Home |
Search |
Today's Posts |
#1
|
|||
|
|||
DBT and science
I'll be brief. I'm a firm believer in DBT for objectively-verifiable
data (such as whether a given drug cures cancer) but given that sound is fugitive, (and that consequently audio perception is inherently non-repeatable, subjective, and not quanifiable) DBT is impossible. Tharefore no debate on whether DBT is useful in audio is warranted. There cannot be any such thing as DBT in audio. One can take two samples of lenses and take pictures with them, which 'fixes' their performance on something outside our bodies: the film. The film or prints made from the film can be studied at leisure and directly compared without having to 'remember' what the one looked like. |
#2
|
|||
|
|||
DBT and science
|
#3
|
|||
|
|||
DBT and science
|
#4
|
|||
|
|||
DBT and science
Michael, I think I see what you are getting at, but don't think
I agree. There is DBT for audio. You may not agree it is the best way or with how sensitive it is versus other ways. But there are double blind tests. Try this scenario. You sign up to be a subject in the local university psychology department's student experiments. You aren't even told what the experiment is testing you for or anything. You are told to sit in a chair until they call your name over the intercom. When called to come into an adjacent room for some testing. Now you don't have to remember your name or what it sounds like. Lets say the real experiment is the aural threshold of hearing your name and responding. They have the 'intercom' speaker behind a black veil put out different odd low level noises. And mix in your name at a known level. Maybe even "Michael we are ready for you". Well if you don't hear it you won't enter the adjacent room. When you do they could have you come and sit for brief 'tests' of your ability to discriminate shades of colored squares at different light levels. Then they send you back into the room to wait for them to ready other color arrangements. You are called again. If you hear it you respond. Do some more tests of color discrimination and wait some more. So on and so forth a few times. Well, they could determine whether or not your threshold were consistent for a few trials. Change to different amplifiers and try that again to get your threshold to hearing your name. And see if it differs with different amplifiers. You would have just participated in a double blind test, assuming the other people doing the test were also blinded. And it has nothing to do with memory etc. It does have to do with your threshold of hearing your name over two differing amplifiers. And whether there is any difference. The color discriminations need not be a real test other than to get your attention elsewhere. You wouldn't even be aware that any listening test was being done. Dennis "Michael Scarpitti" wrote in message . net... I'll be brief. I'm a firm believer in DBT for objectively-verifiable data (such as whether a given drug cures cancer) but given that sound is fugitive, (and that consequently audio perception is inherently non-repeatable, subjective, and not quanifiable) DBT is impossible. Tharefore no debate on whether DBT is useful in audio is warranted. There cannot be any such thing as DBT in audio. One can take two samples of lenses and take pictures with them, which 'fixes' their performance on something outside our bodies: the film. The film or prints made from the film can be studied at leisure and directly compared without having to 'remember' what the one looked like. |
#5
|
|||
|
|||
DBT and science
"Dennis Moore" wrote in message ...
Michael, I think I see what you are getting at, but don't think I agree. There is DBT for audio. You may not agree it is the best way or with how sensitive it is versus other ways. But there are double blind tests. Try this scenario. You sign up to be a subject in the local university psychology department's student experiments. You aren't even told what the experiment is testing you for or anything. You are told to sit in a chair until they call your name over the intercom. When called to come into an adjacent room for some testing. Now you don't have to remember your name or what it sounds like. Lets say the real experiment is the aural threshold of hearing your name and responding. They have the 'intercom' speaker behind a black veil put out different odd low level noises. And mix in your name at a known level. Maybe even "Michael we are ready for you". Well if you don't hear it you won't enter the adjacent room. When you do they could have you come and sit for brief 'tests' of your ability to discriminate shades of colored squares at different light levels. Then they send you back into the room to wait for them to ready other color arrangements. You are called again. If you hear it you respond. Do some more tests of color discrimination and wait some more. So on and so forth a few times. Well, they could determine whether or not your threshold were consistent for a few trials. Change to different amplifiers and try that again to get your threshold to hearing your name. And see if it differs with different amplifiers. You would have just participated in a double blind test, assuming the other people doing the test were also blinded. And it has nothing to do with memory etc. It does have to do with your threshold of hearing your name over two differing amplifiers. And whether there is any difference. The color discriminations need not be a real test other than to get your attention elsewhere. You wouldn't even be aware that any listening test was being done. Dennis Of course there are DBTs in audio. What you describe is a psychometric or psychoacoustic or whatever name the researchers prefer research test. It has a well defined end point: you either hear your name or not. You either hear gaps in frequencies or you don't. You hear the phase change oor you don't. Period. You're doing something different when you compare components blind for the differences in how they reproduce music. The only way you can estimate who is "right" is by statistical criteria: Are there more who do or more who don't and how consistent are they. And there has been enough argument about what statistics are the "right" statistics here in RAHE already. And of course counting if more heard or did not hear the difference is ridiculous because some will hear what others will not. Now you add ABX to the witches' brew and you're doing something not even remotely like the ordinary DBT. First listen to A, then listen to B, then listen to X and compare X with A and B. Different aptitudes, different training. While I'm sure there are oodles of basic research to support the psychometric ABX use , there seems to be zero to support the notion that DBT ABX is a useful