Home |
Search |
Today's Posts |
#241
|
|||
|
|||
Why do the anti-ABX folks not deliver?
|
#243
|
|||
|
|||
Why do the anti-ABX folks not deliver?
(Mkuller) wrote in message news:BXb0b.183627$YN5.135766@sccrnsc01...
(ludovic mirabel) wrote in message : 4) that reporting as the end-outcome the majority vote of ABX untrained/untalented and ignoring the capable minority who DO hear-exactly the way Greenhill had done (see Marcus quotes)is a perversion of "testing". (Bob Marcus) Date: 8/17/03 4:57 PM Pacific Daylight Time Message-id: wrote: "Ignoring the minority who DO have fewer and less severe colds after megadosing vitamin C is a perversion of testing." See how silly that argument is? So stop using it. Bob, your analogy comparing DBTs to medical research is wrongheaded. No, just inexact, as all analogies seem to turn out to be. (Perhaps RAHE should ban them.) I meant to give Mr. Mirabel an example of the kind of cherry-picking and selection bias he practices. If one person in the Greenhill group, or any other group reliably identifies a difference - then it proves a difference exists! Granted. Although Mr. Mirabel cannot point to a single case in which this occurred (absent willful misreading, of course). The fact that the others or a vast majority are not able to reliably identify it under THAT test and THOSE test conditions means next to nothing. It's a null - 0 - meaningless. So why aren't there any blind tests published which show differences? Except for Greenhill's five, and Arny's old article. Let's be fair, Mike. There have been a few. (1) Blind testing seems to filter out subtle audible differences. Even the experts agree that music is not a very sensitive source and it is the only meaningful one to audiophiles. True, but if it's not a sensitive source for blind tests, then it's not a sensitive source for sighted tests either. (Unless you have some theory for how your ear-brain mechanism works differently when you know which cable is connected.) (2) The only people who care about blind testing are the "debunkers" who are trying to show "there are no differences" to the unwashed audiophile masses (most of whom couldn't care less) And just why do you think anyone is "trying" to show this? What have any of us to gain by it? bob |
#244
|
|||
|
|||
Why do the anti-ABX folks not deliver?
|
#246
|
|||
|
|||
Why do the anti-ABX folks not deliver?
|
#247
|
|||
|
|||
Why do the anti-ABX folks not deliver?
|
#248
|
|||
|
|||
Why do the anti-ABX folks not deliver?
(ludovic mirabel) wrote in message news:0He0b.184753$Ho3.25747@sccrnsc03...
(Audio Guy) wrote in message news:G680b.182959$uu5.34042@sccrnsc04... Bob has demonstrated quite well on his own how your "interpretations" of Greenhill's comments are misinterpretations. So let's have no interpreting. Just poor Greenhill's own words, word for word and literally and Marcus own words, word for word and literally. Marcus first: Anyone tempted to take Mirabel's assertion here at face value should consult the original article and read for himself that Greenhill said no such thing. bob This is what Greenhill said reporting the outcome of his cable comparison test: ("The Stereo Review, August 1982): "Final significant conclusion one can draw is that at least one genuine "golden ear" exists. Obviously certain listeners whether through talent, training or experience can hear small differences between components" Apparently the only way you can argue your case is to juxtapose my words with something I was not responding to (and which did not even appear in the post you pulled my comment from). A rather pitiful act of desperation, I'd say. bob |
#249
|
|||
|
|||
Why do the anti-ABX folks not deliver?
(Nousaine) wrote in message news:tcX%a.173930$Ho3.23369@sccrnsc03...
Let me first of all say that I truly appreciate your text. It is what I believe a discussion should be: with serious content, courteously written- not clowning for the audience, picking on single words to score points and listing your opponent's personal deficiences. I also do not doubt your good faith and considerable investment of time and money in this subject. But... Ludovic has posted earlier that he though that past controlled listening research had "hidden" positive scoring subjects with overall null results and that I had been avoiding him about that issue. I recently responded but my post was rejected. So here’s a straight reply to the "hidden audibility" claim. First, there have been no hidden positive subject scoring in any research that I’ve been able to note. In spite of the repeat claims that the Greenhill wire test had high scoring individuals whose results were concealed by averaging there is no data that either confirms or suggests this conclusion. This has been covered quite extensively in this newsgroup and in recent threads. Let me shift the emphasis. It is not on individuals ( such as Greenhill's own "golden ear"). It is on the indisputable fact that the individual performances in the reported tests vary enormously, Example: Your witness Sean Olive's results in his "listening room" pdf (www.revelspeakers.- address quoted from memory). One of his supertrained professionals' was scoring around 30% , a few 50% or thereabouts, most reached (how many repeats?), significant 70% or more. Greenhill: all scored brillianly ( almost all were 100% correct) recognising 1.75db. volume difference between a thick and a thin cable- as long as the pink noise was played. The same people failed to recognise (within Greenhill's statistical validity criteria- ie had much lower "correct" scores) the same volume difference once music replaced pink noise. You or others said that it only proves that pink noise is a better "test" signal. Sorry: doesn't wash with me. The subject is reproduction of MUSIC by audio components- not pink noise. in fact saying that ABX performs better with pink noise than music, amounts to saying that ABX is an inappropriate test for assessing MUSICAL reproduction differences between components. Is a "test" that has such variable individual results an appropriate test for audiophile use? How can you be certain that the failure of most of your subjects to recognise differences was not due to the nature of your test? Especially as you must agree that training in ABXing improves individual results. How does an audiophile know if he had enough training? Next; of the tests I have personally designed, proctored or have been a subject there has been no cadre of ‘those who hear’ compared to the general subject population. IOW there has never been an experiment where a few significant scoring individuals were concealed by the average. Indeed with a reasonably sized sample often a single significant result would be enough to make the overall results positive. Of the tests I’ve designed and conducted subject count has varied from 1 to 31 with sample sizes ranging from 10 to 431. Programs were selected by subjects in the ‘challengeâ€à ƒÂ‚™ experiments where subjects had claimed to already have confirmed differences. Otherwise programs were selected from a collection of 63 musical and voice selections (including many with acoustical instrumentation) that had been chosen because they represent specific reproduction challenges to audio systems. Every experiment was extensively examined for internal significance. Most of them allowed subjects to extend the test with more trials to improve scores. This was seldom asked for and no individual improved his score enough to attain significance with additional trials OR a repeat of the entire experiment. There were NO individual significant scores in any experiment that wasn't significant overall. I'd have to know what you played to them, what were their ages and sex, their ABX training, their musical interests and exposure- a thousand details- before I could say that you ruled out with significant validity the possibility that you missed that rare bird who was not only a great listener but also a great ABX performer. ONE is enough. It is not about pleasing the crowd. It is about "high-end". Something like Mr. Pinkerton- an ABX upholder whom you did not convince that there are no differences between amps. Nor , of course did he convince you. However in the case of Flying Blind (Audio Magazine) a subject who was unable to discern whether a given program contained a confirmed audible level of distortion in 16 weeks of long term listening was able to reliably identify same with a 6 second segment of that program using the ABX technique. But any ofd this notwithstanding I'm still not fully understanding why some individual, some company or some 3rd party has ever been able to confirm "amp" and "wire" sound under any set of conditions with even modest listening bias controls implemented. Even IF someone was withholding the Truth why hasn't someone else willfully or even by accident confirmed it? If these 'differences' are so evident that people like Ludovic are so certain of their existence WHY hasn't some party, interested or otherwise, stumbled across the body instead of just the rumors? It just doesn't seem likely that people as smart as Earl Geddes, Floyd Toole, David Clark, David Rich, Sean Olive, Stan Lip****z, John Vanderkooy, Dan Shanefiled, Rich Cabot, John Eargle, Dick Pierce, et al would have ALL overlooked even the minutest of evidence of true audibility. All your notables to the best of my knowledge did not report component comparisons- where there is NO OBJECTIVE END-POINT- but rather reasearch about perception of KNOWN induced artefact-is the answer right or not. Usually by selected, trained audiences at that. Apples and oranges. Ludovic Mirabel |
#250
|
|||
|
|||
Why do the anti-ABX folks not deliver?
