| Home |
| Search |
| Today's Posts |
|
|
|
#1
|
|||
|
|||
Seeing/hearing and sighted/blind tests
Harry Lavo wrote:
This is a misrepresentation of those members' position. What some of us are saying is that you have to be cognizant of the effects of expectation bias, and take proper steps to control it , if you really want to find out if there are *audible only* differences. We always have said that if the differences are big enough, like those between speakers, then you don't really need DBT's to differentiate them. We don't say that "sight always overrides true differences" (in fact we argue if the audible difference exists in the first place), we are saying that expectation bias is very likely to override subtle differences, and that DBT is the best way to control for expectation bias. In the case of competent amps and speakers, we know that those differences should be subtle at best, from measurements like frequency response, distortion and signal-to-noise ratio tests. What you are saying above is a very reasonable position. Unfortunately, it seems to believed only in the abstract here. When somebody such as Michael comes on saying he can hear differences in amps...there is no questioning him on his listening conditions Actually I asked him whether he level-matched... , no consideration of the age or circuitry of the amps in question (despite one being a digital amp...the one chosen at that). .no discussion of his stated purpose or state of mind. All that happens is that he is told because he listened sighted, he is surely imagining things. He did describe his listening conditions. Maybe you have missed that? Then the turmoil ensues. The turmoil ensues because he refused to believe that expectation bias could lead to false positives when trying to detect differences. Now please answer this: do you agree with Michael on this key point? Do you believe that expectation bias should be controlled for? They should know better, but they don't seem to be able to allow even the possibility that there are real differences and that you might have heard them. No, they do, that's why they recommend the Harry Lavo's and Michael Scarpitti's of this newsgroup to do controlled tests to see if those differences are real. Heck, they even throw in real money to motivate them, in the case of cables. Insisting on a test that the "testees" don' t believe is valid. Nice 'gotcha. Now Harry, how does the word "recommend" become "insist"? As far as I am concerned, you don't have to do any controlled testing. You can pick amps/cables based on whatever criteria. However, when you want to convince others that there is real, audible, difference between them, you should use controlled testing like DBT to make sure that expectation bias (and other stuff like mismatched levels) does not invalidate your listening tests. DBT is the standard methodology on difference detection for such a long time, that I don't see any reason why Michael would have problem with it. So don't get upset...it's a world view of theirs that you are not going to change. All you need to change their world view is to pass the cable DBT test!  . Simple, isn't it?Sure, would greatly simplify the objectivist world-view here if we would just go away and stop challenging the test. The funny thing is no one asked you to go away and stop challenging the test. (In fact we even put up money hoping you would take the test.) And you were the one who said that Michael should go away and find some other topics to discuss... BTW, how would you know that Michael would fail a DBT on amps? But you can ignore them and instead focus on other topics of interest here on the forum. I think the subjectivists actually find this topic of great interest, based on how frequently they post in these threads... You think the objectivists ever let an assertion of heard differences pass without comment or challenge? You realize how many of these threads were started by subjectivists? |
|
#2
|
|||
|
|||
|
Seeing/hearing and sighted/blind tests
"chung" wrote in message
news:h1S0c.95108$Xp.423683@attbi_s54... Harry Lavo wrote: This is a misrepresentation of those members' position. What some of us are saying is that you have to be cognizant of the effects of expectation bias, and take proper steps to control it , if you really want to find out if there are *audible only* differences. We always have said that if the differences are big enough, like those between speakers, then you don't really need DBT's to differentiate them. We don't say that "sight always overrides true differences" (in fact we argue if the audible difference exists in the first place), we are saying that expectation bias is very likely to override subtle differences, and that DBT is the best way to control for expectation bias. In the case of competent amps and speakers, we know that those differences should be subtle at best, from measurements like frequency response, distortion and signal-to-noise ratio tests. What you are saying above is a very reasonable position. Unfortunately, it seems to believed only in the abstract here. When somebody such as Michael comes on saying he can hear differences in amps...there is no questioning him on his listening conditions Actually I asked him whether he level-matched... You did late in the game, but that was not the initial concern of those who responded. , no consideration of the age or circuitry of the amps in question (despite one being a digital amp...the one chosen at that). .no discussion of his stated purpose or state of mind. All that happens is that he is told because he listened sighted, he is surely imagining things. He did describe his listening conditions. Maybe you have missed that? Yes, and listening over stax headphones should probably have raised a few cautionary red flags among the objectivists, since listening on a really