Home |
Search |
Today's Posts |
#41
|
|||
|
|||
Dielectric properties > DBT Debate
"Mkuller" wrote in message
... (Stewart Pinkerton) wrote: 1) Sighted listening is *known* to be useless for the discrimination of subtle sonic differences, and yet you will not even *try* a blind test. Wait, you must mean "Blind listening is *known* to be useless for the discrimination of subtle sonic differences. Come on, get it right. *Everyone* knows DBTs remove subtle sonic differences. Can you show a published DBT that showed any subtle differences? Of course not. What a kidder you are. *snip* I guess we're back to living with no definition of "subtle sonic difference", since the previously quoted Swedish blind test of CD players clearly differentiated between them. It's published, it revealed a difference between the two players and you decided that since the difference was described as a "brightness", it was gross and not subtle. Who's kidding whom? |
#42
|
|||
|
|||
Dielectric properties
"Michael Scarpitti" wrote in message
... *snip* 1) Sighted listening is *known* to be useless for the discrimination of subtle sonic differences, and yet you will not even *try* a blind test. Clearly, you do *not* trust your ears, despite your protestations. Prove that. I can offer counter-evidence. I tried three different products, and the results were unambiguous and consistent, even though the green pen did not work in all instances, it clearly did on others. They were unambiguous and consistent when you knew what you were listening to. If you don't listen blind, how will you ever know that you really heard what you thought you heard? It's like saying I can pick a glass of my favorite Cabernet in a tasting consistently, as long I can see the labels while I'm tasting. 2) The green pen effect was a *joke*, and there is no physical mechanism by which it *can* work. That's false. Kindly explain, then, exactly what mechanism it is that results in any electrical change that results from the application of a green pen to the edges of a CD. |
#44
|
|||
|
|||
Dielectric properties
(Michael Scarpitti) wrote in message ...
(Stewart Pinkerton) wrote in message ... 2) The green pen effect was a *joke*, and there is no physical mechanism by which it *can* work. That's false. Okay, Mr. Scarpitti, the technical ball is squarely in YOUR court, you invited it there, now it's time for you to play it: You have now claimed that there IS a physical mechanism by which the green pen scheme works. This is a TECHNICAL claim, one that lives or dies on its TECHNICAL merits, Please describe the mechanism at work for us. You have also claimed that CDs treated with the green pen have less noise. Noise is a trivially measurable property of a signal. Show us the consistent, repeatable and unambiguous occurance of this reduction in noise. |
#45
|
|||
|
|||
Dielectric properties
Bruce Abrams wrote in message ...
"Michael Scarpitti" wrote in message ... *snip* 1) Sighted listening is *known* to be useless for the discrimination of subtle sonic differences, and yet you will not even *try* a blind test. Clearly, you do *not* trust your ears, despite your protestations. Prove that. I can offer counter-evidence. I tried three different products, and the results were unambiguous and consistent, even though the green pen did not work in all instances, it clearly did on others. They were unambiguous and consistent when you knew what you were listening to. If you don't listen blind, how will you ever know that you really heard what you thought you heard? It's like saying I can pick a glass of my favorite Cabernet in a tasting consistently, as long I can see the labels while I'm tasting. 2) The green pen effect was a *joke*, and there is no physical mechanism by which it *can* work. That's false. Kindly explain, then, exactly what mechanism it is that results in any electrical change that results from the application of a green pen to the edges of a CD. To prove 'bias' has the effect you say, you should try the following (just a suggestion) 1. Get THREE samples of, say, $100 Monster cable 2. Disguise or alter their physical identity so they all look different 3. Conduct 'open' listening tests with a significant number of test subjects 4. Conduct the same test with three or more sample groups 5. Conduct the same test with the equipment undisguised 6. Compare results |
#46
|
|||
|
|||
Dielectric properties
"Michael Scarpitti" wrote in message
... *snip* To prove 'bias' has the effect you say, you should try the following (just a suggestion) 1. Get THREE samples of, say, $100 Monster cable 2. Disguise or alter their physical identity so they all look different 3. Conduct 'open' listening tests with a significant number of test subjects 4. Conduct the same test with three or more sample groups 5. Conduct the same test with the equipment undisguised 6. Compare results I think the test you want to perform is as follows: 1. Take 3 samples of esoteric, "high-end" cable 2. Disguise (or hide) their identity 3. Conduct 'open' listening tests with a number of test subjects 4. Determine how many of the test subjects can accurately and consistently identify which cable is being used at any point in time after any amount of listening. I've performed this test many times with over a dozen interconnects and 8 different test subjects. The results of this test are what convinced me that the high-end cable industry is essentially engaged in selling snow to Eskimos. A few years ago I was auditioning some interconnects...PBJ, KCAG, Audioquest Emerald, DIY microphone cable with Neutrik RCAs and several others. I thought I had switched from PBJ to KCAG the night before and when I sat down to listen, I heard beautiful, crystal clear highs that had no edge or grain. When I got up after an hour or so to switch back to the PBJ to confirm my findings, I saw that I had in fact been listening to PBJ all allong. So I ran out to the local Rat Shack to get some of their premium interconnects and ordered some (with respect to Steve Lamper) dirt cheap Canare mic cable that I terminated with Neutrik RCAs. Then I gave the lot to my wife and we had some "fun with cables" (keeping it 'G' rated). I even invited my brother and several other audiophile friends for a listen. Without knowing which was playing, noone could tell the difference, yet all had claimed to have heard cables sounding differently prior to this experience. Try the same test yourself and save yourself a whole lot of money and useless "did you hear that?!?" time. It's amazing how much more music I actually listen to now that I don't worry about things like cables and magic bricks anymore. |
#47
|
|||
|
|||
Dielectric properties
|
#48
|
|||
|
|||
Dielectric properties
|
#49
|
|||
|
|||
Dielectric properties