(Mkuller) wrote
(1) Blind testing seems to filter out subtle audible differences. Even the experts agree that music is not a very sensitive source and it is the only meaningful one to audiophiles. (Bob Marcus) wrote: True, but if it's not a sensitive source for blind tests, then it's not a sensitive source for sighted tests either. (Unless you have some theory for how your ear-brain mechanism works differently when you know which cable is connected.) Ah, your misguided logic rear its ugly head once again. In sighted tests, most trained listeners can hear and reliably identify differences between amps (for example). Those differences disappear in blind tests. Do you have proof (or just your "logical" inference) that the test itself is not filtering out the differences? I'll leave the theories about ear-brain mechanisms with music and blind tests to the real scientists (not the engineers, statisticians or pseudo-scientists). (2) The only people who care about blind testing are the "debunkers" who are trying to show "there are no differences" to the unwashed audiophile masses (most of whom couldn't care less) And just why do you think anyone is "trying" to show this? What have any of us to gain by it? So you tell us all - why do you and the other pro-DBTers persist in this endless debate? Proving you're right? Winning the debate? Getting in the last post? Regards, Mike |
#251
|
|||
|
|||
Why do the anti-ABX folks not deliver?
|
#252
|
|||
|
|||
Why do the anti-ABX folks not deliver?
Mkuller wrote:
(Mkuller) wrote (1) Blind testing seems to filter out subtle audible differences. Even the experts agree that music is not a very sensitive source and it is the only meaningful one to audiophiles. (Bob Marcus) wrote: True, but if it's not a sensitive source for blind tests, then it's not a sensitive source for sighted tests either. (Unless you have some theory for how your ear-brain mechanism works differently when you know which cable is connected.) Ah, your misguided logic rear its ugly head once again. In sighted tests, most trained listeners can hear and reliably identify differences between amps (for example). Those differences disappear in blind tests. Do you have proof (or just your "logical" inference) that the test itself is not filtering out the differences? Do you have *proof* that it does? Your evidence from sighted listening to amps doesn't begin to be sufficient. That's because we have tons of proof that sighted bias exists and is *common*. We have no proof that 'filtering' of the sort you posit exists. We also have proof that blind testing does not always result in 'no difference'. Components such as speakers can routinely 'pass' blind tests. Components such as cables, cannot. You speak of logic. Your logic is utterly flawed. Your 'evidence' -- that sighted listeners can reliably identify differences between amps -- amoutns to a giant *duh* from a psychoacoustics perspective. All 'explanations' are not created equal'; the exercise of rationality involves evaluating which ones are supported by the existing evidence and which are not. When you have provided *evidence* that 'filtering' exists, it will leave the realm of pure speculation. For now, it's a distinctly less rational explanation than sighted bias, for the disparity between 'perception' and reality. I'll leave the theories about ear-brain mechanisms with music and blind tests to the real scientists (not the engineers, statisticians or pseudo-scientists). Er...you do know taht real scientists acknowledge the existence of perceptual bias every day in their work, right? (most of whom couldn't care less) And just why do you think anyone is "trying" to show this? What have any of us to gain by it? So you tell us all - why do you and the other pro-DBTers persist in this endless debate? Proving you're right? Winning the debate? Getting in the last post? So tell us, Mike -- *who started this thread*? And *why*? -- -S. |
#253
|
|||
|
|||
Why do the anti-ABX folks not deliver?
|
#254
|
|||
|
|||
Why do the anti-ABX folks not deliver?
|
#256
|
|||
|
|||
Why do the anti-ABX folks not deliver?
On Wed, 20 Aug 2003 04:24:08 GMT, Steven Sullivan
wrote: Mkuller wrote: So you tell us all - why do you and the other pro-DBTers persist in this endless debate? Proving you're right? Winning the debate? Getting in the last post? So tell us, Mike -- *who started this thread*? And *why*? Check the header, Steven! I started *this* thread, but it's an obvious rebuttal of Ludovic's 'why does ABX not deliver' attack thread. -- Stewart Pinkerton | Music is Art - Audio is Engineering |
#257
|
|||
|
|||
Why do the anti-ABX folks not deliver?
|
#258
|
|||
|
|||
Why do the anti-ABX folks not deliver?
Steven Sullivan wrote in message news:IJC0b.203356$o%2.94507@sccrnsc02...
Mkuller wrote: (Mkuller) wrote (1) Blind testing seems to filter out subtle audible differences. Even the experts agree that music is not a very sensitive source and it is the only meaningful one to audiophiles. (Bob Marcus) wrote: True, but if it's not a sensitive source for blind tests, then it's not a sensitive source for sighted tests either. (Unless you have some theory for how your ear-brain mechanism works differently when you know which cable is connected.) Ah, your misguided logic rear its ugly head once again. In sighted tests, most trained listeners can hear and reliably identify differences between amps (for example). Those differences disappear in blind tests. Do you have proof (or just your "logical" inference) that the test itself is not filtering out the differences? Do you have *proof* that it does? Your evidence from sighted listening to amps doesn't begin to be sufficient. That's because we have tons of proof that sighted bias exists and is *common*. We have no proof that 'filtering' of the sort you posit exists. We also have proof that blind testing does not always result in 'no difference'. Components such as speakers can routinely 'pass' blind tests. Components such as cables, cannot. You speak of logic. Your logic is utterly flawed. Your 'evidence' -- that sighted listeners can reliably identify differences between amps -- amoutns to a giant *duh* from a psychoacoustics perspective. All 'explanations' are not created equal'; the exercise of rationality involves evaluating which ones are supported by the existing evidence and which are not. When you have provided *evidence* that 'filtering' exists, it will leave the realm of pure speculation. For now, it's a distinctly less rational explanation than sighted bias, for the disparity between 'perception' and reality. I'll leave the theories about ear-brain mechanisms with music and blind tests to the real scientists (not the engineers, statisticians or pseudo-scientists). Er...you do know taht real scientists acknowledge the existence of perceptual bias every day in their work, right? (most of whom couldn't care less) And just why do you think anyone is "trying" to show this? What have any of us to gain by it? So you tell us all - why do you and the other pro-DBTers persist in this endless debate? Proving you're right? Winning the debate? Getting in the last post? So tell us, Mike -- *who started this thread*? And *why*? Mr. Stewart Pinkerton on July 26th under this title. Direct the "why?" question to the right address. Ludovic Mirabel |
#259
|
|||
|
|||
Why do the anti-ABX folks not deliver?
|
#260
|
|||
|
|||
Why do the anti-ABX folks not deliver?
