Home |
Search |
Today's Posts |
#201
|
|||
|
|||
THE ENDLESS DBT DEBATE
Bruce Abrams wrote in message ...
Snip discussion The fact of the matter is that in this test, 2 CD players were evaluated under blind, level matched conditions and found to sound different. This DBT did, in fact, yield a difference between two components. Are you listening Mr. Mirabel? Yes, I am but you're leaving me somewhat puzzled as to what exactly should I listen to. The startling news that trained listeners can hear differences even when ABXing? Who ever said otherwise? Not me. As a matter of fact when I reported that one (ONE!!!) of Greenhill's "expert audiophiles" in a cable listening test was called by the proctor, Greenhill, a "golden ear" because of his brilliant performance that met Greenhill's statistical validity criteria there was a flurry of indignation. With me , the messenger, not Greenhill the source. (Look up the old "Is ABX useful?" thread). Why? Because everybody KNOWS that cables all sound the same, right? Of course talented/trained people will hear differences even when ABXing because the differences are there. Something to clarify, Mr Abrams. And not for the first time. You really must stop confusing ABX/ DBT with other more or less sensible methods. Like this: " The fact of the matter is that in this test, 2 CD players were evaluated under blind, level matched conditions and found to sound different." Last time we talked about it it turned out that you confuse a simple, sensible way of covering the brand name when picking your pianos with switched ABX/DBT for component comparison. A horse of different colour Mr. Abrams. Cover the brand names all you want Mr. Abrams. Pick your piano. As long as you don't tell me that your choice is "scientific" and universally valid. We're talking here about the assertion that ABX DBT is THE one scientific way for an average audiophile to test components for their music reproduction abilities and that if he hears no differences it means more than just that: HE heard no differences. Ludovic Mirabel |
#203
|
|||
|
|||
THE AUDIOPHILE PRESS
"Nousaine" wrote in message news:
OK; perhaps I'm slow, why not list the results again and the measurement method? Stability looks good then. What was the frequency response at the speaker terminals? Dennis I don't know Tom, you seem more intent on picking sides and playing with debating tricks. Alright, another repeat performance. Okay Tom reread the above again. I previously gave you one brand. And the results for example. Which were beyond the +/- .1dB from 20-20,000 you mention. And yes, the measuring was done at the speaker terminal. On Soundlabs to refresh your memory. An odd somewhat tough load of course. But since they are the speakers I have, the response with them is what I would care about naturally. Since you repeat questions answered already, what is your point? That you aren't listening to anyone who might contradict "any 2 ohm stable 200 watt amp is fine". An Adcom 555 measured +.6 dB/-.4 dB at the speaker terminal. It was up at the low end, no surprise with an ESL. Then had a dip between 10-15Khz with a peak also of .6 dB between 16-18khz. Dropping to -.4dB again around 20khz. Not terrible, but not what would could call fully accurate. Some B+K amps were more even until the last couple of octaves, where it had some up and down variations of about .5dB up and .5dB down. A Classe 25 was about .25 dB up at the low end, and .3 dB down at 20 khz. With the variation evenly spread over the whole range. Probably good enough. Though strictly speaking doesn't meet your criteria. A Spectral DMA-50 was about ..15 dB up at 20 hz. Even by about 50 hz, and didn't change through 20 khz. All measurement done at 10 volts 1 khz for the reference level. This should more than satisfy your questions. Then of course you make statements implying I was arguing with your criteria of -/+ .1dB over the audible range. When I wasn't. I was wondering about the seemingly cavalier attitude of assuming what would be fully transparent. Dennis |
#204
|
|||
|
|||
THE AUDIOPHILE PRESS
|
#205
|
|||
|
|||
THE AUDIOPHILE PRESS
"Stewart Pinkerton" wrote in message
... On Tue, 16 Sep 2003 23:35:23 GMT, "Dennis Moore" wrote: An Adcom 555 measured +.6 dB/-.4 dB at the speaker terminal. It was up at the low end, no surprise with an ESL. Then had a dip between 10-15Khz with a peak also of .6 dB between 16-18khz. Dropping to -.4dB again around 20khz. Not terrible, but not what would could call fully accurate. Some B+K amps were more even until the last couple of octaves, where it had some up and down variations of about .5dB up and .5dB down. A Classe 25 was about .25 dB up at the low end, and .3 dB down at 20 khz. With the variation evenly spread over the whole range. Probably good enough. Though strictly speaking doesn't meet your criteria. A Spectral DMA-50 was about .15 dB up at 20 hz. Even by about 50 hz, and didn't change through 20 khz. All measurement done at 10 volts 1 khz for the reference level. This should more than satisfy your questions. Then of course you make statements implying I was arguing with your criteria of -/+ .1dB over the audible range. When I wasn't. I was wondering about the seemingly cavalier attitude of assuming what would be fully transparent. Well, given that the response of your speakers to a *perfect* amplifier would still be up and down by at least 5dB over the 20-20kHz range, I suspect that all of the above amps would be sonically indistinguishable and 'fully transparent'. -- Stewart Pinkerton | Music is Art - Audio is Engineering So since you don't need flat amps without flat speakers, how flat do they need be? And exactly what is the algorithm you go through to decide this issue other than blind testing? If the frequency response variation had been plus or minus 1 dB would you have responded the same? If it had been plus or minus 1.5dB? Where would you draw the line? If unflat speakers make it possible to considerably loosen up the flatness of frequency response needed for transparency then how useful are most of the tests done in a bias-controlled manner since most speakers are going to be less flat than those I have? And while were at it, some tube amps have an output impedance that interacts with most esl's to flatten the frequency response acoustically versus general solid state amps. Which contrary to what is often stated would make a good tube amp more accurate than some of these solid state ones in some uses. BTW, measures with warble tones indicate the speakers are not up and down by 5 decibels. But that would get us into another topic of how do you measure speaker's acoustic output. Dennis |
#206
|
|||
|
|||
THE AUDIOPHILE PRESS
Dennis Moore wrote:
And while were at it, some tube amps have an output impedance that interacts with most esl's to flatten the frequency response acoustically versus general solid state amps. Which contrary to what is often stated would make a good tube amp more accurate than some of these solid state ones in some uses. You've got to be more specific than 'some' and 'most' for that to be really true. Any speaker designer with good sense doesn't design his speaker to match certain amplifiers, given the wide variety of amplifiers that owners use. As I recall, ESL's almost always use some sort of equalization network to help flatten the frequency response anyway because of the natural high resonance of the diaphagm. |
#207
|
|||
|
|||
THE AUDIOPHILE PRESS
"Nousaine" wrote in message
... "Dennis Moore" wrote: "Nousaine" wrote in message news: OK; perhaps I'm slow, why not list the results again and the measurement method? Stability looks good then. What was the frequency response at the speaker terminals? Dennis I don't know Tom, you seem more intent on picking sides and playing with debating tricks. Alright, another repeat performance. Okay Tom reread the above again. I previously gave you one brand. And the results for example. Which were beyond the +/- .1dB from 20-20,000 you mention. And yes, the measuring was done at the speaker terminal. On Soundlabs to refresh your memory. An odd somewhat tough load of course. But since they are the speakers I have, the response with them is what I would care about naturally. Since you repeat questions answered already, what is your point? That you aren't listening to anyone who might contradict "any 2 ohm stable 200 watt amp is fine". An Adcom 555 measured +.6 dB/-.4 dB at the speaker terminal. It was up at the low end, no surprise with an ESL. Then had a dip between 10-15Khz with a peak also of .6 dB between 16-18khz. Dropping to -.4dB again around 20khz. Not terrible, but not what would could call fully accurate. Some B+K amps were more even until the last couple of octaves, where it had some up and down variations of about .5dB up and .5dB down. A Classe 25 was about .25 dB up at the low end, and .3 dB down at 20 khz. With the variation evenly spread over the whole range. Probably good enough. Though strictly speaking doesn't meet your criteria. A Spectral DMA-50 was about .15 dB up at 20 hz. Even by about 50 hz, and didn't change through 20 khz. All measurement done at 10 volts 1 khz for the reference level. This should more than satisfy your questions. Then of course you make statements implying I was arguing with your criteria of -/+ .1dB over the audible range. When I wasn't. I was wondering about the seemingly cavalier attitude of assuming what would be fully transparent. Dennis Thank you. Are you then saying that these amplifiers were identifiable in bias-controlled listening tests? If they were then my criteria still fits well. If they were NOT then the +/- 0.1 dB criteria is too stringent. OK, I'll agree with that. Which puts us right back to the beginning again. The question was which amplifiers could I count on to be sonically transparent. So that no aggravating bias controlled tests are needed. And you told me one thing which I showed through the simplest of measurements in fact couldn't guarantee transparency. So now you change your criteria wondering if they really need to be flat with a tenth of a decibel either way. Yet if someone had posted some positive results with less flatness you would have criticized those results. But in answer to your question above, I am saying that the standard for making a sure an amplifier isn't identifiable has been stated as being flat within plus or minus .1 dB. And some very normal competent amplifiers don't meet that criteria. Which as again takes us back in a big circle to the beginning. Dennis |