In article ,
Fred E. Davis wrote: On 13 Oct 2003 21:53:01 GMT, (Michael Scarpitti) wrote: (Stewart Pinkerton) wrote in message ... ... 2) The green pen effect was a *joke*, and there is no physical mechanism by which it *can* work. That's false. No, it's true. The favorite story is that the green ink 'absorbs stray However, there was the audio CD protection scheme that was broken using a black Sharpie. That "crack" did, however, have a reasonable explanation. One has to be a little careful - the 1970's "golden ear" crowd were in fact correct about problems with early solid-state devices, and it took some engineers who were also avid listeners to pin down the problems. One that I remember: a lot of early solid-state preamps were very nonlinear above 20Khz and had no input filtering, and it takes only a few minutes with a moving coil phono cartridge and a spectrum analyzer to see that this is not true (it's not music, but it produces audible (IM) distortion). Testing amps with 20Hz to 20Khz sound was never going to find this problem. There were similar problems with solid state power amps; for example, full range ESL's would easily drive amps like the Dyna Stereo 400 into current limiting at sound levels where Dyna MKII's would be occasionally clipping. Vacuum tube designs would measure worse using standard tests but would sound better than their solid-state replacements (for example, the Marantz 7 and 7T, and the Dyna PAS-3 and PAT-4). Mike Squires -- Mike Squires (mikes at cs.indiana.edu) 317 233 9456 (w) 812 333 6564 (h) mikes at siralan.org 546 N Park Ridge Rd., Bloomington, IN 47408 |
#50
|
|||
|
|||
Dielectric properties
On 15 Oct 2003 03:43:29 GMT, (Michael Squires)
wrote: In article , Fred E. Davis wrote: On 13 Oct 2003 21:53:01 GMT, (Michael Scarpitti) wrote: (Stewart Pinkerton) wrote in message ... ... 2) The green pen effect was a *joke*, and there is no physical mechanism by which it *can* work. That's false. No, it's true. The favorite story is that the green ink 'absorbs stray However, there was the audio CD protection scheme that was broken using a black Sharpie. That "crack" did, however, have a reasonable explanation. One has to be a little careful - the 1970's "golden ear" crowd were in fact correct about problems with early solid-state devices, and it took some engineers who were also avid listeners to pin down the problems. You are however missing one crucial point here. No one has *ever* demonstrated that the 'green pen' has any effect under *unsighted* conditions. First the observation, *then* the theorising. -- Stewart Pinkerton | Music is Art - Audio is Engineering |
#51
|
|||
|
|||
Dielectric properties
Bruce Abrams wrote in message ...
"Michael Scarpitti" wrote in message ... *snip* To prove 'bias' has the effect you say, you should try the following (just a suggestion) 1. Get THREE samples of, say, $100 Monster cable 2. Disguise or alter their physical identity so they all look different 3. Conduct 'open' listening tests with a significant number of test subjects 4. Conduct the same test with three or more sample groups 5. Conduct the same test with the equipment undisguised 6. Compare results I think the test you want to perform is as follows: 1. Take 3 samples of esoteric, "high-end" cable 2. Disguise (or hide) their identity 3. Conduct 'open' listening tests with a number of test subjects 4. Determine how many of the test subjects can accurately and consistently identify which cable is being used at any point in time after any amount of listening. I've performed this test many times with over a dozen interconnects and 8 different test subjects. The results of this test are what convinced me that the high-end cable industry is essentially engaged in selling snow to Eskimos. A few years ago I was auditioning some interconnects...PBJ, KCAG, Audioquest Emerald, DIY microphone cable with Neutrik RCAs and several others. I thought I had switched from PBJ to KCAG the night before and when I sat down to listen, I heard beautiful, crystal clear highs that had no edge or grain. When I got up after an hour or so to switch back to the PBJ to confirm my findings, I saw that I had in fact been listening to PBJ all allong. So I ran out to the local Rat Shack to get some of their premium interconnects and ordered some (with respect to Steve Lamper) dirt cheap Canare mic cable that I terminated with Neutrik RCAs. Then I gave the lot to my wife and we had some "fun with cables" (keeping it 'G' rated). I even invited my brother and several other audiophile friends for a listen. Without knowing which was playing, noone could tell the difference, yet all had claimed to have heard cables sounding differently prior to this experience. Try the same test yourself and save yourself a whole lot of money and useless "did you hear that?!?" time. It's amazing how much more music I actually listen to now that I don't worry about things like cables and magic bricks anymore. My post was intended to be a sort of DBT on DBT itself, and unless you didn't follow, that should have been clear. You have to establish THAT 'bias' even exists BEFORE you can use it as a explanatory mechanism. Was that not clear? |
#52
|
|||
|
|||
Dielectric properties
|
#53
|
|||
|
|||
Dielectric properties
"Michael Scarpitti" wrote in message
... My post was intended to be a sort of DBT on DBT itself, and unless you didn't follow, that should have been clear. You have to establish THAT 'bias' even exists BEFORE you can use it as a explanatory mechanism. Was that not clear? The existence of bias in product testing has been an accepted fact for decades. That's why there are ALWAYS controls in place and sighted tests are given no validity. Until you accept that as a fact, further argument is pointless. If you wish to convince yourself that you heard something, try hearing it without knowing that you should be. If you're happy simply thinking that you heard something, by all means continue doing what makes you happy. Just don't try to rationalize the decision by trying to convince yourself or anyone else that you know with any degree of certainty what you heard. Only a bias controlled test can prove that. Audio is no more unique in this respect than taste testing, tactile (eg. quality of material in clothing) or other consumer product testing. If I gave you two otherwise identical shampoos that have different scents, your reaction to how your hair felt after washing would be radically different. Similarly, if I gave you two otherwise identical amps to listen that had different cases (perhaps one with huge, heavy, useless heatsinks), your reaction to their respective sounds would likewise be different. |
#54
|
|||
|
|||
Dielectric properties
(Stewart Pinkerton) wrote in message ...