(ludovic mirabel) wrote:
(Nousaine) wrote in message news:tcX%a.173930$Ho3.23369@sccrnsc03... Let me first of all say that I truly appreciate your text. It is what I believe a discussion should be: with serious content, courteously written- not clowning for the audience, picking on single words to score points and listing your opponent's personal deficiences. I also do not doubt your good faith and considerable investment of time and money in this subject. But... Ludovic has posted earlier that he though that past controlled listening research had "hidden" positive scoring subjects with overall null results and that I had been avoiding him about that issue. I recently responded but my post was rejected. So hereÃ_¢Ã_Â_Ã_Â_s a straight reply to the "hidden audibility" claim. First, there have been no hidden positive subject scoring in any research that IÃ_¢Ã_Â_Ã_Â_ve been able to note. In spite of the repeat claims that the Greenhill wire test had high scoring individuals whose results were concealed by averaging there is no data that either confirms or suggests this conclusion. This has been covered quite extensively in this newsgroup and in recent threads. Let me shift the emphasis. It is not on individuals ( such as Greenhill's own "golden ear"). It is on the indisputable fact that the individual performances in the reported tests vary enormously, We'll of course. Individuals guessing will form a distribution with tals. Even when there is no difference a large sample or individuals retested there will be a distribution with tails toeither side. Ludovic wants us to consider any subject who scores better much than 50% as a person who "heard" something whether the individual score or overall score suggests this was the case. Example: Your witness Sean Olive's results in his "listening room" pdf (www.revelspeakers.- address quoted from memory). One of his supertrained professionals' was scoring around 30% , a few 50% or thereabouts, most reached (how many repeats?), significant 70% or more. OK? So what? Given a chance result we'de find a distribution and we'd find a buried Golden Ear according tjo Ludovic's reasoning. Why are there no lucky coins in his experiments? Greenhill: all scored brillianly ( almost all were 100% correct) recognising 1.75db. volume difference between a thick and a thin cable- as long as the pink noise was played. The same people failed to recognise (within Greenhill's statistical validity criteria- ie had much lower "correct" scores) the same volume difference once music replaced pink noise. You or others said that it only proves that pink noise is a better "test" signal. Sorry: doesn't wash with me. T Why not? Noise is the single most effective signals for revealing small audible differences he subject is reproduction of MUSIC by audio components- not pink noise. in fact saying that ABX performs better with pink noise than music, amounts to saying that ABX is an inappropriate test for assessing MUSICAL reproduction differences between components. Is a "test" that has such variable individual results an appropriate test for audiophile use? How can you be certain that the failure of most of your subjects to recognise differences was not due to the nature of your test? Especially as you must agree that training in ABXing improves individual results. How does an audiophile know if he had enough training? Next; of the tests I have personally designed, proctored or have been a subject there has been no cadre of Ã_¢Ã_Â_Ã_Â_those who hearÃ_¢Ã_Â_Ã_Â_ compared to the general subject population. IOW there has never been an experiment where a few significant scoring individuals were concealed by the average. Indeed with a reasonably sized sample often a single significant result would be enough to make the overall results positive. Of the tests IÃ_¢Ã_Â_Ã_Â_ve designed and conducted subject count has varied from 1 to 31 with sample sizes ranging from 10 to 431. Programs were selected by subjects in the Ã_¢Ã_Â_Ã_Â_challengeÃ_¢Ã_Â_Ã_Â_ experiments where subjects had claimed to already have confirmed differences. Otherwise programs were selected from a collection of 63 musical and voice selections (including many with acoustical instrumentation) that had been chosen because they represent specific reproduction challenges to audio systems. Every experiment was extensively examined for internal significance. Most of them allowed subjects to extend the test with more trials to improve scores. This was seldom asked for and no individual improved his score enough to attain significance with additional trials OR a repeat of the entire experiment. There were NO individual significant scores in any experiment that wasn't significant overall. But ..... I'd have to know what you played to them, what were their ages and sex, their ABX training, their musical interests and exposure- a thousand details- before I could say that you ruled out with significant validity the possibility that you missed that rare bird. Soo exactly what are the characteristics of a 'rare bird'? What age group? What training? What musical interests? what thousand details ... can you list them? If not how am Is upposed toknow what meets your criteria? And how my subjects were not qualified. who was not only a great listener but also a great ABX performer. ONE is enough. It is not about pleasing the crowd. It is about "high-end". Something like Mr. Pinkerton- an ABX upholder whom you did not convince that there are no differences between amps. Nor , of course did he convince you. However in the case of Flying Blind (Audio Magazine) a subject who was unable to discern whether a given program contained a confirmed audible level of distortion in 16 weeks of long term listening was able to reliably identify same with a 6 second segment of that program using the ABX technique. But any ofd this notwithstanding I'm still not fully understanding why some individual, some company or some 3rd party has ever been able to confirm "amp" and "wire" sound under any set of conditions with even modest listening bias controls implemented. Even IF someone was withholding the Truth why hasn't someone else willfully or even by accident confirmed it? If these 'differences' are so evident that people like Ludovic are so certain of their existence WHY hasn't some party, interested or otherwise, stumbled across the body instead of just the rumors? It just doesn't seem likely that people as smart as Earl Geddes, Floyd Toole, David Clark, David Rich, Sean Olive, Stan Lip****z, John Vanderkooy, Dan Shanefiled, Rich Cabot, John Eargle, Dick Pierce, et al would have ALL overlooked even the minutest of evidence of true audibility. All your notables to the best of my knowledge did not report component comparisons- So you have to be controlled test or ABX qualified to exist? OK I'll buy that. What aew the qualifications of yourself or reference listeners? where there is NO OBJECTIVE END-POINT- but rather reasearch about perception of KNOWN induced artefact-is the answer right or not. Usually by selected, trained audiences at that. Apples and oranges. Ludovic Mirabel Why don't you give us a list of researcjh that refutes what we already know about amps and wires? |
#261
|
|||
|
|||
Why do the anti-ABX folks not deliver?
|
#262
|
|||
|
|||
Why do the anti-ABX folks not deliver?
Stewart Pinkerton wrote:
On Wed, 20 Aug 2003 04:24:08 GMT, Steven Sullivan wrote: Mkuller wrote: So you tell us all - why do you and the other pro-DBTers persist in this endless debate? Proving you're right? Winning the debate? Getting in the last post? So tell us, Mike -- *who started this thread*? And *why*? Check the header, Steven! I started *this* thread, but it's an obvious rebuttal of Ludovic's 'why does ABX not deliver' attack thread. Yup, the thread I was referring to was the parent 'Why do the anti-ABX folks not deliver' thread, sorry. -- -S. |
#263
|
|||
|
|||
Why do the anti-ABX folks not deliver?
Nousaine wrote:
(Mkuller) wrote: (Mkuller) wrote (1) Blind testing seems to filter out subtle audible differences. Even the experts agree that music is not a very sensitive source and it is the only meaningful one to audiophiles. (Bob Marcus) wrote: True, but if it's not a sensitive source for blind tests, then it's not a sensitive source for sighted tests either. (Unless you have some theory for how your ear-brain mechanism works differently when you know which cable is connected.) Ah, your misguided logic rear its ugly head once again. In sighted tests, most trained listeners can hear and reliably identify differences between amps (for example). You can't say they can reliably identify differences because the 'answers' are known in advance and are directly subject to conditions and internal human bias mechanisms. All he can say is that when listeners know which component is playing... they can 'reliably' say which component is playing. Hence the 'duh' aspect. If he says otherwise, he doesnt' understand what is meant by 'sighted' test. ___ -S. |
#264
|
|||
|
|||
Why do the anti-ABX folks not deliver?
Tom said
(S888Wheel) "debates" endlessly: It takes two to debate Tom. Tom said That's fair enough; but neither you nor any interested party nor Stewart has delivered a smoking gun on amp or wire sound. IF these 'differences' truly existed than there would have been clear verification by now. This is not a new question nor is in soluable. I said If you always choose not to accept evidence of positive tests on the grounds of "uniqueness" of the results, those results will retain their "uniqueness" every time they are presented in a new test. Tom said No they won't. If they get replicated then they become the new standard. But "cold fusion" remains unique. Do we have to accept it for that reason? But events and experiments that remain 'unique' do so for a reason. I said I guess you didn't get the point. They will remain unique if you keep rejecting them for uniqueness. Tom said Please. If there are replicable experiments confirming Stewart's results they will make the new standard. Really? So if Stewert did the same tests and got the same results you would now consider this to be the new standard? Tom said It's exactly like Cold Fusion; Nope. Cold fusion claim was actually put to the test by scientists doing scientifically valid tests. Tom said why hasn't that been confirmed? See above. Tom said And why should we accept it as a 'given' until it has? Who said you should accept cold fusion as real? Especially now that it has failed real scientific investigation. Tom said In this case we've had a couple dozen experiments with other resiults. Yes you have a couple dozen anecdotal tests, scientifically speaking, that have used a multitude of protocols, in some cases introduced multiple variables with no need, in some cases used no varification protocols for other possible cases of nulls and seem to make common unfounded assumptions about listener skills, and a crazy idea that it is an either or propostion when it comes to amps sounding the same or different. Stewerts conclusions do at least have the unique distinction of considering the possibility that some amps may sound the same and some may sound different and that fact may depend on the speakers used. Tom said And we've had a couple+ positive experiments that have confirmed lack of nominal competence. Competence is a matter of opinion in this case. Tom said Why should we give this experiment more weight that the remaining evidence? Who said it should get *more* weight? Why should it get less weight? In your case you seem to be giving it zero weight. It looks like your criteria is the result and only the result. That is bad science 101. Tom said If you have a good reason, other than you are uncomfortable with the other evidence, please tell us. Reason for what? You are the one picking and choosing your anecdotal evidence. I don't see much reason to give Stewert's tests more or less weight than the ones you like. I said You get to keep starting the count at zero with that method. Tom said Not true. No one does that silly game except you. Balony. Cite an example of me doing this. You are still very consistant in mischaracterizing my opinions. You are still 0 for many. Tom said You get one un-replicated and un-documented anecdote and then want us to ignore ALL the other evidence. Balony. Never said I want you or anyone else to ignore any evidence. Try to get the facts straight some time in this debate. Also none of the tests you believe to be valid have been "replicated" and Stewert's tests, as Stewert has said, are documented here on this forum. It is you and only you who wishes to pick and choose your evidence. I have not suggested that *any* evidence be ignored or given more weight. I said Your analogy with cold fusion is way off base. One would not have to do the equivelant of inventing cold fusion to come up with an amp that sounds different. Tom said For sure. Just 'invent' a triode with high output impedance or put a series resistor in the output leg of any competent amplifier and you can get 'difference.' Then we agree that the cold fusion analogy is bogus. Tom said An equalizer is lots easier and much faster. So? See above. I said Discounting any test on the grounds of the results is very unscientific. That is a choice you have made. Not me. Tom said Ok then you tell me; if Stewart's experiment is taken at face value it says that a wide variety of modern amplifiers are sonically indistingushable and those that aren't are "*******" does it not? The corollary becomes: amplifiers sonically transparent to a $$$ Krell are available at far less cost. I said That isn't what Stewert's test tells us it is what Stewert tells us. Tom said Oh; he's not qualified to analyze his own experiment? Sure he is. I never said he wasn't. But his subjective conclusions on what is a "tosser' is valid for his personal listening pleasure and is not a global truth. He never said it was either. |
#265
|
|||
|
|||
Why do the anti-ABX folks not deliver?