test in consumer audio for the average audio consumers. I think this is what Michael Scarpitti is referring to and I think within those limits he is absolutely right. Mixing up these apples , oranges and bananas continues in these column and probably will. Lots of emotional capital behind it... Ludovic Mirabel "Michael Scarpitti" wrote in message . net... I'll be brief. I'm a firm believer in DBT for objectively-verifiable data (such as whether a given drug cures cancer) but given that sound is fugitive, (and that consequently audio perception is inherently non-repeatable, subjective, and not quanifiable) DBT is impossible. Tharefore no debate on whether DBT is useful in audio is warranted. There cannot be any such thing as DBT in audio. One can take two samples of lenses and take pictures with them, which 'fixes' their performance on something outside our bodies: the film. The film or prints made from the film can be studied at leisure and directly compared without having to 'remember' what the one looked like. |
#6
|
|||
|
|||
DBT and science
Actually, what I understood him try to convey was that just because
someone(anyone or any group of grad students for that matter) reached a single conclusion from a single test does not verify the results as etched in stone. This is because on differering occations under different conditions and with different physical or psycological variables, different results could be reached and thus different conclusions could be drawn. As none of those individually could be considered to be truely or more so representative,as the human sense is dynamic, the test would have to go on continuously and throughout all of those variables for some time to have any concrete validity. Now, I don't know about you or Stewart but to me that does not seem so "simplistic" or remotely fun. Even though he did say DBT, I think what he was discussing was the results. Obviously a test can be performed, the key is whether it is relevant. Separating testing the individual and the device is not so easy and I hear everyone go back and forth over what and how things have to be done, what can and can't be heard, etc. with no solid results that everyone can agree upon. One main stumbling block is the equipment used to create a so called "level playing field" for the testing of different types of gear. When many go to great lengths to eliminate these extra controls as they find them to color the sound of signals passed through them, it makes klittle sense to add them other than for convenience. Unfortunately convenience does not validate the results either. Michael seems to speak of the human side of the equation which has it's own variables. I don't think that using a blind test is really irrelevant, but the results could be. So what he seems to be saying is that the single test cannot be said to be conclusive. - Bill www.uptownaudio.com Roanoke VA (540) 343-1250 "Dennis Moore" wrote in message ... Michael, I think I see what you are getting at, but don't think I agree. There is DBT for audio. You may not agree it is the best way or with how sensitive it is versus other ways. But there are double blind tests. Try this scenario. You sign up to be a subject in the local university psychology department's student experiments. You aren't even told what the experiment is testing you for or anything. You are told to sit in a chair until they call your name over the intercom. When called to come into an adjacent room for some testing. Now you don't have to remember your name or what it sounds like. Lets say the real experiment is the aural threshold of hearing your name and responding. They have the 'intercom' speaker behind a black veil put out different odd low level noises. And mix in your name at a known level. Maybe even "Michael we are ready for you". Well if you don't hear it you won't enter the adjacent room. When you do they could have you come and sit for brief 'tests' of your ability to discriminate shades of colored squares at different light levels. Then they send you back into the room to wait for them to ready other color arrangements. You are called again. If you hear it you respond. Do some more tests of color discrimination and wait some more. So on and so forth a few times. Well, they could determine whether or not your threshold were consistent for a few trials. Change to different amplifiers and try that again to get your threshold to hearing your name. And see if it differs with different amplifiers. You would have just participated in a double blind test, assuming the other people doing the test were also blinded. And it has nothing to do with memory etc. It does have to do with your threshold of hearing your name over two differing amplifiers. And whether there is any difference. The color discriminations need not be a real test other than to get your attention elsewhere. You wouldn't even be aware that any listening test was being done. Dennis "Michael Scarpitti" wrote in message . net... I'll be brief. I'm a firm believer in DBT for objectively-verifiable data (such as whether a given drug cures cancer) but given that sound is fugitive, (and that consequently audio perception is inherently non-repeatable, subjective, and not quanifiable) DBT is impossible. Tharefore no debate on whether DBT is useful in audio is warranted. There cannot be any such thing as DBT in audio. One can take two samples of lenses and take pictures with them, which 'fixes' their performance on something outside our bodies: the film. The film or prints made from the film can be studied at leisure and directly compared without having to 'remember' what the one looked like. |
#7
|
|||
|
|||
DBT and science
In article , Michael Scarpitti wrote:
I'll be brief. I'm a firm believer in DBT for objectively-verifiable data (such as whether a given drug cures cancer) but given that sound is fugitive, (and that consequently audio perception is inherently non-repeatable, subjective, and not quanifiable) DBT is impossible. Tharefore no debate on whether DBT is useful in audio is warranted. There cannot be any such thing as DBT in audio. One can take two samples of lenses and take pictures with them, which 'fixes' their performance on something outside our bodies: the film. The film or prints made from the film can be studied at leisure and directly compared without having to 'remember' what the one looked like. I always wonder why those who find flaws in DBT for audio systems do not write up their theories and submit them to the peer reviewed journals that specialize in psycho-acoustics. Your theories would be much better received here with the backing of favorable review. |
#8
|
|||
|
|||
DBT and science
Uptown Audio wrote:
Actually, what I understood him try to convey was that just because someone(anyone or any group of grad students for that matter) reached a single conclusion from a single test does not verify the results as etched in stone. No, he said audio DBT was *impossible* because it involves *memory*. -- -S. |
#9
|
|||
|
|||
DBT and science
Yes, that's a good synopsis. He also alluded to other variables which
cannot be controlled either. Any of which makes it impossible to verify, but not impossible to attempt to make useful. Even his example of a photograph, which is a good one, in that it can be studied at several times and with rest, etc. is still subjective to vision. - Bill www.uptownaudio.com Roanoke VA (540) 343-1250 "Steven Sullivan" wrote in message ... Uptown Audio wrote: Actually, what I understood him try to convey was that just because someone(anyone or any group of grad students for that matter) reached a single conclusion from a single test does not verify the results as etched in stone. No, he said audio DBT was *impossible* because it involves *memory*. -- -S. |
#10
|
|||
|
|||
DBT and science
Perhaps they are not the ones obsessed with dbts.
- Bill www.uptownaudio.com Roanoke VA (540) 343-1250 "Jim West" wrote in message ... In article , Michael Scarpitti wrote: I'll be brief. I'm a firm believer in DBT for objectively-verifiable data (such as whether a given drug cures cancer) but given that sound is fugitive, (and that consequently audio perception is inherently non-repeatable, subjective, and not quanifiable) DBT is impossible. Tharefore no debate on whether DBT is useful in audio is warranted. There cannot be any such thing as DBT in audio. One can take two samples of lenses and take pictures with them, which 'fixes' their performance on something outside our bodies: the film. The film or prints made from the film can be studied at leisure and directly compared without having to 'remember' what the one looked like. I always wonder why those who find flaws in DBT for audio systems do not write up their theories and submit them to the peer reviewed journals that specialize in psycho-acoustics. Your theories would be much better received here with the backing of favorable review. |
#11
|
|||
|
|||
DBT and science
In article , Michael Scarpitti wrote:
That's correct. It's not a 'scientifically' valid test at all. Lens testing performed double-blind (hell, it doesn't even need to be 'blind'!) is a perfect counter-example. You do understand that the results of a test which you administered in non-blind conditions for yourself always giving the results that would lead to your maximum financial gain cannot be offered as evidence that such a test need not be performed double blind? There's no denying the results of such a 'test'. Is such a test 'scientific'? No. You created the images with no independent external review to ensure that those images are the best possible with each device under test. You administered the test under non-blind conditions. You communicated with the test subjects during the test. Any of these would prevent acceptance in a peer-reviewed scientific journal. Frankly, speaking as a consumer, I would never fully trust tests such as the one you described when presented by a place of business. No matter how honest the business, pre-conceived notions and financial incentives can subconsciously (and all too often consciously) affect the outcome. That is why double blind conditions are demanded for rigorous scientific testing. |
#12
|
|||
|
|||
DBT and science
|
#13
|
|||
|
|||
DBT and science
Michael Scarpitti wrote:
Steven Sullivan wrote in message ... Uptown Audio wrote: Actually, what I understood him try to convey was that just because someone(anyone or any group of grad students for that matter) reached a single conclusion from a single test does not verify the results as etched in stone. No, he said audio DBT was *impossible* because it involves *memory*. That's correct. It's not a 'scientifically' valid test at all. You can keep typing this until your keyboard wears out, and it will still be profoundly incorrect. And since your premise is profoundly incorrect, your inferences from it are likely to be as well. Tell me, is it impossible for a blind person to conduct science? Since apparently only visual information can be scientific, I mean. -- -S. |
#14
|
|||
|
|||
DBT and science
(Stewart Pinkerton) wrote:
That's fine, and for gross differences in audio, such as between loudspeakers, you don't need DBTs. This is just ridiculous. What about your "well documented" preconceptions and biases? Do they just disappear when you listen to speakers? How convenient. If you require dbts for any audio component comparison, you should require them for ALL of them to be consistent. At least the dbts with speakers would have a chance at positive results since the differences may be gross. Unlike dbts with other audio components where the flawed test swamps subtle differences. However, for *subtle* differences between audio componnents, sighted listening (such as your Monster comparison) is useless because any *real* differences are swamped by well-documented perceptual bias problems. Perhaps you could provide some of the so-called "well documented" scientific research that shows at exactly what point the audible threshold for perceptual biases swamps *real* differences in sighted listening. This leaves DBTs as the well-proven alternative, capable of resolving subtle - but *real* - sonic differences which are beyond the discriminative capability of sighted testing. I don't believe flawed dbts with music, audio components and untrained audiophiles have the proven sensitivity threshold to be valid for subtle audible differences. IMHO that leaves long term sighted listening (with careful level matching at the switch) as the only alternative for hearing *real* subtle audible differences - those other than gross frequency response and loudness. Regards, Mike |
#15
|
|||
|
|||
DBT and science
Mkuller wrote:
(Stewart Pinkerton) wrote: That's fine, and for gross differences in audio, such as between loudspeakers, you don't need DBTs. This is just ridiculous. What about your "well documented" preconceptions and biases? Do they just disappear when you listen to speakers? How convenient. If you require dbts for any audio component comparison, you should require them for ALL of them to be consistent. At least the dbts with speakers would have a chance at positive results since the differences may be gross. Unlike dbts with other audio components where the flawed test swamps subtle differences. Interesting that you'd attack this point,a nd leave Mr. Scarpatti's rather more amazing assertions uncommented upon. Can I take it that you agree with Mr. Scarpatti's assertions DBT, then? -- -S. |
#16
|
|||
|
|||
DBT and science
Jim West wrote in message ...
