AVS Forum banner

"Early Reflections Are Not Beneficial" article by Ethan Winer

16K views 145 replies 34 participants last post by  mthomas47 
#1 · (Edited)
I'll put this out there without too much comment, but I am interested in what others posting here think of Ethan Winer's new article.

Early Reflections Are Not Beneficial

I've installed first reflection panels per Ethan's instructions and found this to be about the most beneficial thing I've done to improve sound quality in my room. The only improvement that's made more of a difference is adding a subwoofer. As I've got more into audio, I've actually been quite surprised to learn that some leading experts like Floyd Toole don't believe first reflections should necessarily be treated.
 
#2 ·
Have to agree here that treating early reflections is very important. I am lucky that my wife puts up with my 7.2 system in our Family Room. I would love to follow Ethan's advice, but if I start hanging panels in our room, I think it would be pushing things a little too far.
 
#3 ·
I agree with Ethan on this subject and have done so for a long time. I never agreed with Toole's research based upon my own experience and three of the best four systems I've ever heard have had highly controlled directivity or been in very large rooms where the first reflection return has been long.
 
  • Like
Reactions: snpanago
#13 ·
Mr. Winer may have his own commercial agenda (as do Messrs. Toole and Olive), but I have to agree with him when he states:

"Floyd's statements about early reflections defy my own personal experience, and the experience of almost every other audio engineer I know."



I think Mr. Toole's term at JBL tends to bias his sensibilities to conflate attributes of JBL Professional products when transplanted into the living room. Some of these attributes, notably wide power response, are not only of limited value in a home theater, they may be detrimental.

Maybe this is why I like my panels with their limited horizontal and vertical dispersion.
Maybe this is why the inexpensive Behringer 15" 2-way "PA" systems (which beam somewhat like a panel) are finding favor at AVS.

The debate is not really about if "early" reflections should be treated or not. They should. Everybody with their head properly screwed on agrees to that.
Actually....Mr. Toole has expressed the opinion that "some" early reflections are beneficial to the stereo "image," whatever that means. Certainly Amar Bose would agree (if he was still around).
 
#5 ·
The key is to eliminate hard reflective surfaces. I have no obvious treatments, and I have no early reflection issues. But my treatments are there, consisting of an acoustical tile ceiling, carpeted floors, pleated blinds on the windows, all upholstered furniture, book shelves filled with books, walls covered with thick textured wall paper, and an L shaped room. Plus my speakers are line arrays, which have highly controlled vertical directivity. There's more than one way to skin this particular cat.
 
#7 · (Edited)
I'll put this out there without too much comment, but I am interested in what others posting here think of Ethan Winer's new article.

Early Reflections Are Not Beneficial

I've installed first reflection panels per Ethan's instructions and found this to be about the most beneficial thing I've done to improve sound quality in my room. The only improvement that's made more of a difference is adding a subwoofer. As I've got more into audio, I've actually been quite surprised to learn that some leading experts like Floyd Toole don't believe first reflections should necessarily be treated.
You mean first reflection absorption panels, right? :)

But yes, philosophically, I agree with Ethan. In fact, I take the idea further than he does in that my panels reach deeper into the frequency range than his do via being thicker and larger.
 
#9 ·
In fact, I take the idea further than he does in that my panels reach deeper into the frequency range than his do via being thicker and larger.
Yes, I agree that thicker is always better! In any location. Early reflections are mainly about clarity at mid and high frequencies, but thicker panels add more bass trapping which is always useful.

--Ethan
 
#10 ·
The debate is not really about if "early" reflections should be treated or not. They should. Everybody with their head properly screwed on agrees to that. The debate that has been going on lately is about wether absorption, which Ethan has invested a lot of prestige in as well as commercially, is the best *treatment* for early reflections or not.

Stereo reproduction and phantom projections are illusions of the mind. It uses shortcomings in the hearing to create an illusion, just like an optical illusion is used to fool the eye.

Now, in many cases early reflections that reach the listener is detrimental to this illusion. At the same time later coming reflections help cover up many of the flaws of stereo reproduction, So, what is needed is a balance between attenuating early reflections while allowing later reflections to fill the room.

Only absorption will not do that. Absorption will kill the early reflections very effectively, indeed. But applying only absorption for early reflections will also kill the later, psychoacoustically beneficial reflections in most cases. Other treatments for early reflections, such as scattering or diffusing early reflections in combination with some absorbtion will attenuate the direct reflections reaching the listener enough, but still allow the later reflections to occur to complete the illusion to its fullest extent.