"Mkuller" wrote in message
Ah, your misguided logic rear its ugly head once again. In sighted tests, most trained listeners can hear and reliably identify differences between amps (for example). Those differences disappear in blind tests. Talk about misguided logic! In a sighted evaluation anybody with normal sight can reliably identify differences between amps (for example), even with the power turned off no source, and speakers disconnected. Listener training has been conclusively shown to be absolutely not required during sighted evaluations, which should please certain of our posters who have been railing against listening tests that benefit from listener training. Do you have proof (or just your "logical" inference) that the test itself is not filtering out the differences? I have absolute proof that sighted evaluations aren't listening tests, using the very criteria proposed above. All persons with normal sight can reliably identify differences between just about any kind of audio products in a sighted evaluation, even with no power, and no input signal applied to the equipment being evaluated. Loudspeakers need not be used either. Therefore, we can conclude that sighted evaluations aren't listening tests since the results can be the same when listening is intentionally excluded. I'll leave the theories about ear-brain mechanisms with music and blind tests to the real scientists (not the engineers, statisticians or pseudo-scientists). No theories about ear-brain mechanisms are required to prove that sighted evaluations aren't listening tests. (2) The only people who care about blind testing are the "debunkers" who re trying to show "there are no differences" to the unwashed audiophile masses (most of whom couldn't care less) And just why do you think anyone is "trying" to show this? What have any of us to gain by it? So you tell us all - why do you and the other pro-DBTers persist in this endless debate? There appears to be considerable agreement on this point. We pro-DBTers persist in this endless debate because the anti-DBT forces restart it again and again. Proving you're right? It appears that according to the criteria you provided Mr. Kuller, DBTs are the only listening tests around. Winning the debate? When there is only one option, there is no debate to win, right? Getting in the last post? It can't be known a priori whether one gets the last post due to irrelevant, disinterest, or delivery of a overwhelming argument. |
#266
|
|||
|
|||
Why do the anti-ABX folks not deliver?
|
#267
|
|||
|
|||
Why do the anti-ABX folks not deliver?
On Wed, 20 Aug 2003 16:53:05 GMT, Steven Sullivan
wrote: Stewart Pinkerton wrote: On Wed, 20 Aug 2003 04:24:08 GMT, Steven Sullivan wrote: Mkuller wrote: So you tell us all - why do you and the other pro-DBTers persist in this endless debate? Proving you're right? Winning the debate? Getting in the last post? So tell us, Mike -- *who started this thread*? And *why*? Check the header, Steven! I started *this* thread, but it's an obvious rebuttal of Ludovic's 'why does ABX not deliver' attack thread. Yup, the thread I was referring to was the parent 'Why do the anti-ABX folks not deliver' thread, sorry. I do believe you just did it again! :-) -- Stewart Pinkerton | Music is Art - Audio is Engineering |
#268
|
|||
|
|||
Why do the anti-ABX folks not deliver?
On Wed, 20 Aug 2003 16:46:19 GMT, (Mkuller) wrote:
(Nousaine) wrote: Actually it's people like you that continue the endless debate. If you had any evidence that you were right than the 'debate' would stop instantly. Uh, Tom, perhaps you aren't paying attention (or maybe it's just wishful thinking). Both sides have been providing "evidence" of their positions for years and no one on the other side has changed their mind or switched sides. The endless debate continues... It's mental masturbation, and one-upsmanship; nothing more. Actually, that's just the problem. You and your ilk have shown absolutely *zero* evidence for the existence of 'cable sound'. Heck, Tom has even been shafted by cable *manufacturers* who promised that they would show some evidence! -- Stewart Pinkerton | Music is Art - Audio is Engineering |
#269
|
|||
|
|||
Why do the anti-ABX folks not deliver?
Stewart Pinkerton wrote:
On Wed, 20 Aug 2003 16:53:05 GMT, Steven Sullivan wrote: Stewart Pinkerton wrote: On Wed, 20 Aug 2003 04:24:08 GMT, Steven Sullivan wrote: Mkuller wrote: So you tell us all - why do you and the other pro-DBTers persist in this endless debate? Proving you're right? Winning the debate? Getting in the last post? So tell us, Mike -- *who started this thread*? And *why*? Check the header, Steven! I started *this* thread, but it's an obvious rebuttal of Ludovic's 'why does ABX not deliver' attack thread. Yup, the thread I was referring to was the parent 'Why do the anti-ABX folks not deliver' thread, sorry. I do believe you just did it again! :-) I do believe you're right! Parent *of the* etc, is what I meant. Anyway, you get the idea -- replace 'anti-ABX' with 'DBT', and also see the forthcoming post documenting who started what threads about ABX. -- -S. |
#270
|
|||
|
|||
Why do the anti-ABX folks not deliver?