In article , Michael Scarpitti wrote: I'll be brief. I'm a firm believer in DBT for objectively-verifiable data (such as whether a given drug cures cancer) but given that sound is fugitive, (and that consequently audio perception is inherently non-repeatable, subjective, and not quanifiable) DBT is impossible. Tharefore no debate on whether DBT is useful in audio is warranted. There cannot be any such thing as DBT in audio. One can take two samples of lenses and take pictures with them, which 'fixes' their performance on something outside our bodies: the film. The film or prints made from the film can be studied at leisure and directly compared without having to 'remember' what the one looked like. I always wonder why those who find flaws in DBT for audio systems do not write up their theories and submit them to the peer reviewed journals that specialize in psycho-acoustics. Your theories would be much better received here with the backing of favorable review. Sit in a dark room. Have someone flash a red light directly into your eyes. Have another person walk in with either the same light or a very similar one and flash the light into your eyes. Now, tell them apart. This is a scientific test? |
#17
|
|||
|
|||
DBT and science
|
#18
|
|||
|
|||
DBT and science
Michael Scarpitti wrote:
Jim West wrote in message ... In article , Michael Scarpitti wrote: I'll be brief. I'm a firm believer in DBT for objectively-verifiable data (such as whether a given drug cures cancer) but given that sound is fugitive, (and that consequently audio perception is inherently non-repeatable, subjective, and not quanifiable) DBT is impossible. Tharefore no debate on whether DBT is useful in audio is warranted. There cannot be any such thing as DBT in audio. One can take two samples of lenses and take pictures with them, which 'fixes' their performance on something outside our bodies: the film. The film or prints made from the film can be studied at leisure and directly compared without having to 'remember' what the one looked like. I always wonder why those who find flaws in DBT for audio systems do not write up their theories and submit them to the peer reviewed journals that specialize in psycho-acoustics. Your theories would be much better received here with the backing of favorable review. Sit in a dark room. Have someone flash a red light directly into your eyes. Have another person walk in with either the same light or a very similar one and flash the light into your eyes. Now, tell them apart. This is a scientific test? Test of *what*? Difference? Recorvery of the eye from glare? Or what? 'Flashing' a red light directly into somone's eye is not analogous to DBT conditions. Try this: Set up an ABX where the viewer can project the light from A or B onto a screen. Allow them to do this until they are confident that a difference exists between A and B (if they never reach that state, then its pointless to continue: the two lights are the same to that viewer). Then project 'X' onto the screen, which is either A or B.Allow the viewer to switch between X, A, and B as often as they like before they decide what X is. Continue in this fashion using a random presentation of A and B as X. Run the statistics on the results. This would be one example of a scientific test analogous to the ABX form of audio DBT. -- -S. ______ "You're an abuser Sullivan....a base beast with intellect but little intelligence to show for it" -- KENNAH! |
#20
|
|||
|
|||
DBT and science
"Uptown Audio" wrote in message ...
Perhaps they are not the ones obsessed with dbts. Then perhaps, kind sir, you might explain why they start such a large proportion and participate so heavily in the threads on such? If they are NOT the ones obsessed with the topic, why are they so obsessed in starting the discussions and participating in them? |
#21
|
|||
|
|||
DBT and science
In article , Michael Scarpitti wrote:
Jim West wrote in message ... I always wonder why those who find flaws in DBT for audio systems do not write up their theories and submit them to the peer reviewed journals that specialize in psycho-acoustics. Your theories would be much better received here with the backing of favorable review. Sit in a dark room. Have someone flash a red light directly into your eyes. Have another person walk in with either the same light or a very similar one and flash the light into your eyes. Now, tell them apart. This is one of the more interesting red herrings I've encountered on UseNet. I do not care to chase it. Do you have a response to my comments? |
#22
|
|||
|
|||
DBT and science
Jim West wrote in message ...