#208
|
|||
|
|||
THE AUDIOPHILE PRESS
"Dennis Moore" wrote:
"Nousaine" wrote in message ... "Dennis Moore" wrote: "Nousaine" wrote in message news: OK; perhaps I'm slow, why not list the results again and the measurement method? Stability looks good then. What was the frequency response at the speaker terminals? Dennis I don't know Tom, you seem more intent on picking sides and playing with debating tricks. Alright, another repeat performance. Okay Tom reread the above again. I previously gave you one brand. And the results for example. Which were beyond the +/- .1dB from 20-20,000 you mention. And yes, the measuring was done at the speaker terminal. On Soundlabs to refresh your memory. An odd somewhat tough load of course. But since they are the speakers I have, the response with them is what I would care about naturally. Since you repeat questions answered already, what is your point? That you aren't listening to anyone who might contradict "any 2 ohm stable 200 watt amp is fine". An Adcom 555 measured +.6 dB/-.4 dB at the speaker terminal. It was up at the low end, no surprise with an ESL. Then had a dip between 10-15Khz with a peak also of .6 dB between 16-18khz. Dropping to -.4dB again around 20khz. Not terrible, but not what would could call fully accurate. Some B+K amps were more even until the last couple of octaves, where it had some up and down variations of about .5dB up and .5dB down. A Classe 25 was about .25 dB up at the low end, and .3 dB down at 20 khz. With the variation evenly spread over the whole range. Probably good enough. Though strictly speaking doesn't meet your criteria. A Spectral DMA-50 was about .15 dB up at 20 hz. Even by about 50 hz, and didn't change through 20 khz. All measurement done at 10 volts 1 khz for the reference level. This should more than satisfy your questions. Then of course you make statements implying I was arguing with your criteria of -/+ .1dB over the audible range. When I wasn't. I was wondering about the seemingly cavalier attitude of assuming what would be fully transparent. Dennis Thank you. Are you then saying that these amplifiers were identifiable in bias-controlled listening tests? If they were then my criteria still fits well. If they were NOT then the +/- 0.1 dB criteria is too stringent. OK, I'll agree with that. Which puts us right back to the beginning again. The question was which amplifiers could I count on to be sonically transparent. So that no aggravating bias controlled tests are needed. And you told me one thing which I showed through the simplest of measurements in fact couldn't guarantee transparency. So now you change your criteria wondering if they really need to be flat with a tenth of a decibel either way. Yet if someone had posted some positive results with less flatness you would have criticized those results. But in answer to your question above, I am saying that the standard for making a sure an amplifier isn't identifiable has been stated as being flat within plus or minus .1 dB. And some very normal competent amplifiers don't meet that criteria. Which as again takes us back in a big circle to the beginning. Dennis You must have incredibly bad luck. I have purchased and used over a dozen solid state amplifiers over the past 25 years all of which meet the criteria. OTOH I'd also say that IME amplifiers which meet that criteria over the 100 Hz to 10,000 Hz passband will be sonically indistinguishable from each other. The +/- 0.1 criteria is probably more stringent than it needs to be. Also note that in amplifier comparisons amplifiers matched to +/- 0.1 dB will not be sonically identifiable but they may often both have a couple tenths droop at 20 kHz. |
#209
|
|||
|
|||
THE ENDLESS DBT DEBATE
(Mkuller) wrote in message ...
On 8 Sep 2003 22:59:18 GMT, (Mkuller) wrote: I have no intent in engaging in another dbt debate (or in continuing one) with you. I have the following to say about DBT in audio, and that is that since we have to rely on memeory, and memory is not exactly reliable, DBT is in fact impossible, as you cannot compare something perceived directly now with something remembered, accurately. I usually use the 'take it away' criterion. When I auditioned Monster interconnect cables, I compared the $50 ones to the $100 ones. When I put in the $100 ones to replace the $50 ones, I could not be sure I heard a difference. When I put back the cheaper ones, I was certain I did. I repeated this test a number of times, always with the same results: the 'worse' was easier to spot when the 'better' was taken away than when the 'worse' was taken away and the 'better' installed. So, in sum: THERE IS NO WAY TO DO DBT IN AUDIO. IT'S A MYTH. (Stewart Pinkerton) Date: 9/9/03 8:51 AM Pacific Daylight Time Message-id: wrote: I'm sure that you don't. You prefer to assert and run, as you have no basis for your arguments. Neither you nor anyone else has produced evidence that dbts, even if implemented perfectly (which is unlikely for the average audiophile) can show subtle differences in audio components using music as the program. OTOH, many people have shown that sighted bias totally swamps any subtle difference which *may* exist between components, while many major manufacturers use DBTs every day in their research labs because they are not interested in this pathetic 'debate' - they already *know* that DBTs are the *only* way to determine subtle but *real* sonic differences. Yes, we have heard you say this many times - repeating it over and over does not make it true. Please provide pecific evidence that: 1. dbts are appropriate for audiophiles to use in audio component comparison with music as a source. 2. dbts used in #1 above can show subtle audible differences. So far all we have seen actual proof of is that dbts with pink noise can sometimes show gross frequency response and loudness differences when used with audio components. That's it. If you are unsure about what constitutes "actual proof", please refer to the beginning of the "science vs. pseudo-science" thread for the distinction between the two. Subtle differences will disappear and the components will sound more alike than different. Evidence, please, for this *claim*. It should be obvious to anyone who has tried a dbt, used sighted listening and followed the "dbt debates that never end on RAHE". Regards, Mike |
#210
|
|||
|
|||
THE ENDLESS DBT DEBATE
Michael Scarpitti wrote:
snip I have the following to say about DBT in audio, and that is that since we have to rely on memeory, and memory is not exactly reliable, DBT is in fact impossible, as you cannot compare something perceived directly now with something remembered, accurately. Which *IF* accurate, applies equally to any test, sighted or DBT. Hence the use of time proximate switching to minimize such effects. I usually use the 'take it away' criterion. When I auditioned Monster interconnect cables, I compared the $50 ones to the $100 ones. When I put in the $100 ones to replace the $50 ones, I could not be sure I heard a difference. When I put back the cheaper ones, I was certain I did. I repeated this test a number of times, always with the same results: the 'worse' was easier to spot when the 'better' was taken away than when the 'worse' was taken away and the 'better' installed. So, in sum: THERE IS NO WAY TO DO DBT IN AUDIO. IT'S A MYTH. Uhmmm...all of which required you to remember how one cable sounded, some time ago, and mentally compare what you currently hear with that memory. Oh, wait, you yourself said that was impossible, so obviously your cable test was a MYTH by your own admission. Actually, IMO, the myth is that knowing which cable was "worse" (i.e. cheaper) did not affect the test outcome. snip Keith Hughes |
#211
|
|||
|
|||
THE ENDLESS DBT DEBATE
"Harry Lavo" wrote in message . net...