ludovic mirabel wrote:
Steven Sullivan wrote in message news:IJC0b.203356$o%2.94507@sccrnsc02... So tell us, Mike -- *who started this thread*? And *why*? Mr. Stewart Pinkerton on July 26th under this title. Direct the "why?" question to the right address. Youa re having a 'forest for the trees' problem, Ludovic. A perceptual error, perhaps. Allow me to help/ Here is the tree: Stewart spawned this thread off of 'Why DBTs in audio do not deliver' And here is the forest: a survey uniquely-named threads with 'ABX' in their subject lines since Jan 1 2002 (i.e, not including those spawned branches of Subject : X which are named Subject X, but including *renamed* threads that have 'ABX' in the new name) Summary: initial posts (either contained ABX in subject, or was later renamed to contain ABX in subject): 8 anti ABX 4 pro ABX 2 neutral about ABX 1 not about ABX renamed thread, containing 'ABX' 11 anti ABX 3 pro ABX data (in chronological order of parent thread; renamed threads are nested under their parents): // not about ABX From: islakit ) Subject: Help! Date: 2001-12-04 09:57:56 PST not about ABX renamed from Help! From: jim ) Subject: Snobbish replies (was Help!) View this article only Newsgroups: rec.audio.high-end Date: 2001-12-05 09:35:43 PST anti renamed from Snobbish replies From: Ernst Raedecker ) Subject: Hearing capabilities and ABX tests (was: Snobbish replies) Date: 2002-01-06 17:51:01 PST // anti From: ) Subject: testing ABX testing Date: 2002-01-11 13:42:36 PST pro renamed from testing ABX testing From: ) Subject: Categories of Preference (was testing ABX testing) Date: 2002-01-16 11:09:57 PST // anti From: Emerson Wood ) Subject: slower discussion of DBT/ABX Date: 2002-01-12 20:53:01 PST // anti From: ludovic mirabel ) Subject: ABX- THE NEW HORIZONS. Date: 2002-05-18 11:05:27 PST anti renamed from ABX- THE NEW HORIZONS. From: Mkuller ) Subject: ABX- THE VOCAL FEW Date: 2002-09-04 18:12:55 PST anti renamed from: ABX-THE VOCAL FEW From: Mkuller ) Subject: ABX- REMOVING MORE THAN SIGHTED BIAS Date: 2002-09-17 11:35:30 PST pro renamed from: ABX-THE NEW HORIZONS From: ) Subject: Abx and straw Date: 2002-06-04 12:50:26 PST // pro From: Darryl Miyaguchi ) Subject: Sequential ABX test Date: 2002-08-28 09:05:54 PST // anti From: ludovic mirabel ) Subject: HOW TO GET A POSITIVE ABX TEST. Date: 2002-09-24 11:05:34 PST anti renamed from HOW TO GET A POSITIVE ABX TEST. From: BEAR ) Subject: Now: HOW TO GET A LIST OF POSITIVE ABX TESTs. Date: 2002-11-14 10:15:21 PST anti renamed from: HOW TO GET A POSITIVE ABX TEST From: bear ) Subject: Now: WHO CAN GET A POSITIVE ABX TEST? Date: 2002-10-25 16:07:20 PST // pro From: Bruce Abrams ) Subject: DBT, ABX & Blind Reviewing Date: 2002-09-19 12:05:25 PST anti renamed from DBT, ABX and Blind REviewing From: BEAR ) Subject: DBT, ABX & Blind Reviewing - Upside down? Date: 2002-09-19 22:25:42 PST // pro From: Arny Krueger ) Subject: AES papers on CD, ABX, and 24/96 Date: 2002-10-17 11:07:46 PST // anti From: Harry Lavo ) Subject: Significance of Inadible High Frequencies to Musical Enjoyment and Relaxation Date: 2002-11-10 18:51:37 PST anti renamed from Signficance of Inaudible From: BEAR ) Subject: Significance of Inadible Now: non linear distortion & ABX Date: 2002-11-12 17:25:42 PST // anti From: BEAR ) Subject: WHY GET A POSITIVE ABX TEST --- summarized again Date: 2002-11-19 14:33:56 PST anti renamed from WHY GET A POSITIVE ABX TEST --- summarized again From: Mkuller ) Subject: POSITIVE ABX TEST with MUSIC Date: 2002-11-20 15:21:50 PST // anti From: BEAR ) Subject: Popper & ABX Date: 2002-12-03 14:22:02 PST anti renamed from Popper and ABX From: Mkuller ) Subject: ABX Date: 2002-12-08 11:42:05 PST anti From: ludovic mirabel ) Subject: Re Popper and ABX. Follow up to Ferstler's message. Date: 2002-12-27 21:52:42 PS // neutral From: Sven Berendsen ) Subject: Do we need the X in the ABX-Test? Date: 2002-12-14 10:00:57 PST // pro From: Mark Ovchain ) Subject: ABX Tests (and DBT's) are not "insensitive to music". Date: 2002-12-16 11:15:15 PST // neutral From: Juha Posti ) Subject: ABX'ing power cords? Date: 2003-04-07 08:50:21 PST // anti From: Ernst Raedecker ) Subject: Finally ... The Furutech CD-do-something Date: 2003-06-09 19:35:19 PST anti From: Ernst Raedecker ) Subject: Why DBTs in audio do not deliver (was: Finally ... The Furutech CD-do- something) Date: 2003-06-17 22:35:38 PST pro renamed from: Why DBTs in audio do not deliver From: Stewart Pinkerton ) Subject: Why do the anti-ABX folks not deliver? Date: 2003-07-25 14:35:05 PST // |
#271
|
|||
|
|||
Why do the anti-ABX folks not deliver?
(Mkuller) wrote:
(Nousaine) wrote: Actually it's people like you that continue the endless debate. If you had any evidence that you were right than the 'debate' would stop instantly. Uh, Tom, perhaps you aren't paying attention (or maybe it's just wishful thinking). Both sides have been providing "evidence" of their positions for years and no one on the other side has changed their mind or switched sides. The endless debate continues... It's mental masturbation, and one-upsmanship; nothing more. Regards, Mike This is simply not true. No subjectivist has supplied any bias-controlled evidence supporting amp or wire sound that confirms that a nominally competent device (in a room, drivign speakers) have any sound of their own. (Arny Krueger has supplied some evidence that minor differences may be audible with headphones.) What that side gives is unsupported anecdotes and when asked to veryify the sound have never been able to do so with even the simplest bias controls employed. On the other hand. there were 2 dozen+ bias controlled listening tests published by the early 90s. There have been documented tests on wires and capacitors but none that have verified wire sound. It's not a battle of conflicting evidence. The subjectivist (rather Golden-Ear) crowd has no evidence only anecdotes and conjecture. The "debate" rages because the subjectivists have no evidence and are reduced to argument alone. |
#272
|
|||
|
|||
Why do the anti-ABX folks not deliver?
|
#273
|
|||
|
|||
Why do the anti-ABX folks not deliver?
"Mkuller" wrote in message
news:BXb0b.183627$YN5.135766@sccrnsc01... (ludovic mirabel) wrote in message : 4) that reporting as the end-outcome the majority vote of ABX untrained/untalented and ignoring the capable minority who DO hear-exactly the way Greenhill had done (see Marcus quotes)is a perversion of "testing". (Bob Marcus) Date: 8/17/03 4:57 PM Pacific Daylight Time Message-id: wrote: "Ignoring the minority who DO have fewer and less severe colds after megadosing vitamin C is a perversion of testing." See how silly that argument is? So stop using it. Bob, your analogy comparing DBTs to medical research is wrongheaded. If one person in the Greenhill group, or any other group reliably identifies a difference - then it proves a difference exists! The fact that the others or a vast majority are not able to reliably identify it under THAT test and THOSE test conditions means next to nothing. It's a null - 0 - meaningless. That simply isn't true. All it proves is that the individual in question got results that can be expected only once out of 20 tries relying on guesswork alone--approximately 8 correct out of 10 tries. What conclusion would you draw if an individual got 8 out of 10 wrong! Norm Strong |
#274
|
|||
|
|||
Why do the anti-ABX folks not deliver?
(Nousaine) wrote in message .net...
