It's a new experience being accused of spreading audiophile dogma!
I would not consider Earl an audiophle in the usual sense (often a sluir around here) but likes his music.
I did meet Earl at an AES (NYNY ‘99) presentation session.
dave
Movies have audio, and honestly I get more music through movies than otherwise. HT does have different requirements than a purely music sourced audio system.
I am reminded of something Mark Twain once said about another writer ... "That fellow sure knows a lot. Unfortunately, so much of what he knows just isn't true."
The crux of the problem, in a nutshell... Designing a valid test is a hard thing to do. I am married to a test engineer who spent her whole career designing and conducting (and later overseeing) electromagnetic testing. I got to hear all about it every night at dinner... There are thousands of ways to do a test wrong, and only a few ways to do it right.
The most dangerous kind of test error is an invalid test which, by coincidence, gives the correct result to the small set of validation test cases. The engineers and scientists then move forward believing the test was valid. They use the flawed data with confidence since the validation cases came back good. The common analogy we used is "a broken clock shows the correct time once a day"...
j.
So how could the various conflicting statements in this thread be resolved in a scientifically valid manner? Setting aside asking people that know the science or looking at the solutions the people that know the science adopt and instead addressing the topic in a scientific manner as if it was unknown. The scientific method (cut-and-paste from wikipedia setting aside caveats to kick things off):
- Define a question
- Gather information and resources (observe)
- Form an explanatory hypothesis
- Test the hypothesis by performing an experiment and collecting data in a reproducible manner
- Analyze the data
- Interpret the data and draw conclusions that serve as a starting point for a new hypothesis
- Publish results
- Retest (frequently done by other scientists)
Last edited:
Part of the problem here is that someone thinks they have already followed the scientific method.
And here is the publication: https://ethanwiner.com/speaker_isolation.htm
The problem with it is that its wrong.
And here is the publication: https://ethanwiner.com/speaker_isolation.htm
The problem with it is that its wrong.
@Markw4
This is so typical of internet arguments: You say my tests are wrong, but you don't say why they're wrong or what would be the right way to test speaker isolation. So your comment is less than useless. It's just arguing with no substance. Look guys, I tested a typical professional monitor with and without isolation. I accounted for speaker height, and measured every aspect of the sound in the room with a professional microphone and the REW software. There was no meaningful difference with isolation or no isolation. This has gone on now for a dozen pages, and so far not one person has explained what aspect of sound can be heard that won't show up in the REW data. I've asked for that more than once. You guys seem more interested in arguing, and patting yourselves on the back, than actually discussing the issue at hand. One person even claimed that his sighted listening tests are more valid than measuring. Sheesh! Look, maybe I'm wrong. Or maybe Earl is wrong. Argument From Authority is a logical fallacy because sometimes even experts are wrong. But so far I'm the only person who has actually done a test and presented the results! As they say, empirical evidence beats theory every time. Where's your "scientific method" evidence?
This is so typical of internet arguments: You say my tests are wrong, but you don't say why they're wrong or what would be the right way to test speaker isolation. So your comment is less than useless. It's just arguing with no substance. Look guys, I tested a typical professional monitor with and without isolation. I accounted for speaker height, and measured every aspect of the sound in the room with a professional microphone and the REW software. There was no meaningful difference with isolation or no isolation. This has gone on now for a dozen pages, and so far not one person has explained what aspect of sound can be heard that won't show up in the REW data. I've asked for that more than once. You guys seem more interested in arguing, and patting yourselves on the back, than actually discussing the issue at hand. One person even claimed that his sighted listening tests are more valid than measuring. Sheesh! Look, maybe I'm wrong. Or maybe Earl is wrong. Argument From Authority is a logical fallacy because sometimes even experts are wrong. But so far I'm the only person who has actually done a test and presented the results! As they say, empirical evidence beats theory every time. Where's your "scientific method" evidence?
Blocks your path: https://aes2.org/publications/elibrary-page/?id=20942
Since you refuse to reveal a critical element of the protocol by defining 'competent' loudspeakers, would you at least clarify if a 'competent' loudspeaker was used in your trial?
Since you refuse to reveal a critical element of the protocol by defining 'competent' loudspeakers, would you at least clarify if a 'competent' loudspeaker was used in your trial?
@Ethan Winer - Not so fast please. I gave an example involving ITD earlier in the thread. @gedlee said I made a very strong case, which I believe is correct.
What you did in your published study was a very different type of test from how some isolator pads were first used for mixing in recording studios with stereo nearfield monitors. I was very specific about NS-10m nearfields and Recoil Pads (which are made of more than just foam). I mentioned ITD thinking you would already know how it is different from comb filtering. gedlee tried to make it clear too, but you didn't seem to want to understand the difference.
IMHO, this dovetails back to a post by another member at #227. The point is well taken by me. I spent many years as an engineer in some very well known universities, sometimes as non-teaching research faculty, other times as an engineer and engineering manager. I get what post #227 is talking about.
