Home |
Search |
Today's Posts |
#442
|
|||
|
|||
Chung wrote:
Mark DeBellis wrote: What matters, or might well matter to some, is whether the information derived from one source, or its perceptual effect, is the same as that from another source (in blind musical listening). Why should we assume that this reduces to a matter of detecting differences? Because if a difference is not detected by the listener, then the two sources *must* sound the same to the listener. It's really that simple. Well, I *think* my post of Aug. 24, 8:09 pm, contains a counterexample to that claim; please tell me what you think. Probably it is worth stressing that I have temporally extended, or time-distal, cases in mind. To say that two sources "sound the same" to a listener is in a certain way ambiguous. It can mean (1) it seems to the listener that one is just like the other, or (2) the way one source sounds to the listener is just the same as the way the other sounds to the listener. These are not the same thing, because for (1) to be true the listener has to make a judgment of sameness, whereas he need not do so in order for (2) to be true. (2) is true if the properties the listener hears source A as having are the same as the properties he hears source B as having. That is independent of whether he compares the two, whereas (1) is about making a comparison and arriving at a judgment of sameness or difference. In (2), we are saying that sameness holds between what the listener perceives on one occasion and what he perceives on another; in (1), sameness enters into the content of the listener's judgment. Sameness is in a certain sense "external" to the listener's mental state in (2), "internal" in (1). I would resist the claim you make above on either interpretation, actually, because, regarding (1), a person who doesn't detect a difference (in time-distal comparison, say) might simply refrain from judging one way or the other (saying, "I have no idea whether they were the same or different"), and regarding (2), please see the post referred to, because that should illustrate it. (I would also wonder whether "sounding the same" entails having exactly the same perceptual effects, or conveying the same information.) Mark |
#443
|
|||
|
|||
Buster Mudd wrote:
Mark DeBellis wrote: Buster Mudd wrote: I think you keep harping on this notion that our perceptions of these two very different phenomena should somehow be similar. The phenomena (musical memory & audio memory) are similar only in that they are initially triggered by an auditory stimulus; beyond that they are so dissimilar, both in their own content & in the methods by which we perceive them, that to refer to them both as "information we pick up in this way" is to incorrectly characterize them. We *don't* pick up both types of information in the same way once the auditory stimulus has gotten past the ear. I have no idea if our perceptions of them should be similar or different, but given that we have tests for one kind of information, how are we entitled to rely on those tests to tell us that there are no differences in the other kind of information? We're not; but has anyone claimed otherwise? I don't recall anyone in rec.audio.high-end ever asserting that tests...be they ABX, DBT, quick switch, slow switch, monadic, any kind of tests at all...could or couldn't determine differences in our perceptions of *musical content*. Every reference to using statistical testing to identify perceived or imagined differences I've come across here has been pertaining to sonic attributes of audio components. It was the example of SACD vs. CD that I started with, so let me return to that. I thought the idea was that certain tests could tell us that SACD and CD are perceptually equivalent (if they are). If the tests don't tell us that there are no differences in the kind of information that has to do with the longer-term kind of memory, how do they show perceptual equivalence? (By "perceptually equivalent" I mean, more or less, equivalent for all relevant intents and purposes, so that no one could rationally prefer one to the other for its *sonic* attributes.) Mark |
#444
|
|||
|
|||
wrote in message
... Stewart Pinkerton wrote: On 26 Aug 2005 00:45:57 GMT, wrote: Stewart Pinkerton wrote: 'Expectation bias' may EXIST but it is not an EXPLANATION of any given phenomenon. Of course it is. It explains why people think a Krell amp sounds better than a Yamaha. No, it is not an 'EXPLANATION'. You don't understand what an 'explanation' is. Sure I do - you just don't like the explanation. How does a response like this get approved? 'Expectation bias' is NO explanation at all. It is simply assigning a nebulous 'cause', just like saying 'topspin' makes the ball dip. This is true, but insufficient as an EXPLANATION. HOW does topspin make the ball dip? How is of less consequence than the fact that cables DON'T sound different when sound alone is the only tool for identity. Is it so hard to believe that human being simply wish things to sound different, expect them to sound differently, and then perceive it to be so during sighted comparisons? If they really did sound different from each other someone would have been able to tell that in a DBT. The fact is simply that those who have tried have failed. Human psychology being what it is, I have no problem understanding that people believe what they choose, and sometimes perceive what they wish, but the key word is wish. If you trieds the video clip demonstrating the McGurk effect, you should know that hearing is very easily fooled by the eyes. The difference between this and the myth of 'cable sound' is that anyone can strike a ball with spin, and observe the effect. *NO ONE* can listen to two cables and tell them apart by sound alone. I CAN! I HAVE! Then do it again and claim $5K. |
#445
|
|||
|
|||
Jenn wrote:
