Originally Posted by has7738
AES paper you refer to was presented in 1983 with at least one date on a graph of 1981.
That's right. Research dating back 30 years is being ignored day in, day out in this forum and on Ethan's web site. And you wonder why experts are infuriated when they read what is there? And I cringe when you say you want to send people there?
The test device used was a TEF 10, and while I have the greatest affection for it, and it did shed a lot of light, it was very limited in resolution by today's standards. But none of that causes a problem with the paper. The problem is your manner of presenting only selected conclusions from it to support your agenda.
There is no agenda. I am not trying to have a detailed acoustics argument in this thread and am focused on the high level concern of manufacturers using misinformation to sell products as I have been told is the issue. That detailed analysis as I mentioned is in my WSR article and is outlined in the long thread I linked to. Here is the specific post http://www.avsforum.com/t/1413173/does-sound-sounds-better-in-a-room-full-of-furniture-and-stuff-or-without/330#post_22193639
. I quoted those couple of lines to quickly show how wrong misconceptions are here. The article fully supports the statements I quoted. Ethan was present in that thread so my answer to him was just to refresh his memory about that conversation.
Mr. Clark did not do any ABX double blind testing of acoustic treatment for the paper.
I have not asked for "ABX" tests. Not sure why you think ABX tests should be used to detect levels of preference anyway. What I have asked for are controlled
tests as Clark and countless other researchers have performed. They change one and only one thing in the room and perform listening tests/measurements to detect if they correlate. Ethan is changing the discussion to a an empty room versus a room full of acoustic material. That is not a point of argument as empty rooms are not good acoustic environments. So no need to argue about the point that is already in agreement.
What we are asking is why he scares people into putting absorbers on first reflection points where research such as Clark’s and countless other Sanjay is posting and I have in the above thread show the perceptual effects can be very positive. I can’t think of anything worse than getting a customer to spend money on something that would *reduce* their enjoyment of music and make their rooms ugly to boot.
The real problem with the paper is that the data collected does not reflect human hearing well. Data is taken at a single point in space, a single measurement microphone, which presents a view of the system at one point at a time only.
The type of data he collected, is exactly the type of data collected by people here day in and day out. Both impulse and frequency response data. As to waterfall, there are massive pitfalls there. I am working on an article on that showing how another one of Ethan’s conclusions based on that kind of graph doesn’t hold either. I will post that on WBF forum if you are interested. Just quickly, in signal processing you can either have time or frequency resolution. If you make one better, it makes the other worse. A waterfall display unfortunately has both so it can easily show you misinformation in one axis or the other. It requires detailed knowledge of what you are doing to get the right data from it. So no it is not a matter of today’s tools having more resolution. You have to know the problem and limitations of the tools and people do not.
That aside, what you observe about the limitations of acoustic measurements is true in that what is measured doesn’t agree with what we hear. The right solution therefore does not rely on measurements above transition frequencies but rather, uses research of how we hear and listening tests to guide us to what we need to do. This is what happens when we design a theater from scratch anyway. There is nothing to measure before we build the darn thing.
Above transition frequencies psychoacoustics rules whether we like it or not. We don’t get to throw out the first part of that word and say this is an “acoustics” problem as you said. It is not simply so. It is a mix of both. My article on this topic came out during CEDIA. After the show one of the people at a major speaker company sent me a very kind and complimentary note saying he enjoyed reading how psychoacoustic and acoustic worlds collide this way and gave him food for thought. So please don’t convince yourself the messenger, either Clark or I, are wrong. There is nothing wrong here but our preconceived notions built up by manufacturers and others who have not even read the most basic research in this field, or have read them but refuse to accept them without any listening tests of their own to present.
Mr. Clark states in his conclusions that the single test mic hindered getting directional data relating to audibility.
And that statement is perfectly correct yet it is how people measure their rooms around here. Did they teach you in the THX class to use binaural measurements and include HRTF and a model of brain central function to analyze the room? I suspect not.
Human hearing is based on not only two points (ears), but directional and spacial hearing, integration over time, and integration of various system/room anomalies that are integrated within the ear-brain system. The conclusions are, therefore, correct given the measurement technique, but the technique does not correlate well with human hearing, which no doubt has given rise to the apparently startling conclusions you list above. The paper does not in any way present data that supports the conclusion that people prefer untreated over treated rooms. Mr. Clark makes this statement in his conclusions, "Stereo reproduction depends on room reflections. Very dead rooms leave audible comb filtering." Even that statement doesn't support that people like untreated rooms better than treated, because he references "very dead rooms". Proper room treatment would never leave a treated room "very dead".
No one has said the people should have "untreated rooms." That is a made up argument by Ethan. We are talking about specific subset of that topic, namely what you do with early/strong reflections. These reflections cause comb filtering and Clark set out to understand whether the severe anomalies we see on our graphs translate into what we hear. He found that the identical comb filter as shown in the type of measurements advocated here and by Ethan, can have wildly different audible effects. This leads to one and only one conclusion: you must understand how we hear to understand acoustics science. I explain this in the link above (a better version of which is in the WSR article), as do countless other experts such as Dr. Toole. Clark's paper and research in this area is one of many data points.
All of this said, it is good to see that you at least read the paper. You are on your first step toward learning the proper science here. The denial of what it says will go away over time
. That is what happened to me. I came from same school as Ethan. One of the first things that actually caught my eye was a quote from Clark's work that when we simulate exactly the same distortion that happens in the room via the first reflection electronically up stream of the speaker, the audible effect went from very positive to very negative. Right away I knew I was not in Kansas anymore
. I had to learn more. Hundreds of hours later reading research, talking to experts and participating in objective listening tests (double blind in my case), got me to understand that there is a world out there that is far more real than the one portrayed by hungry acoustic vendors and eager posters on forums. So I took the red pill and never looked back.
The option is yours to stay where you are and deny the research or continue down this fruitful path, take the class I mentioned and book references I gave you. Ethan talked about defining difference. This is a "defining" moment for any of you reading this thread! It really is.