On 14 Oct 2003 17:57:40 GMT, (Michael Scarpitti) wrote: Bruce Abrams wrote in message ... "Michael Scarpitti" wrote in message ... 2) The green pen effect was a *joke*, and there is no physical mechanism by which it *can* work. That's false. Kindly explain, then, exactly what mechanism it is that results in any electrical change that results from the application of a green pen to the edges of a CD. To prove 'bias' has the effect you say, you should try the following (just a suggestion) See Bruce Abrams excellent reply (BTW, a similar 'accident' with amps made the scales fall from my own eyes, back when I was a 'true believer'). Now, how about you answer the question? I am not, nor was I ever a 'true believer'. I was and remain a skeptic. I hear what I hear and pay no attention whatsoever to what anyone else claims. I'm Popeye the sailor man. *You* claim that there *is* a mechanism for the 'green pen effect', so exactly what do *you* claim that it is? I claim my car starts every morning, does that mean I have to explain it to you? Uh...nope! |
#56
|
|||
|
|||
Dielectric properties >DBT Debate
"Mkuller" wrote in message
... *snip* If sighted audio components tests have no validity, 100,000 or so audiophiles would not subscribe to Stereophile and other magazines to read them. Obviously most audiophiles have found them useful in some way. This is a false supposition. The "high-end" industry (and the audiophiles that proudly associate themselves with it) has been predicated on the false premise of sighted component testing. Audiophiles who read Stereophile and other mags have rarely, if ever, been exposed to the notion of blind testing. Yes, biases do exist. However, the problem I have with ABX-type DBTs is that it has never been proven scientifically that: 1. they are appropriate for amateur audio enthusiasts to use for comparing audio components using music as the source (just because DBTs are used very differently in other settings is no proof). 2. they don't remove more than *just bias* in the audio equipment comparisons. Until you can prove this, the rest is speculation and hand waving. Blind testing has been an scientifically proven and accepted element of product testing in virtually every industry in the last 100 years. Until you can prove that blind testing removes more than *just bias* in audio testing, the rest is speculation and hand waving. In the mean time, I would suggest "long term observational listening, with carefully matched levels at switching" which I have described here before. While it may not be perfect at filtering out all biases and preconceptions for everyone - you may get some false positives - you will not get all of the false negatives (i.e. null results) you get from ABX. So take your pick - false positives or false negatives. Kindly explain what your objection would be to engaging in the exact same "long term observational listening, with carefully matched levels at switching", only "black boxing" the equipment under review, as I've previously discussed. Nothing would be changed except for the knowledge of which component one was listening to. What *if* the few rigid objectivists here who insist ABX is the only answer - are wrong, and the near 100,000 audiophiles are right? In their hearts the ABXers must ultimately fear this, because as audiophiles they want the *best possible* music reproduction from their systems. They would discover that the *just-as-good-stuff* actually isn't. They would have been fooling themselves into rationalizing the purchase of cheaper equipment that doesn't sound nearly as good as their flawed tests showed. First things first, I've never suggested that ABX is the only answer. Other blind protocols for bias control (such as described above and previously) are certainly better than sighted listening and avoid the potential for false negatives that you claim are inherent in the ABX protocol. On the other hand, if the rest of the audiophiles are wrong, and ABX is *the one true way*, our systems still sound as great as we think they do - we've just spent a few more bucks (and more time listening to cables) than we needed to. Secondly and to your "worst case" point...Lets take the example of someone who spends $3500 on speakers and is convinced to spend $1500 on various interconnects and cables. If he were to engage in a DBT of cables and found them to be audibly indistinguishable and therefore spend the $1500 on better speakers or even on music, wouldn't he be far better off? This is the genesis of my issue with the high-end cable makers. Their dubious (at best) claims of audible superiority invariably result in worse sounding systems for the money. Regards, Mike |
#57
|
|||
|
|||
Dielectric properties
|
#58
|
|||
|
|||
Dielectric properties
On 15 Oct 2003 21:43:52 GMT, (Michael
Scarpitti) wrote: (Stewart Pinkerton) wrote in message ... On 14 Oct 2003 17:57:40 GMT, (Michael Scarpitti) wrote: Bruce Abrams wrote in message ... "Michael Scarpitti" wrote in message ... 2) The green pen effect was a *joke*, and there is no physical mechanism by which it *can* work. That's false. Kindly explain, then, exactly what mechanism it is that results in any electrical change that results from the application of a green pen to the edges of a CD. To prove 'bias' has the effect you say, you should try the following (just a suggestion) See Bruce Abrams excellent reply (BTW, a similar 'accident' with amps made the scales fall from my own eyes, back when I was a 'true believer'). Now, how about you answer the question? I am not, nor was I ever a 'true believer'. I was and remain a skeptic. I hear what I hear and pay no attention whatsoever to what anyone else claims. I'm Popeye the sailor man. Sorry Mike, you are most certainly *not* a skeptic, you remain rooted in your solipsistic belief that you are the only person in the world to whom sighted bias does not apply. It seems pointless to argue with someone so unwilling to accept objective evidence. If you refuse to even *try* a blind test because you are utterly convinced that what you hear in sighted testing is real, that is *not* the mark of a skeptic, it's the mark of a deeply insecure person who is sticking his fingers in his ears and refusing to listen to reason. *You* claim that there *is* a mechanism for the 'green pen effect', so exactly what do *you* claim that it is? I claim my car starts every morning, does that mean I have to explain it to you? Uh...nope! In other words, you have no idea what you're talking about. Fine, we're clear about that. I must say that you are at least consistent. -- Stewart Pinkerton | Music is Art - Audio is Engineering |