good set of headphones lets your hear things that ordinary speakers and room reflections might obscure. Then the turmoil ensues. The turmoil ensues because he refused to believe that expectation bias could lead to false positives when trying to detect differences. Now please answer this: do you agree with Michael on this key point? Do you believe that expectation bias should be controlled for? I believe it should be if you are after scientific proof. Done with a blind, monadic, evaluative test. I think it is a ridiculous burden to put on an audiophile trying to decide for himself what to buy...and it is an equally ridiculous thing to demand that he do before he dare discuss on this newsgroup the sound of products as he heard them on a "kitchen table test". They should know better, but they don't seem to be able to allow even the possibility that there are real differences and that you might have heard them. No, they do, that's why they recommend the Harry Lavo's and Michael Scarpitti's of this newsgroup to do controlled tests to see if those differences are real. Heck, they even throw in real money to motivate them, in the case of cables. Insisting on a test that the "testees" don' t believe is valid. Nice 'gotcha. Now Harry, how does the word "recommend" become "insist"? The insistance is that the test be a blind comparative a-b or a-b-x test rather than a serial, modadic, evaluative test that happens to be blind. As far as I am concerned, you don't have to do any controlled testing. You can pick amps/cables based on whatever criteria. However, when you want to convince others that there is real, audible, difference between them, you should use controlled testing like DBT to make sure that expectation bias (and other stuff like mismatched levels) does not invalidate your listening tests. DBT is the standard methodology on difference detection for such a long time, that I don't see any reason why Michael would have problem with it. He may not have a problem with it. Other than to consider it a waste of time. But he certainly didn't expect to be told that he was wrong, wrong, wrong to have thought he heard different sound characteristics from the amps because he didn't do the test double blind. He might or might not have heard such characteristics...he was given no benefit of the doubt. So don't get upset...it's a world view of theirs that you are not going to change. All you need to change their world view is to pass the cable DBT test! . Simple, isn't it?Sure, would greatly simplify the objectivist world-view here if we would just go away and stop challenging the test. The funny thing is no one asked you to go away and stop challenging the test. (In fact we even put up money hoping you would take the test.) And you were the one who said that Michael should go away and find some other topics to discuss... Nobody asked him to go away. We just suggested he not tear himself apart looking to convince you guys. It looks like he reached the same conclusion on his own. You yourself put the smiley in your statement, meaning you know it is a ridiculous statement that won't be accepted. Why, because you are asking us to "beat" a test that we think is flawed. Meanwhile, my efforts to point out why I and others believe it is flawed and to propose a proper control test have met with very little but stoney silence...as if the issue were never raised. If something uncomfortable comes up, just ignore it, right? The way to prove the test right or wrong is to devise the control test and get on with executing it. I say the best way to get to the other side of the fence is to open the gate. You say the best way is to lower my head and keep charging the fence. Which do you think has the best chance of getting us to the other side? BTW, how would you know that Michael would fail a DBT on amps? I don't know nor did I claim to know that he would. But like Mike Kueller, I believe the test technique itself loads the test in favor of "no difference". And I would never encourage him to undertake a loaded test. But you can ignore them and instead focus on other topics of interest here on the forum. I think the subjectivists actually find this topic of great interest, based on how frequently they post in these threads... You think the objectivists ever let an assertion of heard differences pass without comment or challenge? You realize how many of these threads were started by subjectivists? This one was started not once, but twice, by an objectivist. |
|
#3
|
|||
|
|||
|
Seeing/hearing and sighted/blind tests
Harry Lavo wrote:
"chung" wrote in message news:h1S0c.95108$Xp.423683@attbi_s54... Harry Lavo wrote: This is a misrepresentation of those members' position. What some of us are saying is that you have to be cognizant of the effects of expectation bias, and take proper steps to control it , if you really want to find out if there are *audible only* differences. We always have said that if the differences are big enough, like those between speakers, then you don't really need DBT's to differentiate them. We don't say that "sight always overrides true differences" (in fact we argue if the audible difference exists in the first place), we are saying that expectation bias is very likely to override subtle differences, and that DBT is the best way to control for expectation bias. In the case of competent amps and speakers, we know that those differences should be subtle at best, from measurements like frequency response, distortion and signal-to-noise ratio tests. What you are saying above is a very reasonable position. Unfortunately, it seems to believed only in the abstract here. When somebody such as Michael comes on saying he can hear differences in amps...there is no questioning him on his listening conditions Actually I asked him whether he level-matched... You did late in the