That's what the debate is about. :cool:
 
#14 · (Edited)
First off, thanks for linking the article. Second, Ethan continues to be an incredible and selfless help to innumerable fans of quality audio reproduction. I have no issues with the crux of his article, or his positions in general. He has helped me personally, for which I am very grateful. :cool:

However....

Some of his points about a movie soundtrack's music content equating to a 2ch music setup requirements are at issue I believe. The transient nature of speech and our perception of consonants for example requires a different approach to treatment. Our use of first reflection absorbers are not at issue here, but it is widely accepted that room decay times and reflection energy should be lower for a home theater environment to deal with speech inteligibility.

This being said, improvements made to help with speech clarity serve to better a music experience as well. They are just not as make or break for music because of music's less transient nature. I mention this with regards to the "dead room" dilemma.

I also think he should have had made a distinction between early and late reflections in listening rooms. Late reflections aid ambience whereas early reflections aid intimacy (creates a smaller soundfield impression). The ratio of early to late reflections determines clarity and can be calculated using ETC to quantify clarity index values. Early reflections are destructive with regards to clarity, late ones are good.

This is nitpicky stuff that most people (myself included) wouldn't have thought to consider without the urge to learn more started by guess who.... Ethan Winer. :cool:

P.S. Isn't it cool to be able to say your audio playback system is better than the 'authorities' on audio matters. Whats up with those setups? :p
 
#17 ·
First off, thanks for linking the article. Second, Ethan continues to be an incredible and selfless help to innumerable fans of quality audio reproduction. I have no issues with the crux of his article, or his positions in general. He has helped me personally, for which I am very grateful. :cool:
Thanks. :kiss:

it is widely accepted that room decay times and reflection energy should be lower for a home theater environment to deal with speech inteligibility.
I agree about speech intelligibility, but music clarity is important too. I haven't measured the decay times in my living room in a while, but I imagine it's in the 0.1 to 0.2 second range. So my point is more that stereo music doesn't suffer in a room that has home theater decay times.

As always, anyone near me in western Connecticut is welcome to visit my room, and report their impressions to the forum.

--Ethan
 
#20 ·
I'm still amazed at how much echo you can get out of a 99.8 absorbent space. :eek:

Listen at 1:02 with headphones.


It isn't until they actually close both doors at least 50% of the way, that the echo starts to approach a TRUE zero.

Given this and considering that none of us even have so much as a 60% absorbent space, it makes you wonder just how much reflections we still have in our listening spaces!

I have about 45% absorption, and that is probably considered on the high-side. Here is what REW says:






I've never heard a well set up diffuse space, so I couldn't tell you if I liked it better or not; but I like what I hear in what I've got.

-My 2 cents
 

Attachments

#22 ·
I'm still amazed at how much echo you can get out of a 99.8 absorbent space. :eek:

Listen at 1:02 with headphones.
https://www.youtube.com/watch?v=mXVGIb3bzHI

It isn't until they actually close both doors at least 50% of the way, that the echo starts to approach a TRUE zero.

Given this and considering that none of us even have so much as a 60% absorbent space, it makes you wonder just how much reflections we still have in our listening spaces!

I have about 45% absorption, and that is probably considered on the high-side. Here is what REW says
I've never heard a well set up diffuse space, so I couldn't tell you if I liked it better or not; but I like what I hear in what I've got.

-My 2 cents
If were talking about early reflections, none of those graphs you presented speak to that. Something like a spectrogram with the noted control settings tells the story much better IMO.
 

Attachments

#26 · (Edited)
If you know what to listen for, a handclap is revealing. Not to mention expedient.

You can do a lot without a computer, believe it or not. A 3-D color Waterfall plot, though interesting, does not make room modes any less severe, and does not tell you what precisely to do about them.

A visual inspection (in light of experience) can surmise a lot about that.
 
#28 ·
Sorry, but the hand clap IS useless IMO in the sense that you have no point of reference. That is, if you change something in the room whereby some part of the reverberation is changed just a few DB, you will never notice. Whats more, you wont be able to discern where in the bandwidth it happened. Get a mic and software and do it right.

Maybe in very large spaces (cathedral ?) where the reverb is measured in seconds, you could gain something from such a test. But in small rooms where the reverb time is less than 1/2 a second? Less than 1/4 second?