(ludovic mirabel) wrote: (Nousaine) wrote in message news:tcX%a.173930$Ho3.23369@sccrnsc03... Let me first of all say that I truly appreciate your text. It is what I believe a discussion should be: with serious content, courteously written- not clowning for the audience, picking on single words to score points and listing your opponent's personal deficiences. I also do not doubt your good faith and considerable investment of time and money in this subject. But... Ludovic has posted earlier that he though that past controlled listening research had "hidden" positive scoring subjects with overall null results and that I had been avoiding him about that issue. I recently responded but my post was rejected. So here�_�¢�_�_�_�_s a straight reply to the "hidden audibility" claim. First, there have been no hidden positive subject scoring in any research that I�_�¢�_�_�_�_ve been able to note. In spite of the repeat claims that the Greenhill wire test had high scoring individuals whose results were concealed by averaging there is no data that either confirms or suggests this conclusion. This has been covered quite extensively in this newsgroup and in recent threads. I said: Let me shift the emphasis. It is not on individuals ( such as Greenhill's own "golden ear"). It is on the indisputable fact that the individual performances in the reported tests vary enormously, Nousaine answered: We'll of course. Individuals guessing will form a distribution with tals. Even when there is no difference a large sample or individuals retested there will be a distribution with tails toeither side. Exactly. You've said it. This is a "test" that invariably produces wildly divergent results between participants. This is a "test" that has no objective reference point to check individual results against. So how do you decide who is "right"? The only way that makes sense to me is to establish criterion of validity such as Greenhill's: For the result to count as positive the subject had to be right at least 75% of the time (In his case minimum 12 times out of 15 tries.). But oddly enough instead of acknowledging that the few who met their own criteria were "right"- ie. detected a difference- the proctors of those tests would reverse themselves in their conclusions and take the inept aggregate performance of the majority as the true outcome. And in this strange way they'd get null, negative result every time. Because most people- not all- are inept at ABXing and you have no evidence in each individual case how much of it is due to the nature of your "test". Worse: your subjects don't know either. Question: what is the practical usefulness to an audiophile of a "test" that has a "distribution with tails"? Not very repeatable, is it? Ludovic wants us to consider any subject who scores better much than 50% as a person who "heard" something whether the individual score or overall score suggests this was the case. Ludovic does not and never did hold such moronic views. I'm disappointed that you'd want to attribute them to me. Read above re "statistical validity criteria". On the contrary, Ludovic has been trying to explain to a Mr. Marcus for the last two years that adding up the bad, the average and the good together does not a "positive result" make. Unsuccessfully it would seem because he came back with it this very week. Example: Your witness Sean Olive's results in his "listening room" pdf (www.revelspeakers.- address quoted from memory). One of his supertrained professionals' was scoring around 30% , a few 50% or thereabouts, most reached (how many repeats?), significant 70% or more. OK? So what? Given a chance result we'de find a distribution and we'd find a buried Golden Ear according to Ludovic's reasoning. Why are there no lucky coins in his experiments? The correct guesses between 70 and 80% of a few of his subjects are golden enough, for his test tasks, no? Don't really follow your point. Greenhill: all scored brillianly ( almost all were 100% correct) recognising 1.75db. volume difference between a thick and a thin cable- as long as the pink noise was played. The same people failed to recognise (within Greenhill's statistical validity criteria- ie had much lower "correct" scores) the same volume difference once music replaced pink noise. You or others said that it only proves that pink noise is a better "test" signal. Sorry: doesn't wash with me. T Why not? Noise is the single most effective signals for revealing small audible differences The subject is reproduction of MUSIC by audio components- not pink noise. in fact saying that ABX performs better with pink noise than music, amounts to saying that ABX is an inappropriate test for assessing MUSICAL reproduction differences between components. Is a "test" that has such variable individual results an appropriate test for audiophile use? How can you be certain that the failure of most of your subjects to recognise differences was not due to the nature of your test? Especially as you must agree that training in ABXing improves individual results. How does an audiophile know if he had enough training? Next; of the tests I have personally designed, proctored or have been a subject there has been no cadre of �_�¢�_�_�_�_those who hear�_�¢�_�_�_�_ compared to the general subject population. IOW there has never been an experiment where a few significant scoring individuals were concealed by the average. Indeed with a reasonably sized sample often a single significant result would be enough to make the overall results positive. Of the tests I�_�¢�_�_�_�_ve designed and conducted subject count has varied from 1 to 31 with sample sizes ranging from 10 to 431. Programs were selected by subjects in the �_�¢�_�_�_�_challenge�_�¢� _�_�_�_ experiments where subjects had claimed to already have confirmed differences. Otherwise programs were selected from a collection of 63 musical and voice selections (including many with acoustical instrumentation) that had been chosen because they represent specific reproduction challenges to audio systems. Every experiment was extensively examined for internal significance. Most of them allowed subjects to extend the test with more trials to improve scores. This was seldom asked for and no individual improved his score enough to attain significance with additional trials OR a repeat of the entire experiment. There were NO individual significant scores in any experiment that wasn't significant overall. But ..... I'd have to know what you played to them, what were their ages and sex, their ABX training, their musical interests and exposure- a thousand details- before I could say that you ruled out with significant validity the possibility that you missed that rare bird. So exactly what are the characteristics of a 'rare bird'? What age group? What training? What musical interests? what thousand details ... can you list them? If not how am Is upposed toknow what meets your criteria? And how my subjects were not qualified. O.K. You're testing how different amps handle the cello sound. You get a significant sample (say 10) of experienced chamber music lovers, selected for their proven ABX training and ability. And by pot luck: I answered already: this shy, modest listener Mr. Pinkerton for one. Greenhill's golden ear for two. Keep trying. Since I believe that ABXing is a skill on its own I can't predict who'll be good at it. Krueger should have a few such by now. who was not only a great listener but also a great ABX performer. ONE is enough. It is not about pleasing the crowd. It is about "high-end". Something like Mr. Pinkerton- an ABX upholder whom you did not convince that there are no differences between amps. Nor , of course did he convince you. However in the case of Flying Blind (Audio Magazine) a subject who was unable to discern whether a given program contained a confirmed audible level of distortion in 16 weeks of long term listening was able to reliably identify same with a 6 second segment of that program using the ABX technique. But any ofd this notwithstanding I'm still not fully understanding why some individual, some company or some 3rd party has ever been able to confirm "amp" and "wire" sound under any set of conditions with even modest listening bias controls implemented. Even IF someone was withholding the Truth why hasn't someone else willfully or even by accident confirmed it? If these 'differences' are so evident that people like Ludovic are so certain of their existence WHY hasn't some party, interested or otherwise, stumbled across the body instead of just the rumors? It just doesn't seem likely that people as smart as Earl Geddes, Floyd Toole, David Clark, David Rich, Sean Olive, Stan Lip****z, John Vanderkooy, Dan Shanefiled, Rich Cabot, John Eargle, Dick Pierce, et al would have ALL overlooked even the minutest of evidence of true audibility. All your notables to the best of my knowledge did not report component comparisons- So you have to be controlled test or ABX qualified to exist? OK I'll buy that. What aew the qualifications of yourself or reference listeners? (I amended my answer below for lucidity L.M.) where there is NO OBJECTIVE END-POINT- but rather reasearch about perception of KNOWN induced artefact- the question and answer are straightforward-did he hear right or not. Usually by selected, trained audiences at that. Apples and oranges. Ludovic Mirabel Why don't you give us a list of researcjh that refutes what we already know about amps and wires? Because I have not a clue what the future research will bring. And if it does I may not understand it. My knowledge of electronics is very limited. I think it is up to you to prove that the individual ,human perceptions are directly related to electronic testing . I seem to recall that there still is a lot of controversy amongst the engineers abot what makes an amp. run. It used to be THD, then slew rates , then ability to adjust to low impedances, now other learned tests that are way above my head. I lost count. Ludovic Mirabel |
#275
|
|||
|
|||
Why do the anti-ABX folks not deliver?
mkuller wrote:
Both sides have been providing "evidence" of their positions for years and no one on the other side has changed their mind or switched sides. The endless debate continues... It's mental masturbation, and one-upsmanship; nothing more. Tom Nousaine wrote: This is simply not true. No subjectivist has supplied any bias-controlled evidence supporting amp or wire sound that confirms that a nominally competent device (in a room, drivign speakers) have any sound of their own. (Arny Krueger has supplied some evidence that minor differences may be audible with headphones.) What that side gives is unsupported anecdotes and when asked to veryify the sound have never been able to do so with even the simplest bias controls employed. On the other hand. there were 2 dozen+ bias controlled listening tests published by the early 90s. There have been documented tests on wires and capacitors but none that have verified wire sound. It's not a battle of conflicting evidence. The subjectivist (rather Golden-Ear) crowd has no evidence only anecdotes and conjecture. The "debate" rages because the subjectivists have no evidence and are reduced to argument alone. What you mean to say is "no evidence you will accept". I suppose you haven't presented sufficient evidence to the other side either, or they would be convinced. The only "bias control method" your side accepts is the DBT and over the years many have provided evidence here why DBTs with music are inappropriate and simply do not work to identify subtle audible differences. Your ignoring all of that doesn't make it go away. In fact, believing as you do - that virtually all equipment except speakers sound the same - would seem to disqualify one from claiming to be an audiophile, wouldn't it? Regards, Mike |
#276
|
|||
|
|||
Why do the anti-ABX folks not deliver?
|
#277
|
|||
|
|||
Why do the anti-ABX folks not deliver?
(Bob Marcus) wrote in message ...
