diffusion vs. absorption

[krytikal1]

is it better to do absorption at the first reflection point, or diffusion? or is a matter of taste?



K

[Ethan Winer]

Absorption is better there for sure. At least in normal size rooms. Maybe in a room that's 30 feet wide diffusion would be okay there.

--Ethan

[krytikal1]

thanks ethan.

can you explain a little more detailed why. i believe you, i just have also seen some high end mastering rooms with diffusion at those positions. i was wondering why they decided that was better.

thanks again.

K

[avare]

Originally Posted by krytikal1

thanks ethan.

can you explain a little more detailed why. i believe you, i just have also seen some high end mastering rooms with diffusion at those positions. i was wondering why they decided that was better.

What space are you asking for? the clearly understood guidelines are reflections within 15 ms be at least 10dB down and the classic design standard is no reflections within the first 20 ms. If the the room you are asking for is in our house, I can almost guarantee that the room is not large enough to meet either criteria without absorption.

Please be more specific in your questions.

Andre

[bpape]

Agreed. In most residential sized rooms, absorbtion is the answer for early reflections. Diffusion MAY have it's place if it's a larger space but it's not at reflection points.

Bryan

[krytikal1]

cool. thanks guys. i get the idea. again, i had just seen some rooms that seemed to have some diffusion at early reflection points. thanks much for your reply.

K

[Ethan Winer]

Originally Posted by krytikal1

can you explain a little more detailed why.

Really simple - I tried it.

When RealTraps first came out with our combination diffusor / bass trap I brought two home and put them in front of the existing first-reflection absorbers on each side. The degradation in imaging and clarity was immediately obvious. My wife who is into audio also commented with 10 seconds of listening. It was about the same as having no treatment at all on the sides.

In my living room the side wall absorbers are about 6 feet away from the listening position. That's why I said that maybe in a much wider room diffusors there might be okay. But in a much wider room the reflections would no longer be early just due to the distance. And that larger distance also means the reflections will be much softer, and less damaging to begin with.

--Ethan

[krytikal1]

excellent. thanks ethan. makes sense.

K

[Glenn Kuras]

Originally Posted by Ethan Winer

Really simple - I tried it.

When RealTraps first came out with our combination diffusor / bass trap I brought two home and put them in front of the existing first-reflection absorbers on each side. The degradation in imaging and clarity was immediately obvious. My wife who is into audio also commented with 10 seconds of listening. It was about the same as having no treatment at all on the sides.

In my living room the side wall absorbers are about 6 feet away from the listening position. That's why I said that maybe in a much wider room diffusors there might be okay. But in a much wider room the reflections would no longer be early just due to the distance. And that larger distance also means the reflections will be much softer, and less damaging to begin with.

--Ethan

Ethan,
I would tend to agree, but did you happen to test it and if you did would you mind posting some graphs?

Glenn

[Ethan Winer]

I did not measure that. The degradation compared to using absorption was so obvious it didn't even occur to me to measure.

--Ethan

[Glenn Kuras]

Originally Posted by Ethan Winer

I did not measure that. The degradation compared to using absorption was so obvious it didn't even occur to me to measure.

--Ethan

I have heard people say the same thing about blind testing cables!!!!!!!!! To bad you did not test it. Would be nice to see the graph.

Glenn

[Ethan Winer]

Originally Posted by Glenn Kuras

I have heard people say the same thing about blind testing cables!!!!!!!!!

As I often explain to audiophile potential customers, the improvement from adding bass traps and other acoustic treatment is very obvious, versus the subtle changes people (think they) hear when swapping cables and other components. Blind tests are needed only when there's some question as to whether the change is real or not.

To bad you did not test it. Would be nice to see the graph.

Glenn, I have probably published a dozen acoustic tests in my various articles and web postings. Tell you what - you test this and post the graphs. Deal?