"Stewart Pinkerton" wrote in message ... On Thu, 11 Sep 2003 03:49:01 GMT, "Harry Lavo" wrote: "Steven Sullivan" wrote in message news:uPJ7b.406577$YN5.274362@sccrnsc01... Mkuller wrote: Please provide pecific evidence that: 1. dbts are appropriate for audiophiles to use in audio component comparison with music as a source. One bit of evidence would be that they are used in prodect development by several major audio component manufacturers, as Stewart has noted. Not necessarily. The research is likely to be very directed, e.g. can you hear a difference in the lower midrange, or some other indication of what you are listening *for*. Which speaker, A or B, do you think reproduces her voice best, etc. That is very different than listening to two pieces of equipment and determining how and why they differ in reproducting music on an open-ended basis. Excuse me? The 'subjectivists' have *frequently* said that they choose equipment based on such narrow criteria as listening to a spoken voice, or determining whether a particular singer sounds the same as their memory of a live performance. I can't speak for all audiophiles, but most that I know listen open ended at first, and then over a long period of time alternate open ended listening with specific comparisons designed to illuminate the differences that strike them from the open ended listening. That is a very different situation from being given a very one-dimensional and single-minded task....compare these two things on this one dimension and tell us...... This is indeed the crux of the problem with DBT. Music is complex, and there are many, many, many, many, many, many, many, ways things that consitute the 'listening experience'. What particular aspect of any two items may be different may not be obvious upon simple DBT, but over a period of time, with switching in and out of the variable component, using lots of different music selections, etc, differences will appear. Also, it is impossible to listen to two components at the same time. This differs from say, lens tests, where two prints from two different lenses can be viewed simultaneously and judgements made without relying upon memory. The 'optical image' can be fixed on film, but the 'audio image' cannot. I can hand you two prints made with say, a Leica lens and a Nikon lens, and you can see them together and examine them as long as you want. Yes, optical goods can be evaluated by a DBT, but not audio goods. DBT in audio is not invalid, it is impossible. For one thing, the audiophile can select those aspects of musical reproduction that are most important to him..and ignore others. In gestalt listening, this will be an intuitive, almost unthinking weighting of preferences. Then the specific comparisons can begin.... And many to reach a decision, not just one. And it all takes time. With different music as test input. In different moods, ideally. And all within a known system, unless one is starting with a blank slate. |
#212
|
|||
|
|||
THE ENDLESS DBT DEBATE
On Thu, 02 Oct 2003 22:31:14 GMT, (Michael
Scarpitti) wrote: (Mkuller) wrote in message ... On 8 Sep 2003 22:59:18 GMT, (Mkuller) wrote: I have no intent in engaging in another dbt debate (or in continuing one) with you. I have the following to say about DBT in audio, and that is that since we have to rely on memeory, and memory is not exactly reliable, DBT is in fact impossible, as you cannot compare something perceived directly now with something remembered, accurately. This argument does of course apply even more powerfully to any kind of sighted test, hence it's a pointless argument. I usually use the 'take it away' criterion. When I auditioned Monster interconnect cables, I compared the $50 ones to the $100 ones. When I put in the $100 ones to replace the $50 ones, I could not be sure I heard a difference. When I put back the cheaper ones, I was certain I did. I repeated this test a number of times, always with the same results: the 'worse' was easier to spot when the 'better' was taken away than when the 'worse' was taken away and the 'better' installed. And you did this under blind conditions? So, in sum: THERE IS NO WAY TO DO DBT IN AUDIO. IT'S A MYTH. Well, that's one opinion, but it seems to have no substance....... -- Stewart Pinkerton | Music is Art - Audio is Engineering |
#213
|
|||
|
|||
THE ENDLESS DBT DEBATE
On Fri, 03 Oct 2003 02:37:37 GMT, (Michael
Scarpitti) wrote: DBT in audio is not invalid, it is impossible. You have made this claim before, you have failed to back it up with a logical argument, and major manufacturers use DBTs *every working day* to develop new products. -- Stewart Pinkerton | Music is Art - Audio is Engineering |
#214
|
|||
|
|||
THE ENDLESS DBT DEBATE
Nancy Eilers-Hughes wrote in message news:Xl3fb.482051$Oz4.322232@rwcrnsc54...
Michael Scarpitti wrote: snip I have the following to say about DBT in audio, and that is that since we have to rely on memeory, and memory is not exactly reliable, DBT is in fact impossible, as you cannot compare something perceived directly now with something remembered, accurately. Which *IF* accurate, applies equally to any test, sighted or DBT. No, it does not apply to photographic lens testing. The results are fixed on film and visible at any time. Side-by-side comparisons are possible. Hence the use of time proximate switching to minimize such effects. I usually use the 'take it away' criterion. When I auditioned Monster interconnect cables, I compared the $50 ones to the $100 ones. When I put in the $100 ones to replace the $50 ones, I could not be sure I heard a difference. When I put back the cheaper ones, I was certain I did. I repeated this test a number of times, always with the same results: the 'worse' was easier to spot when the 'better' was taken away than when the 'worse' was taken away and the 'better' installed. So, in sum: THERE IS NO WAY TO DO DBT IN AUDIO. IT'S A MYTH. Uhmmm...all of which required you to remember how one cable sounded, some time ago, and mentally compare what you currently hear with that memory. That's correct. Oh, wait, you yourself said that was impossible, so obviously your cable test was a MYTH by your own admission. I DID NOT claim it was a double blind test, (now did I?), because I believe that to be impossible in the strict sense of these terms FOR COMPARISONS OF AUDIO EQUIPMENT. I believe it was a useful and valid test: could I hear a difference when I switched them in and out. The answer was 'yes'. And the memory is indeed imperfect, so repeated tests over a long listening session were required. The difference was slight, but noticeable and repeatable. Actually, IMO, the myth is that knowing which cable was "worse" (i.e. cheaper) did not affect the test outcome. It did not affect the outcome. I wanted to see which set of cables to use on which components, as I had already purchased them. The best cables were to go between the CD player and the power amp, as I do not use a pre-amp. I own a Sony CD player with a custom-modified variable output stage. I had several sets of cables of various ages and price ranges, some no longer remembered. snip Keith Hughes |
#215
|
|||
|
|||
THE ENDLESS DBT DEBATE
(Michael
Scarpitti) wrote: DBT in audio is not invalid, it is impossible. (Stewart Pinkerton) wrote: You have made this claim before, you have failed to back it up with a logical argument, and major manufacturers use DBTs *every working day* to develop new products. And you have used this weak justification before, too. ALL high end equipment manufacturers use sighted listening in designing their products. A few use sighted listening and DBTs. By this criteria, sighted listening would seem much superior since it is used more widely by equipment manufacturers and designers. Regards, Mike |
#216
|
|||
|
|||
THE ENDLESS DBT DEBATE
"Michael Scarpitti" wrote in message news:2Xifb.
No, it does not apply to photographic lens testing. The results are fixed on film and visible at any time. Side-by-side comparisons are possible. How would lens testing be different? Yes, side by side comparisons are possible. But you can only look at one photo at a time. Switching between them still means a delay and memory is involved all the same. Dennis |
#217
|
|||
|
|||
THE AUDIOPHILE PRESS
(Stewart Pinkerton) wrote in message ...