If you would like to pick up from earlier in the discussion and understand how isolators could work in some situations, I'm sure many readers following this thread could potentially benefit.
EDIT: We could also talk about REW works (as well as other FFT-based spectrum analyzers), and its strengths and weaknesses for showing different types of sometimes audible phenomena...
Last edited:
@Markw4 said. "I spent many years as an engineer in some very well known universities"
More argument from authority. Even if speaker isolation could affect ITD (but how?), isolating both the left and right speakers would apply the same delay and therefore cancel. Hell, just move your head 1/16th of an inch and you have many microseconds of delay. So no, that's not a valid argument. And I did explain many times that a "competent" loudspeaker is one that is sufficiently massive and rigid as to not exhibit cabinet vibrations. If the cabinet doesn't move, then there's nothing to couple to the surface below. How many more times do I have to say that here? My article also makes that very clear. Why you guys are arguing such a patently ridiculous point, and ignoring all my evidence, seems amazing to me. You seem more interested in being right than knowing the truth. Again, where is your data?
As for Bob Katz's article, I'm very familiar with it and it's a joke. Did you read the conclusion? All of the differences were way down in the noise. RT60 change of 15 milliseconds in a room with more than half a second of RT60? He admits only "low level artifacts in the waterfall" with all other changes less than 1 dB. None of that stuff is enough to establish the "obvious improvements" reviewers and users claim.
EDIT AFTER SEEING YOUR EDIT: Yes, isolators could work with flimsy speakers. I'm not sure if they'd make things better or worse, but I can see a difference being possible. And you can be sure that difference would show in REW sweeps. But the Mackie HR series speaker I used for my tests, and the Focal TRIO6 monitors I have now, are not flimsy.
More argument from authority. Even if speaker isolation could affect ITD (but how?), isolating both the left and right speakers would apply the same delay and therefore cancel. Hell, just move your head 1/16th of an inch and you have many microseconds of delay. So no, that's not a valid argument. And I did explain many times that a "competent" loudspeaker is one that is sufficiently massive and rigid as to not exhibit cabinet vibrations. If the cabinet doesn't move, then there's nothing to couple to the surface below. How many more times do I have to say that here? My article also makes that very clear. Why you guys are arguing such a patently ridiculous point, and ignoring all my evidence, seems amazing to me. You seem more interested in being right than knowing the truth. Again, where is your data?
As for Bob Katz's article, I'm very familiar with it and it's a joke. Did you read the conclusion? All of the differences were way down in the noise. RT60 change of 15 milliseconds in a room with more than half a second of RT60? He admits only "low level artifacts in the waterfall" with all other changes less than 1 dB. None of that stuff is enough to establish the "obvious improvements" reviewers and users claim.
EDIT AFTER SEEING YOUR EDIT: Yes, isolators could work with flimsy speakers. I'm not sure if they'd make things better or worse, but I can see a difference being possible. And you can be sure that difference would show in REW sweeps. But the Mackie HR series speaker I used for my tests, and the Focal TRIO6 monitors I have now, are not flimsy.
Not exactly. Isolating both speakers will minimize asynchronous vibrations in the platforms upon which they sit (please recall stereo channels are not always playing the exact same content so they naturally tend to vibrate differently if allowed to do so). Stabilization could plausibly help minimize degradation of ITD localization cues between channels. It doesn't take much to throw off phase coherence between speaker channels when the human auditory system may be sensitive to inter-channel timing differences of as little as few microseconds (as shown in the relevant ITD literature).
EDIT: Also, please recall one of the main purposes of using near field monitors (as I did in my example) is to help minimize effects of local room acoustics.
Last edited:
How could a few microseconds of delay possibly have a meaningful effect on bass fullness and tightness? The main reason people believe they hear a change with isolation is because the speakers are then in a different location.
As for near field monitoring, room effects are still rampant. The waterfall below shows the ringing and response in a typical nearfield setup. (Blue is after adding substantial bass trapping, so the red is what most people have🙂
As for near field monitoring, room effects are still rampant. The waterfall below shows the ringing and response in a typical nearfield setup. (Blue is after adding substantial bass trapping, so the red is what most people have🙂
Attachments
Getting too much into how FFT spectrum analyzers work may risk going too far off topic for this thread. I will leave it to the moderators if a new thread should be split off for that purpose.
The very brief answer to your point about REW sweeps is that you are measuring what are called PSS test signals (periodic steady state), you aren't sweeping to a high enough frequency to detect ITD timing differences (you would need to go up to 100kHz or more), you are using spectral analysis from which phase information has omitted, and you are failing to use stereo speakers to you could have any chance at all of measuring possible ITD performance changes. IOW, you have failed to design sensible tests. You also don't seem to have any idea of the types of problems #227 was referring to.