game, but that was not the initial concern of those who responded. , no consideration of the age or circuitry of the amps in question (despite one being a digital amp...the one chosen at that). .no discussion of his stated purpose or state of mind. All that happens is that he is told because he listened sighted, he is surely imagining things. He did describe his listening conditions. Maybe you have missed that? Yes, and listening over stax headphones should probably have raised a few cautionary red flags among the objectivists, since listening on a really good set of headphones lets your hear things that ordinary speakers and room reflections might obscure. Yes, listening via headphones is more sensitive. So is using pink noise. But even a bigger red flag is the alleged "huge" differences that even someone blind and deaf could discern. And his insistence that he was immune from expectation bias. Then the turmoil ensues. The turmoil ensues because he refused to believe that expectation bias could lead to false positives when trying to detect differences. Now please answer this: do you agree with Michael on this key point? Do you believe that expectation bias should be controlled for? I believe it should be if you are after scientific proof. Good. Let's remember that. Done with a blind, monadic, evaluative test. We had this discussion before, and you could not make decisions once you had to compare. Interesting that Michael told us he actively *compared* the several amps. He did not have any difficulty discerning differences in your so-called "comparative" mode. So why not do the standard DBT? I think it is a ridiculous burden to put on an audiophile trying to decide for himself what to buy...and it is an equally ridiculous thing to demand that he do before he dare discuss on this newsgroup the sound of products as he heard them on a "kitchen table test". You seem to have missed the point. If he said he heard differences between amps, that would be just an anecdote. What raised the discussion level up was his insistence that expectation bias could never affect one's ability to discern difference. He is perfectly welcome to his belief, but when he flat out refutes the existing body of knowledge on human perception, that's where challenges come in. Note that you have erected another strawman again when you said "it is a ridiculous burden to put on an audiophile trying to decide for himself what to buy". How many times do you need to put that up and burn it? Can you tried to be more objective in representing viewpoints of those you don't agree with? They should know better, but they don't seem to be able to allow even the possibility that there are real differences and that you might have heard them. No, they do, that's why they recommend the Harry Lavo's and Michael Scarpitti's of this newsgroup to do controlled tests to see if those differences are real. Heck, they even throw in real money to motivate them, in the case of cables. Insisting on a test that the "testees" don' t believe is valid. Nice 'gotcha. Now Harry, how does the word "recommend" become "insist"? The insistance is that the test be a blind comparative a-b or a-b-x test rather than a serial, modadic, evaluative test that happens to be blind. But we never insist that he has to do that in selecting components. If he wants to prove without doubt to us, of course then he has to use a bias-controlled methodology, and DBT is such a commonly used methodology. BTW, he did not think that he would have problem passing the DBT, unlike in your case where "comparative" causes panic. As far as I am concerned, you don't have to do any controlled testing. You can pick amps/cables based on whatever criteria. However, when you want to convince others that there is real, audible, difference between them, you should use controlled testing like DBT to make sure that expectation bias (and other stuff like mismatched levels) does not invalidate your listening tests. DBT is the standard methodology on difference detection for such a long time, that I don't see any reason why Michael would have problem with it. He may not have a problem with it. Other than to consider it a waste of time. But he certainly didn't expect to be told that he was wrong, wrong, wrong to have thought he heard different sound characteristics from the amps because he didn't do the test double blind. What he was told "wrong, wrong Wrong" was his insistence that expectation bias could not lead to false positives. You believe he's right? He might or might not have heard such characteristics...he was given no benefit of the doubt. I gave him the benefit of the doubt that he truly believes what he heard was real. So don't get upset...it's a world view of theirs that you are not going to change. All you need to change their world view is to pass the cable DBT test! . Simple, isn't it?Sure, would greatly simplify the objectivist world-view here if we would just go away and stop challenging the test. The funny thing is no one asked you to go away and stop challenging the test. (In fact we even put up money hoping you would take the test.) And you were the one who said that Michael should go away and find some other topics to discuss... Nobody asked him to go away. We just suggested he not tear himself apart looking to convince you guys. It looks like he reached the same conclusion on his own. It did not appear that he is tearing himself apart at all! You yourself put the smiley in your statement, meaning you know it is a ridiculous statement that won't be accepted. No, your interpretation is wrong. I put the smiley because I know that you or others know that you will fail the cable test, and will not participate. It is not because the test is ridiculous (why is it ridiculous, Harry, given night and day differences?), it is because the subjectivists are really