On 8 Sep 2003 22:59:18 GMT, (Mkuller) wrote: I have no intent in engaging in another dbt debate (or in continuing one) with you. I'm sure that you don't. You prefer to assert and run, as you have no basis for your arguments. Neither you nor anyone else has produced evidence that dbts, even if implemented perfectly (which is unlikely for the average audiophile) can show subtle differences in audio components using music as the program. OTOH, many people have shown that sighted bias totally swamps any subtle difference which *may* exist between components, while many major manufacturers use DBTs every day in their research labs because they are not interested in this pathetic 'debate' - they already *know* that DBTs are the *only* way to determine subtle but *real* sonic differences. Subtle differences will disappear and the components will sound more alike than different. Evidence, please, for this *claim*. It should be obvious to anyone who has tried a dbt, used sighted listening and followed the "dbt debates that never end on RAHE". What is obvious is that the 'subtle differences' which 'disappear' in DBTs never existed in the real physical world. The classic evidence for this is the 'false sighted' test, where the audience believes that it is comparing two diifferent components, but in reality they are not switched. This does not prevent the vast majority of listeners from 'hearing' quite large claimed differences. OTOH, DBTs *do* reveal subtle but *real* differences which are not detectable by other means. That's why they're used by Revel, JBL, B&W, KEF etc etc. OTOH, many people have shown that sighted bias totally swamps any subtle difference which *may* exist between components, while many major manufacturers use DBTs every day in their research labs because they are not interested in this pathetic 'debate' - they already *know* that DBTs are the *only* way to determine subtle but *real* sonic differences. Nonsense. There is NO debate possible. True DBT isn't even remotely possible for audio evaluation. DBT can be applied only to certain kinds of scientific enquiries, and audio isn't one of them. In drug evaluations, for instance, DBT ensures that neither the doctors nor the patients know whether a placebo or the real drug is being given to the patient. In some cases, even the placebo can be perceived to be effective if the disorder is somewhat subjective (depression, for instance). In such testing, DBT means only that that the participants don't know which drug, if any, is being assigned to which patient. That's all. It does not make the test any more probatory, just less biased. The results of questionaires about mood, for instance, are tabulated and compared. Reports of dry mouth and constipation are reported, and the drug's side effects table is compiled. But no-one is suggesting that this is anything more than data collection. If enough MORE people on drug 'A' report benefits and positive mood than on drug 'B' (placebo), the maker has reason to be encouraged and conduct further tests. In the case of antibiotics, for instance, no DBT is warranted, as neither the researcher nor the bacterium has any way to affect the results if the tests are conducted properly. So, ultimately, DBT has only restricted usage (say to examine spritualists' claims) and audio is not one of them. You can fool some of the people all of the time, and all of the people some of the time, but you can't fool all of the people all of the time. Cables do vary in their performance. Green edging DOES reduce the noise on CDs in some cases. Green platters do almost as much. Spray-on cleaners have done nothing, unfortunately, in my experience. |
#218
|
|||
|
|||
THE ENDLESS DBT DEBATE
On 3 Oct 2003 23:26:58 GMT, "Dennis Moore"
wrote: "Michael Scarpitti" wrote in message news:2Xifb. No, it does not apply to photographic lens testing. The results are fixed on film and visible at any time. Side-by-side comparisons are possible. How would lens testing be different? Yes, side by side comparisons are possible. But you can only look at one photo at a time. Switching between them still means a delay and memory is involved all the same. Quite right, which is why astronomers use a 'flicker' comparator which precisely overlays two images and switches between them, causing any differences in the images to flicker and become easily visible. It is not unlike fast-switching ABX for audio............. -- Stewart Pinkerton | Music is Art - Audio is Engineering |
#219
|
|||
|
|||
THE ENDLESS DBT DEBATE
|
#220
|
|||
|
|||
THE ENDLESS DBT DEBATE
How would lens testing be different? Yes, side by side comparisons
are possible. But you can only look at one photo at a time. Switching between them still means a delay and memory is involved all the same. Lapse in time should not matter. The situation is easy with sound as compared, say, with taste. Its very, very easy to conduct double blind tests of food products, and they work. You can reach conclusions about whether, for example, two kinds of wine taste different from one another. And you can also reach conclusions about which wine the sample group preferred. But with taste, the situation is complicated because tastes tend to linger, and the lingering taste will have an effect on the next thing tasted. That's why people need to cleanse their palates before a new tasting. And all of this takes time. But the testing is not only possible, it is fairly reliable, and its done all of the time. Duffy |
#221
|
|||
|
|||
THE ENDLESS DBT DEBATE
"Dennis Moore" wrote in message
... "Michael Scarpitti" wrote in message news:2Xifb. No, it does not apply to photographic lens testing. The results are fixed on film and visible at any time. Side-by-side comparisons are possible. How would lens testing be different? Yes, side by side comparisons are possible. But you can only look at one photo at a time. Switching between them still means a delay and memory is involved all the same. Dennis Have you ever heard of the "blink microscope"? Zeiss designed this device almost 100 years ago. Its purpose was to discover whether any of 1000's of stars had moved between 2 photographic exposures of the same piece of sky. Memory is not involved; you are looking for movement only. Norm Strong |
#222
|
|||
|
|||
THE ENDLESS DBT DEBATE
"normanstrong" wrote in message ...
"Dennis Moore" wrote in message ... "Michael Scarpitti" wrote in message news:2Xifb. No, it does not apply to photographic lens testing. The results are fixed on film and visible at any time. Side-by-side comparisons are possible. How would lens testing be different? Yes, side by side comparisons are possible. But you can only look at one photo at a time. Switching between them still means a delay and memory is involved all the same. Have you ever heard of the "blink microscope"? Zeiss designed this device almost 100 years ago. Its purpose was to discover whether any of 1000's of stars had moved between 2 photographic exposures of the same piece of sky. Memory is not involved; you are looking for movement only. Indeed it is a stupendously senisitve device. It serves 2 main functions: searching for asteroids and planets, where only 1 point in several thousand or more might move, and searching for supernova or variable stars, where changes ni brightness result in very small changes in the size of the image on the plates due to halation effects. It requires very careful setup, very precise alignment of the two plates being compared, exact matching of the illumination levels of each (gee, JUST like very careful level matching and such in high- resolution audio comparisons). Yet another example of the fact that the ability of human perception to detect small differences is enhanced by time-proximal comparisons: it REDUCES the dependence on memory, contrary to what is oft claimed in the high-end audio realm (and almost no where else). |
#223
|
|||
|
|||
THE AUDIOPHILE PRESS
|
#224
|
|||
|
|||
THE AUDIOPHILE PRESS
(Stewart Pinkerton) wrote in message ...
On 4 Oct 2003 15:39:08 GMT, (Michael Scarpitti) wrote: (Stewart Pinkerton) wrote in message ... OTOH, many people have shown that sighted bias totally swamps any subtle difference which *may* exist between components, while many major manufacturers use DBTs every day in their research labs because they are not interested in this pathetic 'debate' - they already *know* that DBTs are the *only* way to determine subtle but *real* sonic differences. Nonsense. There is NO debate possible. True DBT isn't even remotely possible for audio evaluation. DBT can be applied only to certain kinds of scientific enquiries, and audio isn't one of them. Certainly audio is one of them, your repetition of this baseless claim doesn't alter the facts of the matter. It's not a scientific 'test' at all. You're comparing YOUR MEMORY of the sound with the ACTUAL PERCEIVED sound. Would you consider it a valid test of a cloth's color if I showed you a sample of cloth (far enough away that you could not recognize any pattern in the weave), then removed it, and then showed you another (either the same or a different one) five minutes later? Would you consider that a scientific test of the color of the sample? Of course not! In drug evaluations, for instance, DBT ensures that neither the doctors nor the patients know whether a placebo or the real drug is being given to the patient. Quite so - as with audio DBTs. It ensures no bias, but that's all it does. But just because something is unbiased doesn't mean it's a valid test. It's not probatory. It doesn't do what it claims to do. Only long-term comparative listening can show what differences there are. snip of irrelevant material So, ultimately, DBT has only restricted usage (say to examine spritualists' claims) and audio is not one of them. Utter rubbish, as DBTs are used every working day in the R&D labs of many of the major audio manufacturers. The best known are Meridian, KEF, B&W and Harman International, which includes Revel, JBL, Mark Levinson and Madrigal. That proves nothing whatsoever. The question is: 'Is ABX a scientifically valid test'. The answer is 'No'. You can fool some of the people all of the time, and all of the people some of the time, but you can't fool all of the people all of the time. Quite so, and very few people stay fooled about 'high end' audio..... Cables do vary in their performance. Not in any audible way. Green edging DOES reduce the noise on CDs in some cases. NO, it absolutely does no such thing! As it happens, four of us conducted a carefully set up DBT on this, using six copies of the same CD (one of the group owns a record shop). The result