#59
|
|||
|
|||
Dielectric properties >DBT Debate
|
#60
|
|||
|
|||
Dielectric properties >DBT Debate
"Mkuller" wrote in message
... wrote If sighted audio components tests have no validity, 100,000 or so audiophiles would not subscribe to Stereophile and other magazines to read them. Obviously most audiophiles have found them useful in some way. I'm one of the 100,000. Stereophile is definitely useful in some ways: It's entertaining. It keeps me up to date on developments in the high end audio field. It usually has lab test results that give me an idea of the performance of the component. At the price, I see no reason not to subscribe. How do I feel about the sighted evaluatiions? I don't bother to read them. Norm Strong |
#61
|
|||
|
|||
Dielectric properties >DBT Debate
Mkuller wrote:
Bruce Abrams : wrote: The existence of bias in product testing has been an accepted fact for decades. That's why there are ALWAYS controls in place and sighted tests are given no validity. Until you accept that as a fact, further argument is pointless. If you wish to convince yourself that you heard something, try hearing it without knowing that you should be. If you're happy simply thinking that you heard something, by all means continue doing what makes you happy. Just don't try to rationalize the decision by trying to convince yourself or anyone else that you know with any degree of certainty what you heard. Only a bias controlled test can prove that. Audio is no more unique in this respect than taste testing, tactile (eg. quality of material in clothing) or other consumer product testing. snip If sighted audio components tests have no validity, 100,000 or so audiophiles would not subscribe to Stereophile and other magazines to read them. Obviously most audiophiles have found them useful in some way. True, a community of believers can impart within-group 'validity' to all sorts of strange and objectively unsupported beliefs. That brand of validation is only 'useful' to those who believe. Yes, biases do exist. However, the problem I have with ABX-type DBTs is that it has never been proven scientifically that: 1. they are appropriate for amateur audio enthusiasts to use for comparing audio components using music as the source (just because DBTs are used very differently in other settings is no proof). 2. they don't remove more than *just bias* in the audio equipment comparisons. Until you can prove this, the rest is speculation and hand waving. It's intersting how people like you wield the cudgel of 'scientific proof' with one hand, and the '100,000 audiophiles can't be wrong' cudgel in the other. From a scientific POV there's simply no need to 'prove' the efficacy of DBT; your point #2 is jsut so much hand-waving, given that DBTs have long been the *established standard* for validating audible difference. Any major problems with DBT, as you conjecture for poitn #2, would have to have been noticed by now, given *decades* of DBT use in psychoacoustic science and in developnent of sound-producing and transmitting devices. But alas there is *no* evidence to support conjecture #2. Therefore the onus is on *you* to provide evidence that point #2 is a real *problem*...and not just a mind game. As for your point #1, if you were *really* interested in scientific proof, you'd be advocating the use of scientific methods for component comparison. Your own writing in TAS and here exhibits NO evidence that you are interested in that at all. I conclude that you aren't in the least interested in scientific proof of audible difference. In the mean time, I would suggest "long term observational listening, with carefully matched levels at switching" which I have described here before. Which has *demonstrated*, not conjectural , flaws. While it may not be perfect at filtering out all biases and preconceptions for everyone - you may get some false positives - you will not get all of the false negatives (i.e. null results) you get from ABX. So take your pick - false positives or false negatives. Except..you haven't demonstrated that the negatives *are* false. You are making serious errors of logic here, repeatedly. You are assuming what you need to prove. What *if* the few rigid objectivists here who insist ABX is the only answer - are wrong, and the near 100,000 audiophiles are right? In their hearts the ABXers must ultimately fear this, because as audiophiles they want the *best possible* music reproduction from their systems. They would discover that the *just-as-good-stuff* actually isn't. They would have been fooling themselves into rationalizing the purchase of cheaper equipment that doesn't sound nearly as good as their flawed tests showed. "What if' does not consistute evidence, or proof, alas. But in the same vein: *WHAT IF* a great many audiophile products really *don't* sound different from their less expensive and glamorous counterparts? In their hearts subjectivists must ultimately fear this. On the other hand, if the rest of the audiophiles are wrong, and ABX is *the one true way*, our systems still sound as great as we think they do - we've just spent a few more bucks (and more time listening to cables) than we needed to. If the emperor really has no clothes...what does it say about the courtiers? -- -S. |
#62
|
|||
|
|||
Dielectric properties
Michael Scarpitti wrote:
(Stewart Pinkerton) wrote in message ... On 14 Oct 2003 17:57:40 GMT, (Michael Scarpitti) wrote: Bruce Abrams wrote in message ... "Michael Scarpitti" wrote in message ... 2) The green pen effect was a *joke*, and there is no physical mechanism by which it *can* work. That's false. Kindly explain, then, exactly what mechanism it is that results in any electrical change that results from the application of a green pen to the edges of a CD. To prove 'bias' has the effect you say, you should try the following (just a suggestion) See Bruce Abrams excellent reply (BTW, a similar 'accident' with amps made the scales fall from my own eyes, back when I was a 'true believer'). Now, how about you answer the question? I am not, nor was I ever a 'true believer'. I was and remain a skeptic. I hear what I hear and pay no attention whatsoever to what anyone else claims. I'm Popeye the sailor man. *You* claim that there *is* a mechanism for the 'green pen effect', so exactly what do *you* claim that it is? I claim my car starts every morning, does that mean I have to explain it to you? Uh...nope! Because there *is* a reasonable existing explanation for why that event would occur. Contrast this with your green pen results -- and with the fact that the initial green pen reports were a *hoax*. -- -S. |
#63
|
|||
|
|||
Dielectric properties >DBT Debate
On 16 Oct 2003 18:10:14 GMT, Steven Sullivan wrote:
Your own writing in TAS and here exhibits NO evidence that you are interested in that at all. I conclude that you aren't in the least interested in scientific proof of audible difference. Actually, doesn't the fact that he gets *paid* for writing purple prose about audible difference, ring a few alarm bells concerning his standpoint?............. -- Stewart Pinkerton | Music is Art - Audio is Engineering |
#64
|
|||
|
|||
Dielectric properties >DBT Debate
|
#65
|
|||
|
|||
Dielectric properties >DBT debate
(Stewart Pinkerton)
You have read Bruce's post, you've read my posts, there's 100 years of evidence that bias exists (and totally swamps any subtle differences that *may* exist between components in any sighted test), and you *still* think that because you heard it, it must be true? Sheesh! Stewart, you keep saying this but refuse to provide any evidence. Wow, a hundred years. That sounds like a lot. So you should be able to provide something in the way of evidence to convince the rest of us. At what point do biases swamp subtle differences between components? The exact point please. Would it be 0.1 db in loudness difference? Or would it be 1.75db of loudness difference? Or are humans so biased that it is every single time, even listening to speakers? How do you measure bias? Let's say I have little or no biases - then I should be able to hear subtle audible differences sighted, right? And then if an ABX test prevents me from hearing the subtle differences, the test must be the problem. Right? Please give us an explaination. Regards, Mike |
#66
|
|||
|
|||
Dielectric properties >DBT Debate
Steven Sullivan wrote:
Your own writing in TAS and here exhibits NO evidence that you are interested in that at all. I conclude that you aren't in the least interested in scientific proof of audible difference. (Stewart Pinkerton) wrote: Actually, doesn't the fact that he gets *paid* for writing purple prose about audible difference, ring a few alarm bells concerning his standpoint?............. Always looking for a conspiracy or a sinister motive, eh Stewart. Do you think trying to impeach me will make you look more credible? First, when I was "getting paid" for writing equipment reviews, (at least someone thought enough of my opinions to pay me) I think the most I ever got for a full review was $300 (2-3 months work). No one does it for the money. After reviewing for 15 years, I gave it up in 1999 when I started my third new business since there is a tremendous amount of time involved in doing careful, thorough reviews (countless hours of listening to and switching equipment). So to be accurate you should say, "used to get *paid* for writing... about...audible differences." Regards, Mike |
#67
|
|||
|
|||
Dielectric properties >DBT Debate
mkuller wrote:
If sighted audio components tests have no validity, 100,000 or so audiophiles would not subscribe to Stereophile and other magazines to read them. Obviously most audiophiles have found them useful in some way. Bruce Abrams wrote: This is a false supposition. The "high-end" industry (and the audiophiles that proudly associate themselves with it) has been predicated on the false premise of sighted component testing. Audiophiles who read Stereophile and other mags have rarely, if ever, been exposed to the notion of blind testing. Wrong. It is a subject discussed many times in Stereophile. I would bet that the majority of audiophiles have engaged in a blind comparison of some kind. However, if they are like me, they were amused by the results but found no real correlation between the test and what they hear under ordinary (sighted) conditions. Yes, biases do exist. However, the problem I have with ABX-type DBTs is that it has never been proven scientifically that: 1. they are appropriate for amateur audio enthusiasts to use for comparing audio components using music as the source (just because DBTs are used very differently in other settings is no proof). 2. they don't remove more than *just bias* in the audio equipment comparisons. Until you can prove this, the rest is speculation and hand waving. Blind testing has been an scientifically proven and accepted element of product testing in virtually every industry in the last 100 years. And blind testing is used in clinical psychometric research. So what? None of these are the same as untrained average audiophiles using a source - music - with questionable sensitivity for blind testing to compare audio components. snip Kindly explain what your objection would be to engaging in the exact same "long term observational listening, with carefully matched levels at switching", only "black boxing" the equipment under review, as I've previously discussed. Nothing would be changed except for the knowledge of which component one was listening to. I have no objection whatsoever to your "black box method". In fact I suspect it would be superior to an ABX test because the listener's brain does not have to switch gears to make a decision on whether an unknown sounds like A or B. It is much closer to the way we perform relaxed observational listening to different components. snip First things first, I've never suggested that ABX is the only answer. Other blind protocols for bias control (such as described above and previously) are certainly better than sighted listening and avoid the potential for false negatives that you claim are inherent in the ABX protocol. Agreed. However, a few other posters here seem to feel that ABX is THE only answer. Regards, Mike |