(ludovic mirabel) wrote in message news:0He0b.184753$Ho3.25747@sccrnsc03... (Audio Guy) wrote in message news:G680b.182959$uu5.34042@sccrnsc04... Bob has demonstrated quite well on his own how your "interpretations" of Greenhill's comments are misinterpretations. So let's have no interpreting. Just poor Greenhill's own words, word for word and literally and Marcus own words, word for word and literally. Marcus first: Anyone tempted to take Mirabel's assertion here at face value should consult the original article and read for himself that Greenhill said no such thing. bob bob This is how things went in this thread. Nousaine had said that "NO ONE has ever produced a single experiment that confirms "amplifier" or "wire" sound with even modest bias controls implemented. NO One' Ever!!!!.... ...Deliver some evidence to support your argument and then we can talk' actually IF you deliver the evidence than we won't 'talk' I'll admit you were right. Do it!!!!! Now !!!!! I said in reply : According to your ABX coworker L. Greenhill (Stereo Review August 1982) one of his subjects did that very thing. So much for "NO one! Ever!!!!!" It is your privilege to disagree with Greenhill- you two ABXers fight it out. On Aug 14 (message 62 in Google) -QUOTING THE ABOVE EXCHANGE : (see footnote for requote) Marcus commented: "Elsewhere today you tell us you are a published writer. Fiction a specialty? Anyone tempted to take Mirabel's assertion here at face value should consult the original article and read for himself that Greenhill said no such thing." Thereupon I quoted Greenhill's words, word for word: ("The Stereo Review, August 1982): "Final significant conclusion one can draw is that at least one genuine "golden ear" exists. Obviously certain listeners whether through talent, training or experience can hear small differences between components" (ie. cables in this cable comparison test) And got this reply from Marcus Apparently the only way you can argue your case is to juxtapose my words with something I was not responding to (and which did not even appear in the post you pulled my comment from). A rather pitiful act of desperation, I'd say. As on many previous occasions one wonders if he takes his readers for one of the more moronic jurymen to be befuddled with minimum effort or if he really believes what he's saying. The second possibility is more to his credit.. . I think. Ludovic Mirabel ** " Isn't that funny. No one can ever "prove a negative". So what? We should then ignore the fact that NO ONE has ever produced a single experiment that confirms "amplifier" or "wire" sound with even modest bias controls implemented. NO One' Ever!!!! |
#278
|
|||
|
|||
Why do the anti-ABX folks not deliver?
(Bob Marcus) wrote in message ...
(ludovic mirabel) wrote in message news:0He0b.184753$Ho3.25747@sccrnsc03... (Audio Guy) wrote in message news:G680b.182959$uu5.34042@sccrnsc04... Bob has demonstrated quite well on his own how your "interpretations" of Greenhill's comments are misinterpretations. So let's have no interpreting. Just poor Greenhill's own words, word for word and literally and Marcus own words, word for word and literally. Marcus first: Anyone tempted to take Mirabel's assertion here at face value should consult the original article and read for himself that Greenhill said no such thing. bob bob This is how things went in this thread. Nousaine had said that "NO ONE has ever produced a single experiment that confirms "amplifier" or "wire" sound with even modest bias controls implemented. NO One' Ever!!!!.... ...Deliver some evidence to support your argument and then we can talk' actually IF you deliver the evidence than we won't 'talk' I'll admit you were right. Do it!!!!! Now !!!!! I said in reply : According to your ABX coworker L. Greenhill (Stereo Review August 1982) one of his subjects did that very thing. So much for "NO one! Ever!!!!!" It is your privilege to disagree with Greenhill- you two ABXers fight it out. On Aug 14 (message 62 in Google) -QUOTING THE ABOVE EXCHANGE : (see footnote for requote) Marcus commented: "Elsewhere today you tell us you are a published writer. Fiction a specialty? Anyone tempted to take Mirabel's assertion here at face value should consult the original article and read for himself that Greenhill said no such thing." Thereupon I quoted Greenhill's words, word for word: ("The Stereo Review, August 1982): "Final significant conclusion one can draw is that at least one genuine "golden ear" exists. Obviously certain listeners whether through talent, training or experience can hear small differences between components" (ie. cables in this cable comparison test) And got this reply from Marcus Apparently the only way you can argue your case is to juxtapose my words with something I was not responding to (and which did not even appear in the post you pulled my comment from). A rather pitiful act of desperation, I'd say. As on many previous occasions one wonders if he takes his readers for one of the more moronic jurymen to be befuddled with minimum effort or if he really believes what he's saying. The second possibility is more to his credit.. . I think. Ludovic Mirabel ** " Isn't that funny. No one can ever "prove a negative". So what? We should then ignore the fact that NO ONE has ever produced a single experiment that confirms "amplifier" or "wire" sound with even modest bias controls implemented. NO One' Ever!!!! |
#279
|
|||
|
|||
Why do the anti-ABX folks not deliver?
(ludovic mirabel) wrote:
...snips... Let me shift the emphasis. It is not on individuals ( such as Greenhill's own "golden ear"). It is on the indisputable fact that the individual performances in the reported tests vary enormously, Nousaine answered: We'll of course. Individuals guessing will form a distribution with tals. Even when there is no difference a large sample or individuals retested there will be a distribution with tails toeither side. Exactly. You've said it. This is a "test" that invariably produces wildly divergent results between participants. This is a "test" that has no objective reference point to check individual results against. It doesn't need an objective reference point. It only needs to confirm that subjects had reliable scores that showed they heard a difference. Isn't that your position.....common objective measurements fail to reveal real audible differences? So how do you decide who is "right"? The only way that makes sense to me is to establish criterion of validity such as Greenhill's: For the result to count as positive the subject had to be right at least 75% of the time (In his case minimum 12 times out of 15 tries.). But oddly enough instead of acknowledging that the few who met their own criteria were "right"- ie. detected a difference- the proctors of those tests would reverse themselves in their conclusions and take the inept aggregate performance of the majority as the true outcome. This is certainly not true. There were no individuals in that test that reliably demonstrated an ability to differentiate those wires with music. But if you repeated this test endlessly comparing the same wire to itself you'd still get the random distribution of results. And even if you used the same subjects in a million trials you'd find the distribution of scores but NOT a pattern based on individual performances. And in this strange way they'd get null, negative result every time. Not if there were truly audible differences. If you take your argument to its logical end you would argue that we should only "count" results that conform with your previously held beliefs. Because most people- not all- are inept at ABXing and you have no evidence in each individual case how much of it is due to the nature of your "test". Actually when real differences are present scores quickly become near-universally positive. To expand, when audibility is truly present there is still a distribution BUT most subjects score positively on individual scores. Worse: your subjects don't know either. Question: what is the practical usefulness to an audiophile of a "test" that has a "distribution with tails"? Not very repeatable, is it? Oh yes it IS very repeatable. If there is no audbile difference then you'll get a random distribution. If there is a true difference then you get a distribution that is not random. Both may have tails. Let me give you an example. I helped a friend set-up a blind test that compared two devices the person believed were sonically identical. I auditioned the comparison as set-up and he said "they sound alike, do they not?" I said I think they sound different. He then handed me the remote control for the QSC ABX comparitor. I then ran 3 sessions and scored 21/23, 17/19, 16/20 showing that they did sound different. He, after pulling hair he no longer had, began investigating and discovered that although he had level matched the devices at the black-box output he hadn't level-matched at the speaker terminals and there was a level difference. A level match at the speaker terminals delivered sound that I could no longer reliably identify. The distribution of my scores fell to between 30 and 60 % correct, which one would expect with a non-audible difference. Ludovic wants us to consider any subject who scores better much than 50% as a person who "heard" something whether the individual score or overall score suggests this was the case. Ludovic does not and never did hold such moronic views. I'm disappointed that you'd want to attribute them to me. As much as you claim otherwise what you have specifically asked for is that we accept non-significant, but seemingly high, results on individual subjects as real data "hidden" in the averages. This hasn't ever happened and as far as I can tell has not shown up in any published experimental results. Read above re "statistical validity criteria". On the contrary, Ludovic has been trying to explain to a Mr. Marcus for the last two years that adding up the bad, the average and the good together does not a "positive result" make. Unsuccessfully it would seem because he came back with it this very week. So you would have us only 'count' high scores even if they fall within a distribution that we'd get when people were only guessing. Example: Your witness Sean Olive's results in his "listening room" pdf (www.revelspeakers.