - a phase grating with binary 'filler') and the space array, (a

simple phase grating).

Small correction here...according to Auralex's

website, Russ Berger had a hand in designing three products for

them...Spacecoupler, Spacearray, and the Audiotile.

[mas]

Auralex can say whatever they want. Russ makes no claim to them as diffusion in the current pArt Science catalog.

Auralex also supplied Klipsch with a dead room (dead end dead end) and claim that it offers a diffuse environment that resembles NO room response that Russ has ever designed. In fact, it runs counter to every concept that Russ advocates. They also champion the use of foam for LF absorption. And the irony is that Russ gave us a pre-release hands on presentation and playtime with his launch of his first channel diffusion toys on February, 2007 at the Dallas TEF seminar hosted by Prestonwood Baptist Church. (It was my hope that just a bit of his experience was transfered by osmosis!)

Of course I must admit that we spent much of our time during the seminar divided between goofing around with Mark Seton and trying to get Russ' then new Dell laptop to work, which to this day still does not...Bottomline, Dell's drivers suck. OK, well maybe not all of our time...But I guess that is the difference between reading a website's marketing blurb and talking to the man himself. He certainly doesn't claim it nor does he spec it in installations for diffusion, unlike the use of the space array and the space coupler phase grating.which derived specifically from his use of the concept in his design of small custom 'studios in a closet" over the years.

Does anyone get their data from the technical work as opposed to marketing data sheets?

And considering the depth of the tile pictured (in the original thread photo), would one like to speculate as to its effective diffusive frequencies (assuming it to be purely diffusive instead of absorptive)? The irony is that Russ was perhaps the first advocate of using Schroeder's quadratic and binary series technology developed by RPG, and is in fact responsible for the first in situ testing of them at the SynAudCon seminar that he co-hosted with Don Davis - which resulted in them tossing the hemispherical scatterers used on the back walls of the then new state of the art Oak Ridge Boys studio and replacing them with QRDs.

Oh, but in my adventures I have discovered one thing! The Audiotiles are NOT diffusion! Duh! They are adhesive backed absorption! No wonder I could not find them listed under the Diffusion products!! So much for the affects of running about based upon blind presuppositions and assumptions rather than deductive observation. Doh!

As the AudiotileShockWave components are concentrated, absorption is
maximized and the areas of exposed hard wall surface provide for a reflective surface. As the units are spread out, there is less absorption and greater wall surface reflection, with the only appreciable diffusion is provided by the edge effects of the tiles that provide for a degree of high frequency scattering (as is the case with any raised panel absorption). Greater densities of the product provide for greater absorption. No wonder Russ provides no info on them as diffusors.

The attached picture is an application of the AudioTile at Lakewood Church. And as you will note, it is used in the recording studio recording room to control excessive reflections while not eliminating them. This is a different animal than in the control/mix room.

In any event, I will give Russ a holler and get the skinny on the tiles. Note: Nevermind. I would check if they were being positioned as effective diffusive elements, but I am not calling anyone to ask about the absorptive qualities of foam!