--Ethan

[gainstages]

Originally Posted by krytikal1

cool. thanks guys. i get the idea. again, i had just seen some rooms that seemed to have some diffusion at early reflection points. thanks much for your reply.

K

in some of these places the walls are "false" walls - they are fabric stretched over super absorbing walls, so a little diffusion helps to keep the room sounding a little more "real" like the average listener. Otherwise, there would be no reflections at all, which would not be a realistic listening environment.

typically though, the advice from Ethan and Glenn is correct

[DSD_Mastering]

I just attended the Grammy Studio Summit here in Seattle and Anthony Grimani spoke about room acoustics and stated that according to their calculations, 25% of the wall space should be absorptive and diffusion needs to be at your first reflection points.
I know my control room is not like this though, but my mastering room is.
Don't shoot the messenger


Regards,
Bruce

[jwl]

Originally Posted by Glenn Kuras

I have heard people say the same thing about blind testing cables!!!!!!!!! To bad you did not test it. Would be nice to see the graph.

Glenn

Which graph would you use for this test, Glenn? If we're talking about first-reflection points then frequency response curves wouldn't be as useful, since we're talking about the psychoacoustic perception of an even stereo image. In other words, I think this falls under psychoacoustics rather than acoustics, if that makes sense. Though admittedly I imagine the comb filtering would result in different-looking curves....

But I'm with Ethan, I look forward to seeing how you design and perform your test....

[Glenn Kuras]

Originally Posted by jwl

Which graph would you use for this test, Glenn? If we're talking about first-reflection points then frequency response curves wouldn't be as useful, since we're talking about the psychoacoustic perception of an even stereo image. In other words, I think this falls under psychoacoustics rather than acoustics, if that makes sense. Though admittedly I imagine the comb filtering would result in different-looking curves....

But I'm with Ethan, I look forward to seeing how you design and perform your test....

Yes the comb filtering would look different.

Glenn

[Glenn Kuras]

Originally Posted by Ethan Winer

Glenn, I have probably published a dozen acoustic tests in my various articles and web postings. Tell you what - you test this and post the graphs. Deal?

--Ethan

Na, some of us have to work for a living, plus I don't want to steal your fire. Not all of us get to work at home in there robe, YOU LUCKY DOG!!

Glenn

[Jason Jones]

Diffusion at early reflection points is a little complicated, but research has given some guidelines to follow. Here is what has been shown to matter:

1) The off-axis frequency response of your speakers.

People prefer reflections that have the same frequency response as the original sound. At the sidewall early reflection points you are usually quite a ways off axis. If your speakers aren't flat at the angle of your reflection points, then don't use diffusion.

2) What diffuser you are using.

Is its effective range low enough...how flat is its response...is it 1-D or 2-D.....etc.

3) How live or dead the rest of your room is.
4) The angle of the reflection point with respect to you and your speakers.

People generally prefer reflections that are coming from behind them, behind and to the side or directly to the side....many people find reflections from the front of the room to be problematic.

So diffusion can be used at first reflection points but, usually a combination of #1 and #4 makes it not work well in most situations.

Jason

[Ethan Winer]

Originally Posted by Glenn Kuras

Na, some of us have to work for a living, plus I don't want to steal your fire. Not all of us get to work at home in there robe, YOU LUCKY DOG!!

LOL, yes I admit it - I work at home all day long, usually in my, er, let's just say it's a comfy outfit. I tell people my home office is just like the Playboy Mansion, except without all the babes. But please don't tell Elli I said that!

--Ethan

[Ethan Winer]

Originally Posted by Jason Jones

People prefer reflections that have the same frequency response as the original sound.

I've heard that argument, but it makes no sense to me. If the reflections are as full-range (frequency) as the direct sound, that means the damaging comb filtering also extends over the full range of frequencies. Let's take this to an extreme example and pretend that the off-axis response of a loudspeaker is so terrible that none of the frequencies radiate off to the sides. In that case no treatment is needed at all because nothing reaches the side walls to reflect in the first place. As the off-axis response is improved incrementally, the reflections become more and more damaging. Now, I admit this is only arm-chair theorizing because I've never tested this myself in a room other than my own living room with my particular speakers.