- address quoted from memory). One of his supertrained professionals' was scoring around 30% , a few 50% or thereabouts, most reached (how many repeats?), significant 70% or more. OK? So what? Given a chance result we'de find a distribution and we'd find a buried Golden Ear according to Ludovic's reasoning. Why are there no lucky coins in his experiments? The correct guesses between 70 and 80% of a few of his subjects are golden enough, for his test tasks, no? Don't really follow your point. That's because you don't understand experimental design and analysis. No; 70 to 80% correct will be found in many distributions. So will 20 and 30% scores. You seemingly only want to 'count' those scores that will support your previously held biases. Greenhill: all scored brillianly ( almost all were 100% correct) recognising 1.75db. volume difference between a thick and a thin cable- as long as the pink noise was played. The same people failed to recognise (within Greenhill's statistical validity criteria- ie had much lower "correct" scores) the same volume difference once music replaced pink noise. You or others said that it only proves that pink noise is a better "test" signal. Sorry: doesn't wash with me. Why not? It's pretty obvious from this and other experiments. T Why not? Noise is the single most effective signals for revealing small audible differences The subject is reproduction of MUSIC by audio components- not pink noise. in fact saying that ABX performs better with pink noise than music, amounts to saying that ABX is an inappropriate test for assessing MUSICAL reproduction differences between components. No sir. But the logical fallout from those results is that noise is 'overly-sensitive' to audible difference. And it IS compared to music. Is a "test" that has such variable individual results an appropriate test for audiophile use? How can you be certain that the failure of most of your subjects to recognise differences was not due to the nature of your test? The "test" has the same conditions under which people happily form subjective difference opinions. The "sound" doesn't change only the knowledge over which component is in the chain or is "X" is withheld. It is true that it's much harder to identify which of 2 identical sounding products is driving the loudspeakers when the blindfolds come out. So? None of that changes the "sound being reproduced. If the ability to distinguish an amplifier under blindfolded conditions (either figurately or literally) disappears when bias controls are introduced then it cannot be said that the device sounds "different." If the subject then insists that the devices were sonically different the conclusion can ONLY be attributed to listener bias. Especially as you must agree that training in ABXing improves individual results. How does an audiophile know if he had enough training? When he can 'hear' inaudible differences? Ludovic why not try an ABX test yourself? What you'll find is that 1-2-minutes of "training" is entirely adequate. Have you actually ever taken an ABX or other controlled listening test? pcabx for example? If the answer is "no" I'd say you fall into the same corner that 'audiophiles' always try to back skeptics into. "Have you heard this component? If not HOW DARE YOU question my evaluation?" Next; of the tests I have personally designed, proctored or have been a subject there has been no cadre of Ã?_Ã?ÀšÃ‚¢Ãƒ?_Ã?_Ã?_Ã?_those who hearÃ?_Ã?ÀšÃ‚¢Ãƒ?_Ã?_Ã?_Ã? _ compared to the general subject population. IOW there has never been an experiment where a few significant scoring individuals were concealed by the average. Indeed with a reasonably sized sample often a single significant result would be enough to make the overall results positive. Of the tests IÃ?_Ã?ÀšÃ‚¢Ãƒ?_Ã?_Ã?_Ã?_ve designed and conducted subject count has varied from 1 to 31 with sample sizes ranging from 10 to 431. Programs were selected by subjects in the Ã?_Ã?ÀšÃ‚¢Ãƒ?_Ã?_Ã?_Ã?_challe ngeÃ?_Ã?ÀšÃ‚¢à ƒÆ’?_Ã?_Ã?_Ã?_ experiments where subjects had claimed to already have confirmed differences. Otherwise programs were selected from a collection of 63 musical and voice selections (including many with acoustical instrumentation) that had been chosen because they represent specific reproduction challenges to audio systems. Every experiment was extensively examined for internal significance. Most of them allowed subjects to extend the test with more trials to improve scores. This was seldom asked for and no individual improved his score enough to attain significance with additional trials OR a repeat of the entire experiment. There were NO individual significant scores in any experiment that wasn't significant overall. But ..... I'd have to know what you played to them, what were their ages and sex, their ABX training, their musical interests and exposure- a thousand details- before I could say that you ruled out with significant validity the possibility that you missed that rare bird. So exactly what are the characteristics of a 'rare bird'? What age group? What training? What musical interests? what thousand details ... can you list them? If not how am Is upposed toknow what meets your criteria? And how my subjects were not qualified. O.K. You're testing how different amps handle the cello sound. You get a significant sample (say 10) of experienced chamber music lovers, selected for their proven ABX training and ability. And by pot luck: I answered already: this shy, modest listener Mr. Pinkerton for one. Greenhill's golden ear for two. Keep trying. Since I believe that ABXing is a skill on its own I can't predict who'll be good at it. Krueger should have a few such by now. who was not only a great listener but also a great ABX performer. So how does a person who has not demonstrated and ability to 'hear' inaudible differences qualify under your criteria. All of my ABX subjects have been great listeners and great ABX performers. You want "ABX significant scores" on components that sound the same? Those people don't exist; but are still "great" subjects. ONE is enough. It is not about pleasing the crowd. It is about "high-end". Pinkerton is right "high-end" only means high-price, high-margin and often sub-par performance. Something like Mr. Pinkerton- an ABX upholder whom you did not convince that there are no differences between amps. Nor , of course did he convince you. All you or Mr Pinkerton has to do is deliver some corroborating evidence. However in the case of Flying Blind (Audio Magazine) a subject who was unable to discern whether a given program contained a confirmed audible level of distortion in 16 weeks of long term listening was able to reliably identify same with a 6 second segment of that program using the ABX technique. But any ofd this notwithstanding I'm still not fully understanding why some individual, some company or some 3rd party has ever been able to confirm "amp" and "wire" sound under any set of conditions with even modest listening bias controls implemented. Even IF someone was withholding the Truth why hasn't someone else willfully or even by accident confirmed it? If these 'differences' are so evident that people like Ludovic are so certain of their existence WHY hasn't some party, interested or otherwise, stumbled across the body instead of just the rumors? It just doesn't seem likely that people as smart as Earl Geddes, Floyd Toole, David Clark, David Rich, Sean Olive, Stan Lip****z, John Vanderkooy, Dan Shanefiled, Rich Cabot, John Eargle, Dick Pierce, et al would have ALL overlooked even the minutest of evidence of true audibility. All your notables to the best of my knowledge did not report component comparisons- So you have to be controlled test or ABX qualified to exist? OK I'll buy that. What aew the qualifications of yourself or reference listeners? (I amended my answer below for lucidity L.M.) where there is NO OBJECTIVE END-POINT- but rather reasearch about perception of KNOWN induced artefact- the question and answer are straightforward-did he hear right or not. Usually by selected, trained audiences at that. Apples and oranges. Ludovic Mirabel Why don't you give us a list of researcjh that refutes what we already know about amps and wires? Because I have not a clue what the future research will bring. And if it does I may not understand it. My knowledge of electronics is very limited. I think it is up to you to prove that the individual ,human perceptions are directly related to electronic testing . Why is that incumbent on me? I've only tested the subjective quality of the SOUND that has been attributed to certain components. No electronic testing other than verifying level matchin and flat response into the load, has ever been invollved. I seem to recall that there still is a lot of controversy amongst the engineers abot what makes an amp. run. It used to be THD, then slew rates , then ability to adjust to low impedances, now other learned tests that are way above my head. I lost count. Ludovic Mirabel Grow up. There is no controversey amongst anybody except the engineering/ marketing departments of "high-end" companies. The rest of us know what's true and, except for arguing with people like you, don't waste time, energy or dollars on things that have been shown to not contribute to sound quality improvements. |
#280
|
|||
|
|||
Why do the anti-ABX folks not deliver?
Nousaine wrote:
Steven Sullivan wrote: Nousaine wrote: Holy Cow: do I have to make the observation in full detail every time? Are your claims not based on a normally reverberant environment ? Always meant to ask (hope I haven't asked before) does this exclude headphone listening? -- -S. Doesn't have to. But I know of no one who would claim that his Pass Aleph sounds better on headphones only. The normally reverberant environment is only important in end-usage. If pcabx can elicit subtle differences between a Paraound and Bryston amplifier that can be heard with specially recorded programs in .wav files played back over headphones that's very interesting. But when those same two amplifiers turn out to be indistinguishable from each other when using loudspeakers in a listening room what does that say about the size and the importance of those differences? Not much, but if nothing else this give 'difference' lovers a way to save face -- they can always hope that their amp will sound different under the microscope of headphone listening. -- -S. |
Reply |
Thread Tools | |
Display Modes | |
|
|
Similar Threads | ||||
Thread | Forum | |||
Crazy market saturation! | Car Audio | |||
FAQ: RAM LISTING OF SCAMMERS, SLAMMERS, AND N'EER DO WELLS! V. 8.1 | Audio Opinions | |||
A quick study in very recent RAHE moderator inconsistency | Audio Opinions | |||
System balance for LP? | Audio Opinions | |||
gps install: how to mix its audio (voice prompting) with head unit audio-out? | Car Audio |