In article , Michael Scarpitti wrote: I'll be brief. I'm a firm believer in DBT for objectively-verifiable data (such as whether a given drug cures cancer) but given that sound is fugitive, (and that consequently audio perception is inherently non-repeatable, subjective, and not quanifiable) DBT is impossible. Tharefore no debate on whether DBT is useful in audio is warranted. There cannot be any such thing as DBT in audio. One can take two samples of lenses and take pictures with them, which 'fixes' their performance on something outside our bodies: the film. The film or prints made from the film can be studied at leisure and directly compared without having to 'remember' what the one looked like. I always wonder why those who find flaws in DBT for audio systems do not write up their theories and submit them to the peer reviewed journals that specialize in psycho-acoustics. Your theories would be much better received here with the backing of favorable review. The objectives (introduced artefact recognition) and selection of subjects is completely different in psychometrics from a "test" promoted as feasible for the run-of -the -mill audio consumers (young and old, amplified music and chamber music listeners etc. etc.) The difference is as much as between the rigorous medical drug research testing with objective bodily changes validation and the question and answer psychoacoustic test. Apples, oranges and bananas. Ludovic Mirabel |
#23
|
|||
|
|||
DBT and science
In article , ludovic mirabel wrote:
Jim West wrote in message ... I always wonder why those who find flaws in DBT for audio systems do not write up their theories and submit them to the peer reviewed journals that specialize in psycho-acoustics. Your theories would be much better received here with the backing of favorable review. The objectives (introduced artefact recognition) and selection of subjects is completely different in psychometrics from a "test" promoted as feasible for the run-of -the -mill audio consumers (young and old, amplified music and chamber music listeners etc. etc.) The difference is as much as between the rigorous medical drug research testing with objective bodily changes validation and the question and answer psychoacoustic test. Apples, oranges and bananas. If you are saying that it is unreasonable to expect every consumer to perform a DBT before making a purchase, I would agree. But making the results of controlled DBTs to those consumers who are interested is another thing altogether. In any event, the poster to whom I responded said that DBT for audio systems is unscientific under any conditions. Postive peer review would certainly bolster this statement for me personally. |
#24
|
|||
|
|||
DBT and science
|
#25
|
|||
|
|||
DBT and science
On 6 Oct 2003 21:58:50 GMT, (Michael
Scarpitti) wrote: (Stewart Pinkerton) wrote in message ... On Fri, 03 Oct 2003 04:46:14 GMT, (Michael Scarpitti) wrote: I'll be brief. I'm a firm believer in DBT for objectively-verifiable data (such as whether a given drug cures cancer) but given that sound is fugitive, (and that consequently audio perception is inherently non-repeatable, subjective, and not quanifiable) DBT is impossible. Tharefore no debate on whether DBT is useful in audio is warranted. There cannot be any such thing as DBT in audio. Your argument is simplistic and incorrect. There certainly *is* such a thing as DBT in audio, it's used every day by many of the major manufacturers. In particular time-proximate ABX testing is less reliant on memory than any other method of sonic comparison, so all you are saying is that while you don't feel that DBTs are valid for audio, you implicitly acknowledge that any other comparison method is inferior. Not necessarily. Hearing represents a peculiar set of problems for testing. No one set of 'tests' or 'method' of testing can be considered conclusive. Answer the point, Scarpitti. You claim that DBTs are flawed because they rely on memory, yet they are *less* reliant on memory than any other method of comparing audio components. In particular, sighted listening is easily shown to be useless for subtle difference, as it invariably gives 'false positives' when *nothing* is actually switched. I can tell you this: I own Yamaha NS-1000M speakers, and I sure can hear the difference between them and my former Rogers Studio 1 speakers. I can tell you this: I used to own Yamaha NS1000M speakers, and I sure can tell them apart from any other speaker. The same is true for every speaker I ever owned. Now, just what has this irrelevance to do with amps and cables? -- Stewart Pinkerton | Music is Art - Audio is Engineering |
#26
|
|||
|
|||
DBT and science
(Stewart Pinkerton) wrote:
That's fine, and for gross differences in audio, such as between loudspeakers, you don't need DBTs. (Mkuller) wrote: This is just ridiculous. What about your "well documented" preconceptions and biases? Do they just disappear when you listen to speakers? Pinkerton Certainly not, but I have never failed to obtain 100% scores on the few occasions when I've tried it with speakers. Absolutely meaningless. Kuller How convenient. If you require dbts for any audio component comparison, you should require them for ALL of them to be consistent. Been there, done that. After a while, you discover two basic truths: 1) All speakers sound different You have not proved this - you say that because some do and you've scored 100% that you can imply they all do. Because I've never heard two amps that sound the same, by the same token I can say they ALL sound different. 