#68
|
|||
|
|||
Dielectric properties >DBT Debate
|
#69
|
|||
|
|||
Dielectric properties >DBT Debate
|
#70
|
|||
|
|||
Dielectric properties >DBT Debate
(Mkuller) wrote:
Steven Sullivan wrote: True, a community of believers can impart within-group 'validity' to all sorts of strange and objectively unsupported beliefs. That brand of validation is only 'useful' to those who believe. I suspect that's why the Audio Critic continues to be published. From a scientific POV there's simply no need to 'prove' the efficacy of DBT; your point #2 is jsut so much hand-waving, given that DBTs have long been the *established standard* for validating audible difference. Any major problems with DBT, as you conjecture for poitn #2, would have to have been noticed by now, given *decades* of DBT use in psychoacoustic science and in developnent of sound-producing and transmitting devices. Ok, I think we can all agree that dbts are the *established standard* in psychoacoustic science and research for *decades* now. So you never need to say that again. OK, now from a scientific POV how do you get from the above fact to claiming that ABX is the only way for average audiophiles to identify subtle audible differences with audio components using music as a source - since you claim to understand they dbts are used in completely different ways for completely different things in clinical research. snip No one says that ABX is the only way to conduct bias controlled listening tests. It's simply the most elegant and effective way. A coin, a blanket, an inquiring mind and an honest interest in knowing the truth are all that's needed. That's one of the most important aspects of this line of inquiry .... sophisticated equipment and special test facilities are not needed. If the emperor really has no clothes...what does it say about the courtiers? Perhaps it says, "Stewart, I think we have a problem." Regards, Mike Well said if you substitute "Mike" for the first word. |
#71
|
|||
|
|||
Dielectric properties >DBT Debate
|
#72
|
|||
|
|||
Dielectric properties >DBT debate
|
#73
|
|||
|
|||
Dielectric properties >DBT Debate
Mkuller wrote:
mkuller wrote: If sighted audio components tests have no validity, 100,000 or so audiophiles would not subscribe to Stereophile and other magazines to read them. Obviously most audiophiles have found them useful in some way. Bruce Abrams wrote: This is a false supposition. The "high-end" industry (and the audiophiles that proudly associate themselves with it) has been predicated on the false premise of sighted component testing. Audiophiles who read Stereophile and other mags have rarely, if ever, been exposed to the notion of blind testing. Wrong. It is a subject discussed many times in Stereophile. I would bet that the majority of audiophiles have engaged in a blind comparison of some kind. However, if they are like me, they were amused by the results but found no real correlation between the test and what they hear under ordinary (sighted) conditions. Like I said; you really aren't interested in scientific evidence or results or proof. You're mainly interested in *confirming your sighted perceptions* -- which, according to decades of science, run a significant chance of being *wrong*. But you are right that Stereophile *has* considered the matter of bias on occasion...and waffled on it or taken the utterly unsupported stance that you take -- that 'we don't know why' sighted differences often disappear when subjected to ABX controls. See for example the extraodinary editorial contortions, misrerpresentation of blind testing, and demonstrations of bias preserved in stereophile's archives: http://www.stereophile.com/showarchives.cgi?141 wherein statistical arguments offered by a reader are used to raise the possibility of false negatives....leaving aside the point that for *listeners who have already decided they can hear a difference between two components under sighted conditions*, this would not be a relevant factor. To the extent is deals with DBT at all, Stereophile likes to conduct 'big' DBTs, using audiences full of listeners undesr sometimes questionable conditions. It's not necessary. What's necessary is for Stereophile to test its OWN claims made by individuals associated witht eh magazine. "Audiophiles' and reviewers in Stereophile commonly use language that implies *obvious* audible difference. Those should not be swamped by statistics in a DBT. If they are, the logical response is NOT to question the statistical analysis of the DBT, but to question whether the 'obvious' difference existed at all in the first place. Audiophiles claim that with long term listening they can discern the difference between component A and B. THere is no reason why this should disappear during trials like the ones undergone by Steve Zipser at Sunshine audio, where HIS components, and HIS music were being comapred in HIS environment. Stereophile hides behind the 'you haven't tested enough people' canard -- when all that's required of THEM is for THEIR reviewers to have THEIR claims tested. David Carlstrom challenges John Atkinson to do just that, in the exchange of letters archived above. He points out that the ABX box was invented so that audiophiles could more easily test *themselves* to their satisfaction under DBT conditions...thereby addressing the objections that arise when test results obtained for person X are applied to person Y. And blind testing is used in clinical psychometric research. So what? None of these are the same as untrained average audiophiles using a source - music - with questionable sensitivity for blind testing to compare audio components. snip Except, trained 'audiophiles' HAVE been subjected to listening tests using music -- and the results HAVE accorded with the predictions of psychometric research -- namely, that bias often