was that green pens made no difference. It was only later that I discovered that the whole 'green pen' industry is based on a practical joke played by Jim Johnston on one of the early audio newsgroups. It works. I tried it on two identical copies of a CD. Less hiss and noise were present, and the difference was blatantly obvious when I switched the two CDs. I also have used a green platter. It makes less of a change, but the same kind of change. I also tried a spray-on product. It mmade no change whatsoever that I could hear. If I were disposed to believe on way or then other, I would expect to hear no difference. The fact that I was skeptical means that the green pen made enough change to overcome my skepticism. The fact that I heard a difference means also that I am a carful and sensitive listener. I used Stax Lambda headphones for my listning evaluation. Other transducer products may mask the subtle change. I did not know what to expect when these products were purchased. I heard the most difference with the pen, a little less with the platter, and none with the spray cleaner. This was repeatable and consistent. I ALWAYS heard the distinct improvement with the pen. I ALWAYS heard a similar improvement with the platter, but somewhat less intense. I NEVER heard any difference with the spray cleaner. In the case of the first tow products, the difference was subtle but distinct. If I had been influenced by claims of the mfrs, I should be equally influenced by any of the products. The fact that there were distinct differences among them meant that 'expectation bias' did not enter into it. The bottom line is that there's no possible mechanism for this claimed effect, given how CD replay actually works. Of course there is. Don't be silly! Any physical change in a system can influence it. Green platters do almost as much. That's true, in so far as neither actually does *anything*.......... Since I heard a difference in the effects of these products, I must assume something produced some sonic effect. Whether or not I could pick it out in an ABX test is irrelevant, though I would be very surprised if I could not. |
#225
|
|||
|
|||
THE AUDIOPHILE PRESS
Michael Scarpitti wrote:
(Stewart Pinkerton) wrote in message ... On 4 Oct 2003 15:39:08 GMT, (Michael Scarpitti) wrote: (Stewart Pinkerton) wrote in message ... OTOH, many people have shown that sighted bias totally swamps any subtle difference which *may* exist between components, while many major manufacturers use DBTs every day in their research labs because they are not interested in this pathetic 'debate' - they already *know* that DBTs are the *only* way to determine subtle but *real* sonic differences. Nonsense. There is NO debate possible. True DBT isn't even remotely possible for audio evaluation. DBT can be applied only to certain kinds of scientific enquiries, and audio isn't one of them. Certainly audio is one of them, your repetition of this baseless claim doesn't alter the facts of the matter. It's not a scientific 'test' at all. You're comparing YOUR MEMORY of the sound with the ACTUAL PERCEIVED sound. So? How does that make it NOT a scientific test at all? Are you saying that no test involving memory can be scientific? Would you consider it a valid test of a cloth's color if I showed you a sample of cloth (far enough away that you could not recognize any pattern in the weave), then removed it, and then showed you another (either the same or a different one) five minutes later? Would you consider that a scientific test of the color of the sample? Of course not! It certainly could be, depending on what was being tested. Do you actually assert that all the papers ever published in the field of psychoacoustics, not to mention perceptual psychology in general, are NOT scientific? In drug evaluations, for instance, DBT ensures that neither the doctors nor the patients know whether a placebo or the real drug is being given to the patient. Quite so - as with audio DBTs. It ensures no bias, but that's all it does. But just because something is unbiased doesn't mean it's a valid test. It's not probatory. It doesn't do what it claims to do. Only long-term comparative listening can show what differences there are. But long term comparative SIGTED listening tests are susceptible to bias. And DBTs can be as long term as you like too. So what's your beef? snip of irrelevant material So, ultimately, DBT has only restricted usage (say to examine spritualists' claims) and audio is not one of them. Utter rubbish, as DBTs are used every working day in the R&D labs of many of the major audio manufacturers. The best known are Meridian, KEF, B&W and Harman International, which includes Revel, JBL, Mark Levinson and Madrigal. That proves nothing whatsoever. The question is: 'Is ABX a scientifically valid test'. The answer is 'No'. Strange, then, that scientists use it, don't you think? Are you preparing a paper to set the field straight? Clearly a mistake has been made. conducted a carefully set up DBT on this, using six copies of the same CD (one of the group owns a record shop). The result was that green pens made no difference. It was only later that I discovered that the whole 'green pen' industry is based on a practical joke played by Jim Johnston on one of the early audio newsgroups. It works. I tried it on two identical copies of a CD. Less hiss and noise were present, and the difference was blatantly obvious when I switched the two CDs. You had to use memory, then, didn't you? Not very scientific. I also have used a green platter. It makes less of a change, but the same kind of change. I also tried a spray-on product. It mmade no change whatsoever that I could hear. If I were disposed to believe on way or then other, I would expect to hear no difference. The fact that I was skeptical means that the green pen made enough change to overcome my skepticism. Hardly. The fact that I heard a difference means also that I am a carful and sensitive listener. Hardly. I did not know what to expect when these products were purchased. I heard the most difference with the pen, a little less with the platter, and none with the spray cleaner. This was repeatable and consistent. I ALWAYS heard the distinct improvement with the pen. I ALWAYS heard a similar improvement with the platter, but somewhat less intense. I NEVER heard any difference with the spray cleaner. In the case of the first tow products, the difference was subtle but distinct. If I had been influenced by claims of the mfrs, I should be equally influenced by any of the products. The fact that there were distinct differences among them meant that 'expectation bias' did not enter into it. Are you aware that the whole green pen thing was a JOKE started by an objectivist audiophile? The bottom line is that there's no possible mechanism for this claimed effect, given how CD replay actually works. Of course there is. Don't be silly! Any physical change in a system can influence it. Good gracious. Green platters do almost as much. That's true, in so far as neither actually does *anything*.......... Since I heard a difference in the effects of these products, I must assume something produced some sonic effect. Other assumptions are possible, and must be ruled out. Whether or not I could pick it out in an ABX test is irrelevant, though I would be very surprised if I could not. You'd be the first. Go for it. -- -S. |
#226
|
|||
|
|||
THE AUDIOPHILE PRESS
(Michael Scarpitti) wrote in message ...
(Stewart Pinkerton) wrote in message ... On 4 Oct 2003 15:39:08 GMT, (Michael Scarpitti) wrote: (Stewart Pinkerton) wrote in message ... Scarpitti said: OTOH, many people have shown that sighted bias totally swamps any subtle difference which *may* exist between components, while many major manufacturers use DBTs every day in their research labs because they are not interested in this pathetic 'debate' - they already *know* that DBTs are the *only* way to determine subtle but *real* sonic differences. Pinkerton answered: Nonsense. There is NO debate possible. True DBT isn't even remotely possible for audio evaluation. DBT can be applied only to certain kinds of scientific enquiries, and audio isn't one of them. Certainly audio is one of them, your repetition of this baseless claim doesn't alter the facts of the matter. Scarpitti: It's not a scientific 'test' at all. You're comparing YOUR MEMORY of the sound with the ACTUAL PERCEIVED sound. Would you consider it a valid test of a cloth's color if I showed you a sample of cloth (far enough away that you could not recognize any pattern in the weave), then removed it, and then showed you another (either the same or a different one) five minutes later? Would you consider that a scientific test of the color of the sample? Of course not! In drug evaluations, for instance, DBT ensures that neither the doctors nor the patients know whether a placebo or the real drug is being given to the patient. Pinkerton Quite so - as with audio DBTs. Double Blind component comparison and the ABX protocol are two different things. The practice of using the two interchangeably is common. Blind comparing of components to conceal their origin, looks etc is a sensible manoeuvre to prevent bias. But as Mr. Scarpitti says the comparison of consecutive snippets works for some and it does not for others. And those others may THINK they are as good as anyone but they do not KNOW it. So at home, for an audio consumer it is not and it can not be a valid comparison. They may LEARN to hear in prolonged comparison what they did not hear before. Or not. Who can tell? "Double' has only been thrown in in imitation of the original medical drug testing protocol. It makes perfect sense the I witnessed the enthusiasm of the testing physician pervert the test.. In one case an enthusiastic physician surrounded by the attendant train of nurses and junior medical staff approached his first subject, long bedridden with deforming rheumatoid arthritis , bearing a loaded syringe and announcing a "new promising drug". The patient waited one minute and then jumped out of bed and took his first few steps in years. Sadly this flash in a pan did not recur. If things like that were not common people would not be discarding crutches and wheelchairs in the places of pilgrimage (but never a wooden leg as G.B. Shaw nastily observed). In most cases when comparing components at home with assistance and reasonable precautions single blind is quite sufficient. ABX protocol adds one more complication. You have to memorise A, then B, then X and compare X with your memory of A and B. When comparing components for what they'll do for us playing MUSIC we have to use a complex MUSICAL signal. You collect a motley crew of "audiophiles" and ask them if they heard a difference or not. Who is "right"? The few who heard the difference or the majority who did not? If you average the results you almost guarantee the "no difference" verdict for anything under the sun. This is an entirely different situation from psychometric or psychoacoustic research. where a simple artefact of one kind or another is used for a selected, trained audience. You hear it or you do not. Period. You can check and validate. At that the individuals diverge widely. The trained, as a group, performed TWO HUNDRED SEVENTY times better than the untrained in Sean Olive loudspeaker comparison group.