2) No cables sound different Wrong. Many do - some are designed to. I hate to quote Firesign Theater and say "everything you know is wrong!", but it looks like everything you promote so vociforously here on RAHE is questionable at best. Pinkerton However, for *subtle* differences between audio componnents, sighted listening (such as your Monster comparison) is useless because any *real* differences are swamped by well-documented perceptual bias problems. Kuller Unlike dbts with other audio components where the flawed test swamps subtle differences. Funny, but you completely ignored this paragraph from my previous post: "Perhaps you could provide some of the so-called "well documented" scientific research that shows at exactly what point the audible threshold for perceptual biases swamps *real* differences in sighted listening." Pinkerton Absolute rubbish! Where is the *flaw* in DBTs, aside from the fact that they don't support your preconceptions? They're used every day in the industry, because they have proven to be the *most* sensitive method for revealing subtle - but *real* - sonic differences. Like the "ideal amplifier" or the "perpetual motion" machine, flawed but *sensitive* dbts are the stuff of objectivist/engineering fantasy. True they may be used by two or three companies out of the few hundreds that produce high end products. So what? They are only as sensitive as the program material, application and protocols makes them; i.e. for audio components, music and amateurs they are flawed because their lack of sensitivity overshadows *real* audible differences resulting in neverending *null* results. Which prove nothing. Regards, Mike I don't believe flawed dbts with music, audio components and untrained audiophiles have the proven sensitivity threshold to be valid for subtle audible differences. IMHO that leaves long term sighted listening (with careful level matching at the switch) as the only alternative for hearing *real* subtle audible differences - those other than gross frequency response and loudness. Mike, this is absolute nonsense, and very easily disproven. Take your 'carefully matched' sighted listening, take it to the max by not actually switching anything, and your audience will still 'hear' lots of differences. That's why sighted listening is *useless* for subtle difference - because it produces false positives with *no* difference. -- Stewart Pinkerton | Music is Art - Audio is Engineering |
#27
|
|||
|
|||
DBT and science
Probably because of the supply, certainly not the demand. It is
difficult to carry-on conversations about audio or for new readers to feel comfortable posting anything because of the ridiclous number of posts by those that have motives other than the enjoyment of music ( I shudder to think of the variety). There should be something along the lines of a tec.audio.measurement group to carry-on that sort of theme in. It has so clearly become of non-interest and a distraction to otherwise polite and helpful posters that the group has had to specifically monitor threads for just that sort of thing. A shame really. Of course everyone is free to either offer other ideas or continue pounding downward this spiral. BTW, thank you for refering to me as a "kind sir". For a minute there, I thought it to be merely personal, sarcastic and a violation of the charter and of the larger groups respect for civility. - Bill www.uptownaudio.com Roanoke VA (540) 343-1250 "Dick Pierce" wrote in message ... "Uptown Audio" wrote in message ... Perhaps they are not the ones obsessed with dbts. Then perhaps, kind sir, you might explain why they start such a large proportion and participate so heavily in the threads on such? If they are NOT the ones obsessed with the topic, why are they so obsessed in starting the discussions and participating in them? |
#28
|
|||
|
|||
DBT and science
(Stewart Pinkerton) wrote in message ...
On 6 Oct 2003 21:58:50 GMT, (Michael Scarpitti) wrote: (Stewart Pinkerton) wrote in message ... On Fri, 03 Oct 2003 04:46:14 GMT, (Michael Scarpitti) wrote: I'll be brief. I'm a firm believer in DBT for objectively-verifiable data (such as whether a given drug cures cancer) but given that sound is fugitive, (and that consequently audio perception is inherently non-repeatable, subjective, and not quanifiable) DBT is impossible. Tharefore no debate on whether DBT is useful in audio is warranted. There cannot be any such thing as DBT in audio. Your argument is simplistic and incorrect. There certainly *is* such a thing as DBT in audio, it's used every day by many of the major manufacturers. In particular time-proximate ABX testing is less reliant on memory than any other method of sonic comparison, so all you are saying is that while you don't feel that DBTs are valid for audio, you implicitly acknowledge that any other comparison method is inferior. Not necessarily. Hearing represents a peculiar set of problems for testing. No one set of 'tests' or 'method' of testing can be considered conclusive. Answer the point, Scarpitti. You claim that DBTs are flawed because they rely on memory, yet they are *less* reliant on memory than any other method of comparing audio components. Not at all. Besides, some things are revealed only with long-term listening. In particular, sighted listening is easily shown to be useless for subtle difference, as it invariably gives 'false positives' when *nothing* is actually switched. I can tell you this: I own Yamaha NS-1000M speakers, and I sure can hear the difference between them and my former Rogers Studio 1 speakers. I can tell you this: I used to own Yamaha NS1000M speakers, and I sure can tell them apart from any other speaker. The same is true for every speaker I ever owned. Now, just what has this irrelevance to do with amps and cables? That one's ears are sensitive. |