afraid to learn that without sight information, they do not have that touted ability to discriminate. Why, because you are asking us to "beat" a test that we think is flawed. You said it's flawed because of that mystical comparative vs evaluative dilemma. Others said that it is flawed because the test period is too long, too short, snippets too long, too short, while in academics and industry DBT's are used day in and day out. Meanwhile, my efforts to point out why I and others believe it is flawed and to propose a proper control test have met with very little but stoney silence...as if the issue were never raised. Stoney silence? You seem to have very selective memory. If you are looking for posts supporting your proposal only, yeah, I guess that was stoney silence .If something uncomfortable comes up, just ignore it, right? Please reread the responses to your proposal. The way to prove the test right or wrong is to devise the control test and get on with executing it. You seem to want to prove an established methodology is wrong. Yeah, go ahead and prove it. I say the best way to get to the other side of the fence is to open the gate. You say the best way is to lower my head and keep charging the fence. Which do you think has the best chance of getting us to the other side? Nil, because you made up your mind a long time ago not to go there! BTW, how would you know that Michael would fail a DBT on amps? I don't know nor did I claim to know that he would. But like Mike Kueller, I believe the test technique itself loads the test in favor of "no difference". And I would never encourage him to undertake a loaded test. Maybe because in most cases, there is really no difference? But you can ignore them and instead focus on other topics of interest here on the forum. I think the subjectivists actually find this topic of great interest, based on how frequently they post in these threads... You think the objectivists ever let an assertion of heard differences pass without comment or challenge? You realize how many of these threads were started by subjectivists? This one was started not once, but twice, by an objectivist. And subjectivists would never let an assertion by the other side pass without comment or challenge, so there you go... |
|
#4
|
|||
|
|||
|
Seeing/hearing and sighted/blind tests
On Tue, 02 Mar 2004 18:08:49 GMT, "Harry Lavo"
wrote: "chung" wrote in message news:h1S0c.95108$Xp.423683@attbi_s54... He did describe his listening conditions. Maybe you have missed that? Yes, and listening over stax headphones should probably have raised a few cautionary red flags among the objectivists, since listening on a really good set of headphones lets your hear things that ordinary speakers and room reflections might obscure. Indeed yes, they are very sensitive tools, which is why no one has denied that these *might* be real differences. It has however been pointed out that he can't *know* this because sighted listening is *proven* to be useless for determining subtle sonic differences, and it's also been pointed out that it's not a very competent comparison of power amps, which should of course be tested with a representative loudspeaker load, not merely headphones. It's actually *more* likely that the amps would sound the same under such a light load....... -- Stewart Pinkerton | Music is Art - Audio is Engineering |
|
#7
|
|||
|
|||
|
Seeing/hearing and sighted/blind tests
|
|
#8
|
|||
|
|||
|
Seeing/hearing and sighted/blind tests
(Nousaine) wrote in message ...
(Michael Scarpitti) wrote: What you're suggesting is that merely knowing the names of these products -- and ONLY that -- will 'create' a whole sonic 'signature' for each! Big over-generalization. No one has ever suggested that. Of course you have. That's preposterous on its face. Remember, I knew nothing about any of these amps whatsoever, except what in one case: what I read in the Harmon-Kardon literature. That makes one wonder why you were 'looking.' But who cares. Just for the record I'm used to seeing the name Harman spelled without an "o". Am I missing something? I really did not spend much time looking at the amps. I just hooked them up and listened. If I were 'prone' to hear differences between them that may or may not have there (your claim), that same 'bias' (your term) should be equally present regardless of my 'knowledge' of which amp I was listening to. Could be; but there are other possible bias factors involved such as input sensitivities or otherwise un-matched levels or dissimilar programs. In other words, why should I be MORE 'biased' (your term) by the name 'PS Audio' than 'A'? Don't know; never said that you were. Appearance is another good mechanism. But to answer the specific question more specifically much of human bias (not tobe confused with acoustical bias) is held at the subconscious level and technically not directly 'known' to the subject at a conscious level. So what? I'm only concerned with acoustically-based results. Which is why I prefer listening that does everything it can, given a set-up, to reduce sound quality and acoustical performance to acoustical factors as much as possible. I think you misunderstood. Suppose you keep me from seeing either of two amps, so I don't know their brand names. You call one 'A' and one 'B'. You hook up 'A' for me to my Stax Lambda transformer set-up, and I listen. Then you switch to 'B', and I listen. How on Earth does that make any difference? Why is 'A' different from 'Bryston' or 'Sony'? If I had no more opinion or knowledge of 'Bryston' or 'Sony' than I do of 'A' or 'B', how could it possibly matter? Obviously, it cannot. I must point out that on several occasions I have made listening tests that showed no difference. I discussed this before. So, I am quite capable of detecting no difference as well as some difference. |
Hybrid Mode