People generally prefer reflections that are coming from behind them, behind and to the side or directly to the side....many people find reflections from the front of the room to be problematic.

Yes, I agree with this. Of course, the first reflection points on the side walls are always in front of the listener! At least for loudspeakers in a 2-channel setup as opposed to the surrounds in a multichannel setup.

--Ethan

[Ethan Winer]

Originally Posted by DSD_Mastering

Anthony Grimani spoke about room acoustics and stated that according to their calculations, 25% of the wall space should be absorptive and diffusion needs to be at your first reflection points.

I find this interesting because I consider Anthony one of the more knowledgeable proponents of room treatment. I've also been amused to see a growing denial, for lack of a better word, to the use of any treatment at reflection points. Floyd Toole of Harmon International fame published a paper saying that in tests people preferred the sound of bare walls at reflection points. The only explanation I can think of is the people were inexperienced listeners. The improvement in focusing when reflections are absorbed is undeniable to me, and likewise the improved ability to hear much smaller changes in EQ and panning and ambience effects when mixing.

--Ethan

[avare]

I've also been amused to see a growing denial, for lack of a better word, to the use of any treatment at reflection points. Floyd Toole of Harmon International fame published a paper saying that in tests people preferred the sound of bare walls at reflection points.

I have seen this trend. I am not saying it is not happening, just that I have not seen it.

The first thought comes to mind about Toole's paper (do you have a link or a title I research with) is how large were the rooms? 20 years ago the thinking was 20 to 30 ms before reflections. With the EBU/ITU/tak your pick standards using 15 ms, the rooms do have to be as large to criteria.

Andre

[Jason Jones]

Originally Posted by Ethan Winer

I've heard that argument, but it makes no sense to me. If the reflections are as full-range (frequency) as the direct sound, that means the damaging comb filtering also extends over the full range of frequencies. Let's take this to an extreme example and pretend that the off-axis response of a loudspeaker is so terrible that none of the frequencies radiate off to the sides. In that case no treatment is needed at all because nothing reaches the side walls to reflect in the first place. As the off-axis response is improved incrementally, the reflections become more and more damaging. Now, I admit this is only arm-chair theorizing because I've never tested this myself in a room other than my own living room with my particular speakers.

--Ethan

I understand your argument. But if that were the case then an anechoic chamber would be the perfect listening room as it has NO reflections. You would never under any circumstances want to use a diffuser, because that is just more reflections and more comb-filtering. However, I have never, ever heard anyone say that.

In studies - with lots of different people with different levels of experience and "blinded" conditions, and often in real life, people like to have some reflections - from the side-behind in particular.

Don't discount the effect of speakers with poor off-axis response. Especially if you have any toe-in, your first reflection points are going to be WAY-off-axis.

Jason

[Bierce85]

Originally Posted by Ethan Winer

The only explanation I can think of is the people were inexperienced listeners. The improvement in focusing when reflections are absorbed is undeniable to me, and likewise the improved ability to hear much smaller changes in EQ and panning and ambience effects when mixing.

--Ethan

agreed. someone who's just listening for fun might prefer the "wall of sound" created by a loudspeaker system in an untreated room where as someone making critical decisions needs to hear each speaker as clearly as possible.

[Bierce85]

Originally Posted by Jason Jones

I understand your argument. But if that were the case then an anechoic chamber would be the perfect listening room as it has NO reflections.

as far as accuracy is concerned I think this is correct but that would be a pretty lousy sounding room, wouldn't it?

[Jason Jones]

Originally Posted by Bierce85

agreed. someone who's just listening for fun might prefer the "wall of sound" created by a loudspeaker system in an untreated room where as someone making critical decisions needs to hear each speaker as clearly as pos