In this day and age where you can get a accurate measurement rig for less than $100, why even bother with a archaic means of evaluation?
 
#29 ·
Sorry, but the hand clap IS useless IMO in the sense that you have no point of reference. That is, if you change something in the room whereby some part of the reverberation is changed just a few DB, you will never notice. Whats more, you wont be able to discern where in the bandwidth it happened. Get a mic and software and do it right.
I am not trying to be argumentative, or dismissing computers as superfluous tools. I logged quite a few hours during my professional career on EASE.

I would say, though, that if "the reverberation is changed just a few dB, you will never notice" -- well, then, it would seem that the "change" in question was ineffective or not substantial enough to be worthwhile.
 
#31 ·
It was way easier for me to learn by using measurements and plots in my room. I was able to compare my results to others and acoustic reference texts. I found it pretty fast too.

If I know what "goal" to strive for in terms of a measured result, I can acheive it and be satisfied. Otherwise I find I second guess and fall back into bad habits of tweaking and puttering. It wastes time I find.

It also helps me correlate perceived sound with measured numbers. I can describe what I am hearing in objective terms and it means something to ohers if I need to ask for help. :)

I still handclap in a new room, or to test a spot I'm visiting. But not once I get down to it.
 
#32 ·
There are so many things wrong with this article... it's in the weeds in the first sentence.

This article is meant mainly as a rebuttal to those who believe that early reflections enhance sound quality in a typical home-sized listening room. (emphasis added)
Belief has no place in science.

You are free to accept or reject a researcher's results and conclusions, no one is asking you to believe anything. However if you choose to reject their results, you can only do so by presenting contrary data, or an alternate analysis supporting a different conclusion. I taught Astronomy at a local Christian college, which had its fair share of young-earth Creationists. There is no data supporting Creationist theories, so we had a short discussion.

I bring up Creationism because this piece shares a lot with Creationist publications. First, you trivialize the reality of what you're discussing, replace it with your preferred fabrication, then use a bunch of specious arguments to support the fabrication. Anything can be proven beyond reasonable doubt... if the reader's gullible.

One proponent of early reflections is loudspeaker expert Dr. Floyd Toole, whose latest book "Sound Reproduction" has been deservedly well received....
Having trivialized the research, you now trivialize the researcher.

Dr. Toole is an audio researcher. Calling him a loudspeaker expert ignores decades of his career. "His book" is neither "his" nor a "book" in the conventional sense. Most books with 17 pages of references are called "text books" and used for teaching... which is what this work does. And like most text books, the unique content is sparse. Toole is describing the research of others for the most part; only 17 of the references list him in the authors. This work is a synthesis of a lifetime of research, far more than a "book."

Shall I continue with the second paragraph? No, there are bigger issues with this piece.

Let's compare some rooms. Your room is sub-optimal; you have too many surfaces, and an odd number to boot. The arched ceiling is the problem, and your only option is to heavily treat to get rid of the odd room reflections. No one would intentionally start here, and those who have started with non-parallel walls had to fix that first, before they could get good sound.

Floyd's room is a far superior layout. The only skewed wall (ceiling) is tilted slightly, beneficial for smearing room modes. An opening to the rest of the house smears modes as well, splitting that wall's modal resonance energy into 1/2 wave and 1/4 wave resonaces.The room is well lit, with large windows (rear wall is in his book), so you know it's above grade and likely will leak bass. Those windows have draperies, and the floor is carpeted, so there's a lot more absorption than Ethan notes. In fact, that looks like leather upholstered furniture, so you know the room has enough high-end absorption (or he'd have cloth upholstery). He has passive acoustic features you don't point out (bookcase, lower front wall) and I've read where he used Sound Field Management to set up the room's bass response, using 4 subwoofers, sized, placed and phased for optimum listening area response. Given he's a cutting edge researcher at a large firm, I expect his room set-up was far beyond what professionals were capable of at that time.

So while it's puzzling that Dr. Toole's own listening room doesn't meet even minimal home theater standards, it seems relevant in assessing what he considers good sound.
Again, if the research is unassailable, tear down the researcher. Your statement is false on the initial presumption that Dr. Toole's room doesn't meet minimal standards, and your personal methods are revealed when you then extend this falsehood to criticism of his professional judgment. ALL BY EYE ALONE!!! Sight bias would appear to be your SOP, and as to your rampant confirmation bias...

The Facebook quote seals it; Facebook is your authority! Facebook? And you want to be taken seriously????

Next you tear into Audioholics room.