#29
|
|||
|
|||
DBT and science
|
#30
|
|||
|
|||
DBT and science
On 7 Oct 2003 22:41:42 GMT, "Uptown Audio"
wrote: Probably because of the supply, certainly not the demand. It is difficult to carry-on conversations about audio or for new readers to feel comfortable posting anything because of the ridiclous number of posts by those that have motives other than the enjoyment of music ( I shudder to think of the variety). There should be something along the lines of a tec.audio.measurement group to carry-on that sort of theme in. Another classic strawman argument, by an audio dealer who must also be presumed to 'have motives other than the enjoyment of music' . Why is it that we keep seeing this 'meter beater' argument, when it's clear that DBTs are a *listening* test? -- Stewart Pinkerton | Music is Art - Audio is Engineering |
#31
|
|||
|
|||
DBT and science
On 7 Oct 2003 22:45:11 GMT, (Michael
Scarpitti) wrote: (Stewart Pinkerton) wrote in message ... On 6 Oct 2003 21:58:50 GMT, (Michael Scarpitti) wrote: (Stewart Pinkerton) wrote in message ... On Fri, 03 Oct 2003 04:46:14 GMT, (Michael Scarpitti) wrote: I'll be brief. I'm a firm believer in DBT for objectively-verifiable data (such as whether a given drug cures cancer) but given that sound is fugitive, (and that consequently audio perception is inherently non-repeatable, subjective, and not quanifiable) DBT is impossible. Tharefore no debate on whether DBT is useful in audio is warranted. There cannot be any such thing as DBT in audio. Your argument is simplistic and incorrect. There certainly *is* such a thing as DBT in audio, it's used every day by many of the major manufacturers. In particular time-proximate ABX testing is less reliant on memory than any other method of sonic comparison, so all you are saying is that while you don't feel that DBTs are valid for audio, you implicitly acknowledge that any other comparison method is inferior. Not necessarily. Hearing represents a peculiar set of problems for testing. No one set of 'tests' or 'method' of testing can be considered conclusive. Answer the point, Scarpitti. You claim that DBTs are flawed because they rely on memory, yet they are *less* reliant on memory than any other method of comparing audio components. Not at all. Besides, some things are revealed only with long-term listening. There's no time limit on DBTs. Answer the point, Scarpitti. You claim that DBTs are flawed because they rely on memory, yet they are *less* reliant on memory than any other method of comparing audio components. And now you want to claim that *long term* comparisons are better? In particular, sighted listening is easily shown to be useless for subtle difference, as it invariably gives 'false positives' when *nothing* is actually switched. I can tell you this: I own Yamaha NS-1000M speakers, and I sure can hear the difference between them and my former Rogers Studio 1 speakers. I can tell you this: I used to own Yamaha NS1000M speakers, and I sure can tell them apart from any other speaker. The same is true for every speaker I ever owned. Now, just what has this irrelevance to do with amps and cables? That one's ears are sensitive. Indeed so, and most sensitive when they operate alone, without prior *knowledge* of which component is connected. Why is it that people like you cry "trust your ears", and yet you refuse to trust *only* your ears? -- Stewart Pinkerton | Music is Art - Audio is Engineering |
#32
|
|||
|
|||
DBT and science
(Mkuller) wrote:
What about your "well documented" preconceptions and biases? Do they just disappear when you listen to speakers? (Stewart Pinkerton) wrote: Certainly not, but I have never failed to obtain 100% scores on the few occasions when I've tried it with speakers. Absolutely meaningless. Nope, it proves the point that the differences are gross. By *your own* criteria, you must admit this. I am becoming tired of your continual use of sophistry to avoid meaningful debate. In fact, I'm becoming tired of this whole ludicrous forum, where ducking, diving and distortion have become the norm for the 'subjectivists'. Stewart - here's the rest of my post you failed to answer - once again. " If you require dbts for any audio component comparison, you should require them for ALL of them to be consistent. At least the dbts with speakers would have a chance at positive results since the differences may be gross. Unlike dbts with other audio components where the flawed test swamps subtle differences. Perhaps you could provide some of the so-called "well documented" scientific research that shows at exactly what point the audible threshold for perceptual biases swamps *real* differences in sighted listening. I don't believe flawed dbts with music, audio components and untrained audiophiles have the proven sensitivity threshold to be valid for subtle audible differences. IMHO that leaves long term sighted listening (with careful level matching at the switch) as the only alternative for hearing *real* subtle audible differences - those other than gross frequency response and loudness." It appears to me you are no longer interested in discussing this topic since I'm raising questions you are unable to answer. My conclusion is that you are not interested in finding the TRUTH ("you can't handle the truth") but in using debating tactics and getting in the last word. Regards, Mike |
#34
|
|||
|
|||
DBT and science
Jim West wrote in message ...