(see the thread). And these loudspeakers must have sounded VERY different. Two had droopy bass and one hybrid (that I had heard) had two distinct radiating sound sources: the woofer and the rest. (no names!). I did not buy it. Not to mention the intra group differences between individuals. Please, note that Sean Olive DID NOT choose to use the ABX protocol. He went double blind only. I accept that blind listening is advantageous and may be a necessity for some in this market- driven society. I accept that proximate, ie brief snippet AFTER a brief snippet, listening is better for those who are good at comparing CONSECUTIVE signals. I myself prefer to compare blind SIMULTANEOUSLY left-right with random changes. But one way or the other it may be advantageous for an audio consumer to compare blind, at home, before buying. I accept that psychoacousticians are happy with the ABX protocol or its derivatives. THEY ARE DOING SOMETHING COMPLETELY DIFFERENT FROM COMPONENT COMPARISON FOR THEIR ABILITY TO REPRODUCE MUSIC. I accept that some manufacturers use undefined, unreported double blind (ABX?) comparisons. (see note below) What I do not accept is that an ABX component comparison for an average , untrained audiophile is THE TEST. I do not believe that it proves anything if such a person hears no difference. There may be none or else he is no good at ABX. How can he tell? He can't even tell if with training he may not improve enough to hear what he did not hear before. The challenges to anyone, including audio reviewers, to take ABX test are absurd. One can be a discriminating listener but terrible at ABXing -like for instance the writer. Scarpitti It ensures no bias, but that's all it does. But just because something is unbiased doesn't mean it's a valid test. It's not probatory. It doesn't do what it claims to do. Only long-term comparative listening can show what differences there are. So, ultimately, DBT has only restricted usage (say to examine spritualists' claims) and audio is not one of them. Pinkerton: Utter rubbish, as DBTs are used every working day in the R&D labs of many of the major audio manufacturers. The best known are Meridian, KEF, B&W and Harman International, which includes Revel, JBL, Mark Levinson and Madrigal. I wonder when Mr. Pinkerton will stop calling on "manufacturers" to defend his views The first step in any scientific-aspiring debate is to supply a proper reference as used in any branch of science. Reference points one to the printed material that contains the details of: First; the purpose of the test: A manufacturers' purpose could be to find out eg. "Are we doing worse than Yamaha?" 1) how the subjects and the controls were selected 2) what statistical criteria were used 3) what kind of signal was used (pink noise?, music?) 3) what was compared with what including the details of ancillary equipment. 4) what were the results, tabulated, with NUMBERS. 5) otcome with interpretation . I pointed out several times that this kind of "reference" is just gossip. When we truly find a reference to read such as Sean Olive JAES article it turns out that 1) He did not use ABX 2) there were tremendous differences in performance between his subjects Telling untrained, audio consumers with different musical interests and experience that an ABX test will solve their buying problems because manufacturers use it is an irrelevancy.. Of course, if one has nothing better... The curious thing is that , of all people, it is those fulminating most volubly against "high-end snake oil" should call on its prestige at their convenience. Ludovic Mirabel That proves nothing whatsoever. The question is: 'Is ABX a scientifically valid test'. The answer is 'No'. You can fool some of the people all of the time, and all of the people some of the time, but you can't fool all of the people all of the time. Quite so, and very few people stay fooled about 'high end' audio..... Cables do vary in their performance. Not in any audible way. Green edging DOES reduce the noise on CDs in some cases. NO, it absolutely does no such thing! As it happens, four of us conducted a carefully set up DBT on this, using six copies of the same CD (one of the group owns a record shop). The result was that green pens made no difference. It was only later that I discovered that the whole 'green pen' industry is based on a practical joke played by Jim Johnston on one of the early audio newsgroups. It works. I tried it on two identical copies of a CD. Less hiss and noise were present, and the difference was blatantly obvious when I switched the two CDs. I also have used a green platter. It makes less of a change, but the same kind of change. I also tried a spray-on product. It mmade no change whatsoever that I could hear. If I were disposed to believe on way or then other, I would expect to hear no difference. The fact that I was skeptical means that the green pen made enough change to overcome my skepticism. The fact that I heard a difference means also that I am a carful and sensitive listener. I used Stax Lambda headphones for my listning evaluation. Other transducer products may mask the subtle change. I did not know what to expect when these products were purchased. I heard the most difference with the pen, a little less with the platter, and none with the spray cleaner. This was repeatable and consistent. I ALWAYS heard the distinct improvement with the pen. I ALWAYS heard a similar improvement with the platter, but somewhat less intense. I NEVER heard any difference with the spray cleaner. In the case of the first tow products, the difference was subtle but distinct. If I had been influenced by claims of the mfrs, I should be equally influenced by any of the products. The fact that there were distinct differences among them meant that 'expectation bias' did not enter into it. The bottom line is that there's no possible mechanism for this claimed effect, given how CD replay actually works. Of course there is. Don't be silly! Any physical change in a system can influence it. Green platters do almost as much. That's true, in so far as neither actually does *anything*.......... Since I heard a difference in the effects of these products, I must assume something produced some sonic effect. Whether or not I could pick it out in an ABX test is irrelevant, though I would be very surprised if I could not. |
#227
|
|||
|
|||
THE AUDIOPHILE PRESS
(Stewart Pinkerton) wrote in message ...
On 4 Oct 2003 15:39:08 GMT, (Michael Scarpitti) wrote: (Stewart Pinkerton) wrote in message ... OTOH, many people have shown that sighted bias totally swamps any subtle difference which *may* exist between components, while many major manufacturers use DBTs every day in their research labs because they are not interested in this pathetic 'debate' - they already *know* that DBTs are the *only* way to determine subtle but *real* sonic differences. Nonsense. There is NO debate possible. True DBT isn't even remotely possible for audio evaluation. DBT can be applied only to certain kinds of scientific enquiries, and audio isn't one of them. Certainly audio is one of them, your repetition of this baseless claim doesn't alter the facts of the matter. In drug evaluations, for instance, DBT ensures that neither the doctors nor the patients know whether a placebo or the real drug is being given to the patient. Quite so - as with audio DBTs. snip of irrelevant material So, ultimately, DBT has only restricted usage (say to examine spritualists' claims) and audio is not one of them. Utter rubbish, as DBTs are used every working day in the R&D labs of many of the major audio manufacturers. The best known are Meridian, KEF, B&W and Harman International, which includes Revel, JBL, Mark Levinson and Madrigal. You can fool some of the people all of the time, and all of the people some of the time, but you can't fool all of the people all of the time. Quite so, and very few people stay fooled about 'high end' audio..... Cables do vary in their performance. Not in any audible way. Green edging DOES reduce the noise on CDs in some cases. NO, it absolutely does no such thing! As it happens, four of us conducted a carefully set up DBT on this, using six copies of the same CD (one of the group owns a record shop). The result was that green pens made no difference. It was only later that I discovered that the whole 'green pen' industry is based on a practical joke played by Jim Johnston on one of the early audio newsgroups. The bottom line is that there's no possible mechanism for this claimed effect, given how CD replay actually works. Green platters do almost as much. That's true, in so far as neither actually does *anything*.......... If what (I thought?) 'I heard' is due to 'bias', how do you explain the CONSISTENT differences I heard among three CD treatments (green pen, green platter, and spray cleaner). If 'bias' is at work, how does the 'bias' ALWAYS know to make the green pen the most 'effective' and the green platter slightly less, and the spray cleaner completely ineffective?How sophisticated my listening bias must be to be able to do that, when before I tried any of these products I had no reason to believe they would work at all? If anything, my experience with the first one I tried (green pen) should predispose me to think the others would work as well or better. That they did not work as well or at all seems to be evidence that something other than 'bias' is at work. 'Bias' cannot also account for the consistency between trials. I tried the spray cleaner over and over again, and heard nothing, despite my wanting to.... The green pen was a no-brainer. I heard less noise. The green platter produced a similar effect, but in lesser amounts. The spray cleaner produced no effect. How can 'bias' account for this in repeated trials? |
#228
|
|||
|
|||
THE AUDIOPHILE PRESS
Steven Sullivan wrote in message ...