leads listeners astray. Kindly explain what your objection would be to engaging in the exact same "long term observational listening, with carefully matched levels at switching", only "black boxing" the equipment under review, as I've previously discussed. Nothing would be changed except for the knowledge of which component one was listening to. I have no objection whatsoever to your "black box method". In fact I suspect it would be superior to an ABX test because the listener's brain does not have to switch gears to make a decision on whether an unknown sounds like A or B. It is much closer to the way we perform relaxed observational listening to different components. snip There is no reason -- NONE -- that an ABX protocol could not be used with 'long term observational listening'. In any comparison protocol, including sighted, eventually the listener's brain has to 'switch gears' to 'make a decision': does A sound the same or different from B? There is no reason an ABX cannot mimic your preferred method of listening, with only ONE condition altered: that you do not know which component is playing, when you make that decision. Agreed. However, a few other posters here seem to feel that ABX is THE only answer. NO ONE HERE has ever made that claim. However, a few posters here seem to think that sighted perception *hasn't* been demonstrated conclusively to be highly prone to error... notwithstanding decades worth of evidence. -- -S. |
#74
|
|||
|
|||
Dielectric properties >DBT Debate
Mkuller wrote:
Steven Sullivan wrote: True, a community of believers can impart within-group 'validity' to all sorts of strange and objectively unsupported beliefs. That brand of validation is only 'useful' to those who believe. I suspect that's why the Audio Critic continues to be published. Except, the Audio Critic seems willing to subject its claim to scientific testing -- so your suspicion would appear to be unfounded. From a scientific POV there's simply no need to 'prove' the efficacy of DBT; your point #2 is jsut so much hand-waving, given that DBTs have long been the *established standard* for validating audible difference. Any major problems with DBT, as you conjecture for poitn #2, would have to have been noticed by now, given *decades* of DBT use in psychoacoustic science and in developnent of sound-producing and transmitting devices. Ok, I think we can all agree that dbts are the *established standard* in psychoacoustic science and research for *decades* now. So you never need to say that again. One can only hope. OK, now from a scientific POV how do you get from the above fact to claiming that ABX is the only way for average audiophiles to identify subtle audible differences with audio components using music as a source - since you claim to understand they dbts are used in completely different ways for completely different things in clinical research. snip No one here has made that claim about ABX, least of all me. So you need never offer up that strawman again, thanks. And your claims about DBTs being used in 'completely different ways' is belied by the facts that the AES publishes papers that involve DBTs, and that audio manufacturers use DBTs. So someone MUST be making that connection between academic research into audio perception, and identification of differeences between engineered components. It really isn't that big a leap, Mike. If the emperor really has no clothes...what does it say about the courtiers? Perhaps it says, "Stewart, I think we have a problem." Stewart would surely be among those pointing and laughing at the emperor. The courtiers would be *ever* so angry at him. -- -S. |
#75
|
|||
|
|||
Dielectric properties >DBT debate
"Mkuller" wrote in message
... *snip* Or are humans so biased that it is every single time, even listening to speakers? How do you measure bias? You don't measure bias. You eliminate it from the equation via blind testing. Let's say I have little or no biases - then I should be able to hear subtle audible differences sighted, right? And then if an ABX test prevents me from hearing the subtle differences, the test must be the problem. Right? Please give us an explaination. Let's try this...there's a wonderful new anti-depression drug and it's being tested. Everyone who is told what they're taking, and gets the drug reports a dramatic improvement in their symptoms. That's it. No further studies undertaken. Would you take such a medication? The FDA certainly wouldn't approve it. Why? Because of the bias inherent in sighted tests. What if the control group, who got a placebo, reported exactly the same results as the folks getting the drug? You'd have to conclude that the drug was no more effective than the placebo. Translating to the audio world, without the willingness to agree that blind testing is necessary, you have essentially agreed to take the drug which has been proven to be no more effective than a placebo, because you have refused to test it. You have absolutely no evidence whatsoever to show that audio is any different from any other consumer product. |
#76
|
|||
|
|||
Dielectric properties >DBT Debate
"Mkuller" wrote in message
... *snip* Kindly explain what your objection would be to engaging in the exact same "long term observational listening, with carefully matched levels at switching", only "black boxing" the equipment under review, as I've previously discussed. Nothing would be changed except for the knowledge of which component one was listening to. I have no objection whatsoever to your "black box method". In fact I suspect it would be superior to an ABX test because the listener's brain does not have to switch gears to make a decision on whether an unknown sounds like A or B. It is much closer to the way we perform relaxed observational listening to different components. snip If you have no objection to blind testing under my "black box method", it is disingenuous to continue to espouse the notion that only sighted listening can be the final arbiter as to the existence of "subtle audible differences." Why do you suppose that such a method is not used, in fact editorially demanded, by magazines? (Imagine the reaction to a wine taste test where the wines all known in advance. It would have no credibility whatsoever.) The answer can only be that if the possibility of imagined audible differences and preferences is eliminated, and it is "discovered" that there is no such thing as cable sound, and there are no audible differences between "nominally competent" amplifiers, there would be nothing to write about, nothing to advertise, and hence no money to be made. Bruce |