In article , ludovic mirabel wrote: Jim West wrote in message ... I always wonder why those who find flaws in DBT for audio systems do not write up their theories and submit them to the peer reviewed journals that specialize in psycho-acoustics. Your theories would be much better received here with the backing of favorable review. The objectives (introduced artefact recognition) and selection of subjects is completely different in psychometrics from a "test" promoted as feasible for the run-of -the -mill audio consumers (young and old, amplified music and chamber music listeners etc. etc.) The difference is as much as between the rigorous medical drug research testing with objective bodily changes validation and the question and answer psychoacoustic test. Apples, oranges and bananas. If you are saying that it is unreasonable to expect every consumer to perform a DBT before making a purchase, I would agree. But making the results of controlled DBTs to those consumers who are interested is another thing altogether. In any event, the poster to whom I responded said that DBT for audio systems is unscientific under any conditions. Postive peer review would certainly bolster this statement for me personally. Two questions: How will you select your test subjects for a "controlled DBT"? I'd want to know , what age, what training and ABX aptitude they have shown and last but not least what kind of music they've been exposed to. The car audio lovers wouldn't value my opinion. And vice versa. Secondly- please find a reference to ONE SINGLE positive panel component comparison that was using ABX. Such a thing does not exist. They were ALL said by their proctors to have negative outcome. (Review "ABX- is it useful?" and "ABX-the new horizons" threads in RAHE) Ludovic Mirabel. |
#35
|
|||
|
|||
DBT and science
ludovic mirabel wrote:
Jim West wrote in message ... In article , ludovic mirabel wrote: Jim West wrote in message ... I always wonder why those who find flaws in DBT for audio systems do not write up their theories and submit them to the peer reviewed journals that specialize in psycho-acoustics. Your theories would be much better received here with the backing of favorable review. The objectives (introduced artefact recognition) and selection of subjects is completely different in psychometrics from a "test" promoted as feasible for the run-of -the -mill audio consumers (young and old, amplified music and chamber music listeners etc. etc.) The difference is as much as between the rigorous medical drug research testing with objective bodily changes validation and the question and answer psychoacoustic test. Apples, oranges and bananas. If you are saying that it is unreasonable to expect every consumer to perform a DBT before making a purchase, I would agree. But making the results of controlled DBTs to those consumers who are interested is another thing altogether. In any event, the poster to whom I responded said that DBT for audio systems is unscientific under any conditions. Postive peer review would certainly bolster this statement for me personally. Two questions: How will you select your test subjects for a "controlled DBT"? Audiophile magazine reviewers who routinely reports audible differences in amps and cables and CD transports to the public, would be a logical choice. Let's test *their* claims, since their claims drive a significant part of the high-end industry. And what the heck, let's throw in those who strenuously object, here , to skeptical questions about their claims of difference. You included. Any objections, Mr. Mirabel? -- -S. ______ "You're an abuser Sullivan....a base beast with intellect but little intelligence to show for it" -- KENNEH! |
#36
|
|||
|
|||
DBT and science
Steven Sullivan wrote in message ...
ludovic mirabel wrote: Two questions: How will you select your test subjects for a "controlled DBT"? Mr. Sullivan: Audiophile magazine reviewers who routinely reports audible differences in amps and cables and CD transports to the public, would be a logical choice. Let's test *their* claims, since their claims drive a significant part of the high-end industry. And what the heck, let's throw in those who strenuously object, here , to skeptical questions about their claims of difference. You included. Who "asked me skeptical questions" about which my "claims of difference"? How about a quote?- I understand paraphrase is not acceptable in RAHE. Any objections, Mr. Mirabel? None whatsoever. I'm happy you asked me that question. All I have to do is to requote the next sentence in my posting. Exactly where you cut me off. It was: "I'd want to know , what age, what training and ABX aptitude they have shown and last but not least what kind of music they've been exposed to. ". Aptitude first: I happen to be hopeless at ABXing. Panel selection second: Did the reviewers,- whom you say you despise but apparently continue to read, (why on earth?)- claim to be good at your "controlled DBT"? If so when and where? Panel selection third: Would you oblige and quote evidence that your "controlled DBT" ( a cryptonim for ABX, I presume) doesn't interfere with perceptions of many subjects ( including myself and presumably some reviewers) I documented that 80% of "expert audiophiles" in Greenhill's cable test failed to identify 1,75 db volume difference when ABXing and 40% of supertrained professionals in S.Olive's Revelspeakers "listening room" test, and most of his untrained subjects in his loudspeaker test (see the recent thread) similarly failed at relatively simple tasks- recognising frequency bumps and dips and distinguishing unlike loudspeakers. Do you have documentation to the contrary?. So my results on comparing anything whatsoever by ABX would be guaranteed to make you happy:"They all sound the same". Checking it would be a waste of time. But no, no objection. Always anxious to please. But we don't know about you Mr. Sullivan. Could it be that you too get the same answer in all of your ABX research? How would you know when you're wrong if all you have been getting were "It all sounds the same"? Could that be the cause of your irritation with those who rely on other methods? How about trying to listen with your ears without ABX in the way? Who knows? You might get to like it. Ludovic Mirabel |