Michael Scarpitti wrote: (Stewart Pinkerton) wrote in message ... On 4 Oct 2003 15:39:08 GMT, (Michael Scarpitti) wrote: (Stewart Pinkerton) wrote in message ... OTOH, many people have shown that sighted bias totally swamps any subtle difference which *may* exist between components, while many major manufacturers use DBTs every day in their research labs because they are not interested in this pathetic 'debate' - they already *know* that DBTs are the *only* way to determine subtle but *real* sonic differences. Nonsense. There is NO debate possible. True DBT isn't even remotely possible for audio evaluation. DBT can be applied only to certain kinds of scientific enquiries, and audio isn't one of them. Certainly audio is one of them, your repetition of this baseless claim doesn't alter the facts of the matter. It's not a scientific 'test' at all. You're comparing YOUR MEMORY of the sound with the ACTUAL PERCEIVED sound. So? How does that make it NOT a scientific test at all? Are you saying that no test involving memory can be scientific? Would you consider it a valid test of a cloth's color if I showed you a sample of cloth (far enough away that you could not recognize any pattern in the weave), then removed it, and then showed you another (either the same or a different one) five minutes later? Would you consider that a scientific test of the color of the sample? Of course not! It certainly could be, depending on what was being tested. 'The color of the cloth', is what I explicity stated! |
#229
|
|||
|
|||
THE AUDIOPHILE PRESS
At that the individuals diverge widely. The trained, as a group,
performed TWO HUNDRED SEVENTY times better than the untrained in Sean Olive loudspeaker comparison group.(see the thread). Must have been that wine for dinner. The correct number is TWENTY SEVEN times. No doubt the well wishers will point it out. In spades. The joke is that it was I myself who posted the correct figure in the "Sean Oliver" thread. Ludovic Mirabel |
#230
|
|||
|
|||
THE AUDIOPHILE PRESS
On 6 Oct 2003 22:00:08 GMT, (Michael
Scarpitti) wrote: (Stewart Pinkerton) wrote in message ... On 4 Oct 2003 15:39:08 GMT, (Michael Scarpitti) wrote: Green edging DOES reduce the noise on CDs in some cases. NO, it absolutely does no such thing! As it happens, four of us conducted a carefully set up DBT on this, using six copies of the same CD (one of the group owns a record shop). The result was that green pens made no difference. It was only later that I discovered that the whole 'green pen' industry is based on a practical joke played by Jim Johnston on one of the early audio newsgroups. The bottom line is that there's no possible mechanism for this claimed effect, given how CD replay actually works. Green platters do almost as much. That's true, in so far as neither actually does *anything*.......... If what (I thought?) 'I heard' is due to 'bias', how do you explain the CONSISTENT differences I heard among three CD treatments (green pen, green platter, and spray cleaner). If 'bias' is at work, how does the 'bias' ALWAYS know to make the green pen the most 'effective' and the green platter slightly less, and the spray cleaner completely ineffective? Oh dear, are we reduced to explaining the basics - again? You believed that green was some kind of magic colour (it isn't), so you believed that something happened. Here's a clue - nothing happened. Once you reach an initial conclusion under sighted conditions, you will typically just reinforce it with each trial. This is quite unlike blind testing. How sophisticated my listening bias must be to be able to do that, when before I tried any of these products I had no reason to believe they would work at all? If anything, my experience with the first one I tried (green pen) should predispose me to think the others would work as well or better. That they did not work as well or at all seems to be evidence that something other than 'bias' is at work. 'Bias' cannot also account for the consistency between trials. I tried the spray cleaner over and over again, and heard nothing, despite my wanting to.... Well, that's hardly a surpirise! The green pen was a no-brainer. I heard less noise. The green platter produced a similar effect, but in lesser amounts. The spray cleaner produced no effect. How can 'bias' account for this in repeated trials? Easily, it happens all the time. Get over it and try it again under blind conditions. Green pens do *not* work, indeed they *can't* work, because there is no mechanism for them to work. -- Stewart Pinkerton | Music is Art - Audio is Engineering |
#231
|
|||
|
|||
THE ENDLESS DBT DEBATE
Michael Scarpitti wrote:
Nancy Eilers-Hughes wrote in message news:Xl3fb.482051$Oz4.322232@rwcrnsc54... Michael Scarpitti wrote: snip I have the following to say about DBT in audio, and that is that since we have to rely on memeory, and memory is not exactly reliable, DBT is in fact impossible, as you cannot compare something perceived directly now with something remembered, accurately. Which *IF* accurate, applies equally to any test, sighted or DBT. No, it does not apply to photographic lens testing. The results are fixed on film and visible at any time. Side-by-side comparisons are possible. Which of course is irrelevant, and totally outside the topic under discussion. Your posited limitation applies equally to sighted (NOT visual) comparisons of audio equipment. Hence the use of time proximate switching to minimize such effects. I usually use the 'take it away' criterion. When I auditioned Monster interconnect cables, I compared the $50 ones to the $100 ones. When I put in the $100 ones to replace the $50 ones, I could not be sure I heard a difference. When I put back the cheaper ones, I was certain I did. I repeated this test a number of times, always with the same results: the 'worse' was easier to spot when the 'better' was taken away than when the 'worse' was taken away and the 'better' installed. So, in sum: THERE IS NO WAY TO DO DBT IN AUDIO. IT'S A MYTH. Uhmmm...all of which required you to remember how one cable sounded, some time ago, and mentally compare what you currently hear with that memory. That's correct. Thank you. You agree then that your premise is false. Oh, wait, you yourself said that was impossible, so obviously your cable test was a MYTH by your own admission. I DID NOT claim it was a double blind test, (now did I?), because I believe that to be impossible in the strict sense of these terms FOR COMPARISONS OF AUDIO EQUIPMENT. It is of no consequence whether the test was SBT, DBT, or any other protocol. You just agreed, above, that sonic memory was required to perform any comparison test, including the one you performed, yet your claim is, quote; "as you cannot compare something perceived directly now with something remembered, accurately". I believe it was a useful and valid test: could I hear a difference when I switched them in and out. The answer was 'yes'. And the memory is indeed imperfect, so repeated tests over a long listening session were required. The difference was slight, but noticeable and repeatable. Repetition of a 'flawed' test imparts no additional validity. Since there are plenty of perceptual biases that can affect the outcome of sighted tests, you cannot, based on your premise, *ever* conclude that components sound different or, indeed, alike. Actually, IMO, the myth is that knowing which cable was "worse" (i.e. cheaper) did not affect the test outcome. It did not affect the outcome. I wanted to see which set of cables to use on which components, as I had already purchased them. The best cables were to go between the CD player and the power amp, as I do not use a pre-amp. I own a Sony CD player with a custom-modified variable output stage. I had several sets of cables of various ages and price ranges, some no longer remembered. Well, again, based on your own assertions, you cannot know if the outcome was affected. That's the whole point of the discussion. All your claimed "deficiencies" relative to DBT's are also limitations of all sighted tests (note I did not say visual comparisons). Sighted tests just have the additional deficiency of no bias controls (at a minimum). Keith Hughes |
#232
|
|||
|
|||
THE AUDIOPHILE PRESS
(Stewart Pinkerton) wrote in message ...