... the Audioholics home theater demo room below is another example of an inferior listening environment built by someone who advises others what to do.
THIS IS A PROFESSIONALLY TREATED ROOM. You failed to recognize professional acoustic treatments. Audioholics did exactly as they advise. At least the acoustic impact of their room's construction challenges are easily addressed.

There's one person you didn't mention, a well known audio professional whose internet presence includes several residences, all of which feature optimized loudspeaker installations. Linkwitz' site on surround sound shows a full set of room photos. I see a lot more in common with Dr. Toole's room that with yours. Walls are square, ceiling's flat, lots of glass and book cases... given his work with dipole speakers, one would not expect a reflection-free environment, and all he's trying to do is replicate the live listening experience in his home!

From a technical standpoint, your credentials pale against actual industry researchers, so what makes you think they're the ones that are wrong?

From a discourse standpoint, I expect disreputable behavior from politicians... shysters among the many reputable lawyers... and snake oil salesmen. You are what you do, and you should be ashamed of this piece. That you are not, tells the tale.

Have fun,
Frank
 
#34 · (Edited)
There are so many things wrong with this article... it's in the weeds in the first sentence.

This article is meant mainly as a rebuttal to those who believe that early reflections enhance sound quality in a typical home-sized listening room. (emphasis added)
Belief has no place in science.

You are free to accept or reject a researcher's results and conclusions, no one is asking you to believe anything. However if you choose to reject their results, you can only do so by presenting contrary data, or an alternate analysis supporting a different conclusion. I taught Astronomy at a local Christian college, which had its fair share of young-earth Creationists. There is no data supporting Creationist theories, so we had a short discussion.

I bring up Creationism because this piece shares a lot with Creationist publications. First, you trivialize the reality of what you're discussing, replace it with your preferred fabrication, then use a bunch of specious arguments to support the fabrication. Anything can be proven beyond reasonable doubt... if the reader's gullible.

One proponent of early reflections is loudspeaker expert Dr. Floyd Toole, whose latest book "Sound Reproduction" has been deservedly well received....
Having trivialized the research, you now trivialize the researcher.

Dr. Toole is an audio researcher. Calling him a loudspeaker expert ignores decades of his career. "His book" is neither "his" nor a "book" in the conventional sense. Most books with 17 pages of references are called "text books" and used for teaching... which is what this work does. And like most text books, the unique content is sparse. Toole is describing the research of others for the most part; only 17 of the references list him in the authors. This work is a synthesis of a lifetime of research, far more than a "book."
In science, the results of one study are relatively meaningless compared to scientific consensus. This relates to the fallacy of changing one's diet, for example, after every study comes out suggest eating this or that is the key to a healthy life (especially as these studies are reported in the media). Are you suggesting that there is anything resembling a legitimate scientific consensus regarding Toole's research on first reflections? If so, how have so many highly logical people (like Winer) been convinced otherwise?

Granted, I think Winer's case would be more convincing if he had a published, peer reviewed, bias-controlled study supporting every single claim he makes about listening rooms, but he's a highly rational person advocating an approach to audio that has a tremendous amount of support amongst knowledgable and skeptical engineers of many varieties. It's not like Winer's just making this stuff up for his own amusement. It seems to me that one could easily make the case that Toole is actually the one making an extraordinary claim. And I find it totally absurd to compare Winer's ideas to creationism. You much really be trying to get under his skin with that one!
 
#63 ·
Scientifically, you can measure the effect early reflections have on the measurement data. Whether you like that effect is the subjective part.
Maybe some of what happened in this thread was simply sighted listening bias from viewing photos of listening rooms over an internet connection.

Without defining what ones criteria is, the topic cant be explored properly.
When Frank claims his room needs no absorption and Ethan claims his room sounds great with lots of absorption I believe them both despite them being obviously quite different approaches. They are different rooms. If I were to move all their equipment into my room neither setup would work out well here. My room is different too.

Per Floyd, the shorter the delay the greater the improvement in speech intelligibility
Speech intelligibility suffers in a large concert hall. I do not hear anyone complaining about the sound of their favorite concert hall.

Speech intelligibility can also be improved by adding distortion. Should we do that?

If a pair of 901s were facing the listening area, the bulk of their drivers (8 out of 9) were facing the front wall, producing reflections from the same direction as the direct sound from the front soundstage.
With stereo toe-in, most of the drivers are sending their reflected sound toward the side wall, not straight back at the listener.