#77
|
|||
|
|||
Dielectric properties >DBT Debate
|
#78
|
|||
|
|||
Dielectric properties >DBT debate
Nousaine wrote:
(Mkuller) wrote (Stewart Pinkerton) You have read Bruce's post, you've read my posts, there's 100 years of evidence that bias exists (and totally swamps any subtle differences that *may* exist between components in any sighted test), and you *still* think that because you heard it, it must be true? Sheesh! Stewart, you keep saying this but refuse to provide any evidence. Wow, a hundred years. That sounds like a lot. So you should be able to provide something in the way of evidence to convince the rest of us. Here's one: "Can You Trust Your Ears" 91st AES Convention Preprint 3177. But if it doesn't confirm Mike's sighted perceptions, it won't convince him -- he's basically *said* as much here, as have Mssrs. Atkinson et. al. in Stereophile. The mantra is : "Thousands of audiophiles can't be wrong'. Whereas science says: of course they can. -- -S. |
#79
|
|||
|
|||
Dielectric properties >DBT Debate
"Bruce Abrams" wrote in message
... "Mkuller" wrote in message ... *snip* Kindly explain what your objection would be to engaging in the exact same "long term observational listening, with carefully matched levels at switching", only "black boxing" the equipment under review, as I've previously discussed. Nothing would be changed except for the knowledge of which component one was listening to. I have no objection whatsoever to your "black box method". In fact I suspect it would be superior to an ABX test because the listener's brain does not have to switch gears to make a decision on whether an unknown sounds like A or B. It is much closer to the way we perform relaxed observational listening to different components. snip If you have no objection to blind testing under my "black box method", it is disingenuous to continue to espouse the notion that only sighted listening can be the final arbiter as to the existence of "subtle audible differences." Why do you suppose that such a method is not used, in fact editorially demanded, by magazines? (Imagine the reaction to a wine taste test where the wines all known in advance. It would have no credibility whatsoever.) The answer can only be that if the possibility of imagined audible differences and preferences is eliminated, and it is "discovered" that there is no such thing as cable sound, and there are no audible differences between "nominally competent" amplifiers, there would be nothing to write about, nothing to advertise, and hence no money to be made. Bruce I have given the black box idea a lot of thought, even before you brought it up. I think it might be worthwhile for comparing two comparable amps of similar size and weight, that is all. For example, tube amps would have to be more vented, so tube amp vs solid state is impractical. And when you have vents, then you most likely have the opportunity to look inside a bit, so cover is blown. Even to keep the speaker hookups honest, you would have to connect the amps to terminals on the box and from there to the speakers, so simple boxes would not do. Finally, somebody would have to come into your house and move the boxes so you couldn't guess at weight, and you'd have to sign an oath to keep "hands off". So any test would have to be extended over time, which brings up how do you rate? My conclusion: you would ask indirect questions as well as direct each week, e.g. did you feel like playing music much this week, when you did play music, were you able to enjoy it, etc. ect. in addition to how do you rate the sound for: (you name it). And then the "mood responses" would have to be averaged over time and statistically analyzed. Certainly a test for the average audiophile making a purchasing decision, right? |
#80
|
|||
|
|||
Dielectric properties >DBT debate
"Nousaine" wrote in message
... (Mkuller) wrote (Stewart Pinkerton) You have read Bruce's post, you've read my posts, there's 100 years of evidence that bias exists (and totally swamps any subtle differences that *may* exist between components in any sighted test), and you *still* think that because you heard it, it must be true? Sheesh! Stewart, you keep saying this but refuse to provide any evidence. Wow, a hundred years. That sounds like a lot. So you should be able to provide something in the way of evidence to convince the rest of us. Here's one: "Can You Trust Your Ears" 91st AES Convention Preprint 3177. At what point do biases swamp subtle differences between components? The exact point please. Would it be 0.1 db in loudness difference? Or would it be 1.75db of loudness difference? They would swamp truly audible differences when subjects were hearing the same thing twice in a row. Why don't you divulge the exact point where open-listening passes signals that are masked when bias controls are implemented? Or are humans so biased that it is every single time, even listening to speakers? How do you measure bias? Let's say I have little or no biases - then I should be able to hear subtle audible differences sighted, right? And then if an ABX test prevents me from hearing the subtle differences, the test must be the problem. Right? Please give us an explaination. If you can "hear" subtle differences that disappear when bias controls are implemented it can ONLY mean that the differences you claimed to hear weren't acoustically based. You keep forgetting to add: "or that the test mechanism utilized does not do a good job of allowing that(those) particular difference(s) to be heard. " Which is another possibility. Just to be complete. Good science and all that, whot? :-) |
Reply |
|
Thread Tools | |
Display Modes | |
|
|
Similar Threads | ||||
Thread | Forum | |||
cabling explained | Car Audio |