On 6 Oct 2003 22:00:08 GMT, (Michael Scarpitti) wrote: I heard something DESPITE my expecting NOT to at first (the green pen), and heard nothing when I expected to (the spray-on cleaner). That's sure some very tricky 'bias'! It blows your assertion out of the water. It was consistent and repeatable. You know, 'scientific'. (Stewart Pinkerton) wrote in message ... On 4 Oct 2003 15:39:08 GMT, (Michael Scarpitti) wrote: Green edging DOES reduce the noise on CDs in some cases. NO, it absolutely does no such thing! As it happens, four of us conducted a carefully set up DBT on this, using six copies of the same CD (one of the group owns a record shop). The result was that green pens made no difference. It was only later that I discovered that the whole 'green pen' industry is based on a practical joke played by Jim Johnston on one of the early audio newsgroups. The bottom line is that there's no possible mechanism for this claimed effect, given how CD replay actually works. Green platters do almost as much. That's true, in so far as neither actually does *anything*.......... If what (I thought?) 'I heard' is due to 'bias', how do you explain the CONSISTENT differences I heard among three CD treatments (green pen, green platter, and spray cleaner). If 'bias' is at work, how does the 'bias' ALWAYS know to make the green pen the most 'effective' and the green platter slightly less, and the spray cleaner completely ineffective? Oh dear, are we reduced to explaining the basics - again? You believed that green was some kind of magic colour (it isn't), so you believed that something happened. Here's a clue - nothing happened. Once you reach an initial conclusion under sighted conditions, you will typically just reinforce it with each trial. This is quite unlike blind testing. How sophisticated my listening bias must be to be able to do that, when before I tried any of these products I had no reason to believe they would work at all? If anything, my experience with the first one I tried (green pen) should predispose me to think the others would work as well or better. That they did not work as well or at all seems to be evidence that something other than 'bias' is at work. 'Bias' cannot also account for the consistency between trials. I tried the spray cleaner over and over again, and heard nothing, despite my wanting to.... Well, that's hardly a surpirise! The green pen was a no-brainer. I heard less noise. The green platter produced a similar effect, but in lesser amounts. The spray cleaner produced no effect. How can 'bias' account for this in repeated trials? Easily, it happens all the time. Get over it and try it again under blind conditions. Green pens do *not* work, indeed they *can't* work, because there is no mechanism for them to work. |
#233
|
|||
|
|||
THE AUDIOPHILE PRESS
(Michael Scarpitti)
(Stewart Pinkerton) wrote in message ... On 6 Oct 2003 22:00:08 GMT, (Michael Scarpitti) wrote: I heard something DESPITE my expecting NOT to at first (the green pen), and heard nothing when I expected to (the spray-on cleaner). That's sure some very tricky 'bias'! Much of human bias is systemic, common to humans, and much of it may be held at the subconscious level. So to 'say' that you held a particular bias is subject to verification. That's why we use bias controls. You did not so you can't just relate what your bias was and that you sometime held it in check with willpower. For example humans are prone to report 'differences' a large percetage of the time when given a pair of identical sound clips ("Can You Trust Your Ears" AES Preprint 3177, 1991 AES Convention). So its relatively easy to 'hear' non-extant differences the first time when auditioning 2 identical sounds and then simply repeat that decision every time going forward under open conditions. This is why people actually make a living selling audio products that have no audio contribution other than basic commodity grade playback capability. It blows your assertion out of the water. It was consistent and repeatable. You know, 'scientific'. Actually when anyone, me included, have tried to duplicate these results with bias controls implemented have never found them to have audible consequence. As Mr Pinkerton notes there is no mechanism for green-ink and green accessories to have an audible effect. |
#234
|
|||
|
|||
THE AUDIOPHILE PRESS
Michael Scarpitti wrote:
(Stewart Pinkerton) wrote in message ... On 6 Oct 2003 22:00:08 GMT, (Michael Scarpitti) wrote: I heard something DESPITE my expecting NOT to at first (the green pen), and heard nothing when I expected to (the spray-on cleaner). That's sure some very tricky 'bias'! It blows your assertion out of the water. It was consistent and repeatable. You know, 'scientific'. No, Michael. Consistent and repeatable without controls for false positives is not scientific. -- -S. ______ "You're an abuser Sullivan....a base beast with intellect but little intelligence to show for it" -- KENNEH! |
#235
|
|||
|
|||
THE AUDIOPHILE PRESS
On 7 Oct 2003 22:44:27 GMT, (Michael
Scarpitti) wrote: (Stewart Pinkerton) wrote in message ... On 6 Oct 2003 22:00:08 GMT, (Michael Scarpitti) wrote: I heard something DESPITE my expecting NOT to at first (the green pen), and heard nothing when I expected to (the spray-on cleaner). That's sure some very tricky 'bias'! It blows your assertion out of the water. It was consistent and repeatable. You know, 'scientific'. Nope. I would merely have to *not* place the 'green pen' disc in the player, while telling you that it *was* there, for you to 'hear' all the same effects. It's not a test when you know the answer..... Speaking of science, I note that you conveniently ignore my point that there is no mechanism for green (or any other colour) pens to have any physical effect on CD replay. -- Stewart Pinkerton | Music is Art - Audio is Engineering |
#236
|
|||
|
|||
THE AUDIOPHILE PRESS
(Nousaine) wrote in message ...
(Michael Scarpitti) (Stewart Pinkerton) wrote in message ... On 6 Oct 2003 22:00:08 GMT, (Michael Scarpitti) wrote: I heard something DESPITE my expecting NOT to at first (the green pen), and heard nothing when I expected to (the spray-on cleaner). That's sure some very tricky 'bias'! Much of human bias is systemic, common to humans, and much of it may be held at the subconscious level. So to 'say' that you held a particular bias is subject to verification. That's why we use bias controls. You did not so you can't just relate what your bias was and that you sometime held it in check with willpower. For example humans are prone to report 'differences' a large percetage of the time when given a pair of identical sound clips ("Can You Trust Your Ears" AES Preprint 3177, 1991 AES Convention). So its relatively easy to 'hear' non-extant differences the first time when auditioning 2 identical sounds and then simply repeat that decision every time going forward under open conditions. This is why people actually make a living selling audio products that have no audio contribution other than basic commodity grade playback capability. It blows your assertion out of the water. It was consistent and repeatable. You know, 'scientific'. Actually when anyone, me included, have tried to duplicate these results with bias controls implemented have never found them to have audible consequence. As Mr Pinkerton notes there is no mechanism for green-ink and green accessories to have an audible effect. I heard it, every time. Less noise. |
#237
|
|||
|
|||
THE AUDIOPHILE PRESS
|
#238
|
|||
|
|||
THE AUDIOPHILE PRESS
On 8 Oct 2003 14:48:32 GMT, Stewart Pinkerton wrote:
Speaking of science, I note that you conveniently ignore my point that there is no mechanism for green (or any other colour) pens to have any physical effect on CD replay. On the contrary, I've heard that black pens can hurt or help CD replay, depending on how they're used.... :-) http://www.wired.com/news/technology...,52665,00.html ...Is an example of how they can "help" though in this case I think it would make my ears hurt more.... -alan -- Alan Hoyle - - http://www.alanhoyle.com/ "I don't want the world, I just want your half." -TMBG Get Horizontal, Play Ultimate. |
#239
|
|||
|
|||
THE AUDIOPHILE PRESS
(Stewart Pinkerton) wrote in message ...
On 8 Oct 2003 14:55:38 GMT, (Michael Scarpitti) wrote: As Mr Pinkerton notes there is no mechanism for green-ink and green accessories to have an audible effect. I heard it, every time. Less noise. No, you didn't. You just *thought* you did. Please understand that the 'green pen' effect is a myth, in fact it all started as a practical joke. It doesn't do anything, because it *cannot* do anything. Try it again under blind conditions. The effecet was so substantial I'll not bother. No, thank you. |
#240
|
|||
|
|||
THE AUDIOPHILE PRESS
On 9 Oct 2003 00:50:35 GMT, (Michael
Scarpitti) wrote: (Stewart Pinkerton) wrote in message ... On 8 Oct 2003 14:55:38 GMT, (Michael Scarpitti) wrote: As Mr Pinkerton notes there is no mechanism for green-ink and green accessories to have an audible effect. I heard it, every time. Less noise. No, you didn't. You just *thought* you did. Please understand that the 'green pen' effect is a myth, in fact it all started as a practical joke. It doesn't do anything, because it *cannot* do anything. Try it again under blind conditions. The effecet was so substantial I'll not bother. No, thank you. None so blind, as those who will not see. If your attitude is that "I heard it, so it *must* be real", then I can only say that : a) You are just plain wrong b) You have no interest in advancing your knowledge of Hi-Fi music reproduction c) Any time you'd like to drop in, I'll be happy to demonstrate that you most certainly did *not* hear any substantial difference -- Stewart Pinkerton | Music is Art - Audio is Engineering |