Diffusion

[Buckle-meister]

Hello all,

I was thinking a while ago about diffusion and what it does and wondered the following:

If an arbitary listening position in a room has a flat (or as good as flat) response, is there any need for diffusion? Will it make the slightest bit of difference?

 

[Savant]

Good diffusion will make improvements to the time response. Strong reflections get replaced with relatively weaker reflective energy spread out over time. In general, a "flat" amplitude (frequency) response before and after installing diffusion is not uncommon.

 

[Buckle-meister]

Good diffusion will make improvements to the time response. Strong reflections get replaced with relatively weaker reflective energy spread out over time. In general, a "flat" amplitude (frequency) response before and after installing diffusion is not uncommon.

I presume then from the last sentence above that even with a flat response, the addition of diffusion must therefore cause an audible difference (otherwise why would one bother). Are we saying that tonally, the sound will be the same (flat), but that any ringing present will tend to reduce as a consequence of there being "weaker reflective energy" from having been better dispersed by the diffusion?

 

[sploo]

Sounds to me that you could measure it by looking at the impulse response with ETF - the reflected impulses would be weaker and more spread out. Assuming I've got the right end of the stick of course .

From what I've read, diffusion is more an issue for larger spaces (though I'm noting this as a question as much as a statement).

 

[bpape]

Diffusion can be useful in smaller spaces provided you have the room for something effective below a couple thousand Hz and that you can sit far enough away from it to allow it to work. It takes some distance for the sound to spread out.

The reason in most cases (IMO) that diffusion doesn't change much in terms of frequency response is that to get a diffusor that is effective down below 300Hz (where most of your severe FR anomolies will happen), the diffusor would have to be huge. The size is directly related to the length of the waves you're trying to diffuse.

Also, some types of diffusors have not only a bottom limit, but also a top limit where they stop being effective.

Bryan

 

[JoeE SP9]

Adding diffusion to the front wall (behind my speakers) seemed to open up the soundstage and increase the impression of depth. It doesn't seem to change frequency response. This is with ESL's. I don't know if boxes will benefit as much or in the same way.

 

[Savant]

I presume then from the last sentence above that even with a flat response, the addition of diffusion must therefore cause an audible difference (otherwise why would one bother). Are we saying that tonally, the sound will be the same (flat), but that any ringing present will tend to reduce as a consequence of there being "weaker reflective energy" from having been better dispersed by the diffusion?

Mid and high frequency "ringing" might be reduced, but diffusion tends not to be an effective solution for modal problems. If the room is large enough and the diffusors are large enough, then modal problems can be addressed. But in home theaters, this tends not to be the case.

 

[Savant]

Diffusion can be useful in smaller spaces provided you have the room for something effective below a couple thousand Hz and that you can sit far enough away from it to allow it to work. It takes some distance for the sound to spread out.

I have found that this depends on the diffusor. E.g., a decent distance from a traditional QRD-type diffusor is usually warranted—I would usually suggest no closer than about 6-8 feet. But there are some diffusors where distance doesn't appear to matter as much.

I also find that the distance "rule" does not tend to apply when diffusors are used on the ceiling. Due to what I believe is a weakened human hearing localization for ceiling reflections—relative to wall reflections—I have used diffusors to great (sometimes astounding) effect on even 7-8 foot high ceilings; i.e., where the listener-to-diffusor distance was well below 6 feet. I prefer diffusion, rather than absorption (or reflection ), for ceilings in many home theater applications, though this is subjective and open to debate. It is my opinion that this sort of approach jives well with the general desire to bring the movie-going experience into the home; i.e., into a smaller space than would "normally" be used for viewing and listening.

 

[Buckle-meister]

Still confused I'm afraid

Mid and high frequency "ringing" might be reduced, but diffusion tends not to be an effective solution for modal problems.

Then I don't understand what you meant by...

Good diffusion will make improvements to the time response.

...as I thought you meant ringing. Could you elaborate please?

 

[bpape]

Excellent points Jeff.

In your opinion, would one of the 'non-dependent' types be a poly? I've found that they tend to be less dependent on distance. But then, they're not true diffusors in some people's minds since they don't scatter randomly but evenly.

Bryan

 

[Savant]

Bryan,

Depends on what you mean by "poly." Different people have different definitions. If you are talking about semicylinders or semispheres, I would say they might could work, but you have to watch out for the comb-filtering associated with those types of devices.

What I've heard more success with in small room applications are random and pseudo-random devices. "Random" devices tend to be custom made. Either strategic use of bookcases, etc.; or actual custom-made devices. One of the better sounding recording rooms I've come across was in a studio in Colorado where all the diffusors were custom-designed, random combinations of different sizes of wood strips. The best photos I could find on the net were here. The band is pictured in front of the diffusors in several photos. (To the best of my knowledge, the diffusor design pictured was the studio owner's concept and design.) (Also, the very bottom pic is an RPG product from a different room or studio, I think.) (Anyway, you get the picture... )

 

[bpape]

Jeff.