My impression of the 901 is that its full range drivers have a clean sound and the sound is definitely bigger than the room. In that respect I felt it was ahead of its time. It is also finicky to place, like a dipole. YMMV.

This debate is sort of ironic given the new Atmos ceiling bounce approach. We have come full circle. Wasn't the 301 and 501 etc. also doing a ceiling bounce?

I have come to believe lateral reflections DO sound good and I prefer them to not having them. I just prefer the later variety over the early one.
Localization to the rear is very poor and sound from that location takes on an 'inside the head' quality that splashes the whole sonic field. When people advised me to deaden my rear wall behind the MLP to eliminate early reflections off it, that helped a lot.

My frontrear dimension is the shorter one. If I face right so that my frontrear dimension aligns with the longer dimension of my room, the sound improves despite the asymmetry of the sound stage. Additional delay in front and rear seems to improve the intelligibility despite nothing at all changing except the orientation of my head-related transfer function.

That would seem to vindicate Toole in that when the short dimension is sideside as usually preferred, the side reflections are earlier than those arriving from front and rear. It appears that having two ears to the sides biases us toward side reflections.

ISD works fine in a very large space or with a finely tuned stereo setup like yours. For surround sound, especially like my 11.1 in tiny asymmetrical space, ISD is unworkable. Too many sources of sound, awful angles, obstructions, irregularities, impossible to steer those reflections.

So instead I let Neo:X steer the ambiance for me without relying on any reflections for spaciousness or intelligibility. Seems to work so far, though a work in progress.

i like to use wide channels to deal with recreating desirable reflections:
http://www.audyssey.com/blog/practical-guide-audyssey-dsx
I tried this approach. It overlays a 'boxy' club sound and the way they resorted to piggybacking it on Dolby did not seem to help it out any. It also gains up the overall level and I took that as a form of cheating via psychoacoustic marketing bias. Finally, in a lightly treated asymmetrical room like mine, the synthetic late reflections got buried in the room early reflections.

I moved on to Neo:X because its ambiance steering seems to preserve most of the natural character of the program without competing with the walls. I spent many hours comparing all the surround algorithms on my receiver and kept coming back to Neo even preferring it over native channels.

Yes, Frank has shown himself to be a hater with an agenda. It's pointless to even try to reason with people like that, so why bother.
Frank and Ethan's participation on the forums has been incredibly helpful to me. I wish you two could :kiss: and make up. Call it Christmas. Good tidings and joy, all that rot.

:)
 
#41 ·
Sure. Lots of examples of such studies on many topics.

Build a room that allows you to control reflections (there are such rooms, and concert halls, in existence -- or just put treatments behind acoustic fabric and take them up or down) and run a bunch of listening tests. See which they prefer. The problem is a lot of people are going to prefer the reflections since it will make the sound "fuller" even though it is inaccurate. It will take some training and listening so they know what to listen for and what it is doing to the image. A lot of people like the sound of vinyl even though it is inferior in most respects to digital sources. I like ice cream even though research proves it is bad for me. And so forth...

The impact of early reflections is pretty well-known and I have not personally been in a recording studio that does not dampen them in the mixing room. In my experience with my own systems and setting up numerous others controlling early reflections always yielded a better, more stable stereo image without the interference effects that make fixed instruments wander in position up and down the scale and such.

YMMV - Don
 
#90 · (Edited)
Sure. Lots of examples of such studies on many topics.

Build a room that allows you to control reflections (there are such rooms, and concert halls, in existence -- or just put treatments behind acoustic fabric and take them up or down) and run a bunch of listening tests. See which they prefer. The problem is a lot of people are going to prefer the reflections since it will make the sound "fuller" even though it is inaccurate. It will take some training and listening so they know what to listen for and what it is doing to the image. A lot of people like the sound of vinyl even though it is inferior in most respects to digital sources. I like ice cream even though research proves it is bad for me. And so forth...

Read Toole's book and his papers. Read what he actually writes. Stop blathering. He more than once notes the reported difference from research, between reflection preferences of professional audio engineers (mixing/mastering) and consumers, for example. He reports results as statistical preferences, not absolute predictions.

So far, of what I've read here, a lot of you people don't know what you're criticising. Shame on you.

As far as Toole vs. Winer, Toole has the better grasp of the body of research on his side. Anecdotal evidence is anecdotal. It's best used as a signpost for what might be worth further investigating, not as convincing evidence.
 
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top