Yes - I was referring to the pseudo-arc type polys. I've also recently seen some other diffusors that combine the random strip type you pictured but with parts of them not so random and emulating an arc but with a series of flat surfaces.

As for the comb filtering, I guess it's a matter of where you want it. A flat wall in a small room will give you comb filtering too if parallel with another flat wall.

Bryan

 

[Buckle-meister]

Psst! Savant; post number 9 mate.

 

[Savant]

Robbie,

Well-implemented and -designed diffusion tends to improve the time response. For all intents and purposes, this is synonymous with a reduction in "ringing" in the contexts of mids and highs. Since "ringing" is often (mis?)used to describe low frequency phenomena in this forum, I thought I'd clarify. What part did you not understand? Did I confuse?

 

[Buckle-meister]

Well-implemented and -designed diffusion tends to improve the time response. For all intents and purposes, this is synonymous with a reduction in "ringing" in the contexts of mids and highs. Since "ringing" is often (mis?)used to describe low frequency phenomena in this forum, I thought I'd clarify. What part did you not understand? Did I confuse?

Ah, well, I must admit that I tend to think of ringing only in the lower frequencies. So mid and higher frequencies can ring too? Hmm. I figured that with them being absorbed so much more in air than low frequencies, that ringing wouldn't occur.

But I digress. What I was confused about was that you had written:

Mid and high frequency "ringing" might be reduced, but diffusion tends not to be an effective solution for modal problems...

To me, the above quote seems fairly skeptical of diffusion having much of an impact on mid and high frequences, and also low frequencies too since modal problems are mentioned. Thus I concluded that time-response must refer to something other than ringing and I wondered what that might be. See?

 

[Ethan Winer]

> So mid and higher frequencies can ring too? <

You bet! That's exactly what flutter echo is.

 

[Savant]

If we agree that "ringing" is a colloquialism for resonance, whether diffusion will mitigate "ringing" will depend on the proper selection and implementation of the diffusion to address the problem specifically. The time response is the behavior of sound over time. This could include "ringing." Hence the "might" in my previous comment.

The most common effect diffusion has on the time response is, as I mentioned earlier, a conversion of (e.g.) a strong specular reflection into relatively lower energy reflections disbursed over a wider range of time.

Continuing the above assumption regarding the use of "ringing," a flutter echo would not fall completely into this category since it is not strictly a resonance; it is a transient phenomenon. That said, diffusion can still help attenuate flutter echo problems since the time (transient) response is being addressed. There is often a "ringing" sound associated with flutter for wall-wall or floor-ceiling spacings less than about 25 feet. This should not be confused with resonance, either literally or physically. More discussion on flutter echoes.

 

[Buckle-meister]

Ok, thanks. Interesting link by the way.

 

[Ethan Winer]

Jeff,

> a flutter echo would not fall completely into this category since it is not strictly a resonance; it is a transient phenomenon. <

I recall you and I went around on this once before.

It seems to me anything that repeats at a regular rate based on a physical attribute implies resonance. In this case the regular rate is determined by the wall to wall spacing, so I don't see why flutter echo is any different from a room mode defined by the same spacing. With flutter echo the sound source is typically hand claps or another sound having more midrange content than bass content. But either way the repeat rate (fundamental frequency) is the same, and all that varies is the duty cycle of the source sound. In this context, duty cycle means a pulse wave that's relatively narrow compared to the fundamental frequency. Like this:

<img src="http://www.sfu.ca/sonic-studio/handbook/Graphics/Pulse.gif">

So why is this repeat rate different from a regular mode's repeat rate?

--Ethan

 

[Savant]

Jeff,

> a flutter echo would not fall completely into this category since it is not strictly a resonance; it is a transient phenomenon. <

I recall you and I went around on this once before.

Yeah...you'd think you'd have learned by now!

It seems to me anything that repeats at a regular rate based on a physical attribute implies resonance.

Yes. I didn't say it wasn't dependent on resonance. There's just more to a flutter echo than resonance. The dependence on normal modes is covered, albeit briefly, in the thread linked above. (There are also a couple references for the inquiring mind.) Note also the underlined portion of my previous comment, emphasized by me so as to avoid carrying on about this.

In this case the regular rate is determined by the wall to wall spacing, so I don't see why flutter echo is any different from a room mode defined by the same spacing.

If one were to focus strictly on a flutter echo's dependence on wall-wall or ceiling-floor spacing, then they would appear to be quite similar.

With flutter echo the sound source is typically hand claps or another sound having more midrange content than bass content.

Which begins to explain the many differences, yes?

So why is this repeat rate different from a regular mode's repeat rate?

It's not.