In article , Chung wrote: Mark DeBellis wrote: Chung wrote: Mark DeBellis wrote: I think you are saying the same thing that I have been trying to say, I don't think so. and you put it very well. It may be that there are differences between the presentations, but you can't detect them. You certainly reduce your ability to detect subtle differences if the presentations are far apart in time. That is why we believe that quick switching is the most effective method. It is others who have (apparently) been arguing that if you can't detect a difference, there can't *be* a difference between the presentations. If you cannot dectect a difference using quick switching under blind conditions, while you previously could detect differences under sighted conditions, then it is very likely that you simply cannot detect differences, regardless of whether there is any detectible difference. What is detectible to someone else may not be to you. And, of course, there are differences between presentations that simply are not detectible. For instance, a 0.01 dB difference in level. By a "difference between two presentations" I would understand this to mean a difference between what you perceive in one and what you perceive in the other, not just that there is some physical difference out there. No, I simply mean some physical difference in the sound waves received by your ears. OK, thank you then for the clarification. I agree with pretty much everything you say. I'm just saying something different. What matters, or might well matter to some, is whether the information derived from one source, or its perceptual effect, is the same as that from another source (in blind musical listening). Why should we assume that this reduces to a matter of detecting differences? Because if a difference is not detected by the listener, then the two sources *must* sound the same to the listener. It's really that simple. Actually to be more fully correct, if a difference is not detected by the listener under the given test conditions, then they must sound the same under those conditions. Actually, the qualification you added does not make my statement any "more truly correct". What you were trying to say is that there might be some test conditions where the listener could detect differences. If that is the case, then the listener simply can detect differences under those test conditions, and of course the two will therefore not sound the same to that listener under those test conditions. Rather obvious, is it not? |
#446
|
|||
|
|||
"Jenn" wrote in message
... In article , Chung wrote: Mark DeBellis wrote: Chung wrote: Mark DeBellis wrote: I think you are saying the same thing that I have been trying to say, I don't think so. and you put it very well. It may be that there are differences between the presentations, but you can't detect them. You certainly reduce your ability to detect subtle differences if the presentations are far apart in time. That is why we believe that quick switching is the most effective method. It is others who have (apparently) been arguing that if you can't detect a difference, there can't *be* a difference between the presentations. If you cannot dectect a difference using quick switching under blind conditions, while you previously could detect differences under sighted conditions, then it is very likely that you simply cannot detect differences, regardless of whether there is any detectible difference. What is detectible to someone else may not be to you. And, of course, there are differences between presentations that simply are not detectible. For instance, a 0.01 dB difference in level. By a "difference between two presentations" I would understand this to mean a difference between what you perceive in one and what you perceive in the other, not just that there is some physical difference out there. No, I simply mean some physical difference in the sound waves received by your ears. OK, thank you then for the clarification. I agree with pretty much everything you say. I'm just saying something different. What matters, or might well matter to some, is whether the information derived from one source, or its perceptual effect, is the same as that from another source (in blind musical listening). Why should we assume that this reduces to a matter of detecting differences? Because if a difference is not detected by the listener, then the two sources *must* sound the same to the listener. It's really that simple. Actually to be more fully correct, if a difference is not detected by the listener under the given test conditions, then they must sound the same under those conditions. Thank you, Jenn, for clarifying the issue so beautifully. As Mark has been arguing, the "conditions" of a quick, switch, comparative double blind abx test are quite different from extended sequential monadic listening done in a normal listening environment through various listening sessions. Since listening to music is subjective and has a strong emotional component, we cannot be sure we are hearing or measuring the same thing (Mark's main point) and we certainly *can* be sure that "under those conditions" are not the same. |
#447
|
|||
|
|||
On 27 Aug 2005 00:19:45 GMT, wrote:
Stewart Pinkerton wrote: On 26 Aug 2005 00:45:57 GMT, wrote: Stewart Pinkerton wrote: 'Expectation bias' may EXIST but it is not an EXPLANATION of any given phenomenon. Of course it is. It explains why people think a Krell amp sounds better than a Yamaha. No, it is not an 'EXPLANATION'. You don't understand what an 'explanation' is. Sure I do - you just don't like the explanation. How does a response like this get approved? The moderator felt it was clearly the truth? :-) 'Expectation bias' is NO explanation at all. It is simply assigning a nebulous 'cause', just like saying 'topspin' makes the ball dip. This is true, but insufficient as an EXPLANATION. HOW does topspin make the ball dip? HOW? HOW does 'expectation bias' produce aural phenomena. HOW? Does it matter, when the *effect* is real in each case? The Bernoulli effect is an explanation of why a tennis ball curves more than gavity predicts when struck with spin. You cannot simply say 'the ball curves because it is struck with spin'. The does not explain the curving path. The difference between this and the myth of 'cable sound' is that anyone can strike a ball with spin, and observe the effect. *NO ONE* can listen to two cables and tell them apart by sound alone. I CAN! I HAVE! No, you can't, and you certainly haven't. If you could, why haven't you collected the $5,000 prize? What 'reality' are you talking about? The real one - all nominally competent cables sound the same. Begs the question. 'Nominally competent' is BY YOUR DEFINITION incapable of sounding different. Nominally competent means that when you measure at the speaker terminals, the cables measure within 0.2dB of each other at 1kHz and 10kHz. That's the *only* precondition, and I'm not aware of any audio cable at any price (plus of course good old Home Depot zipcord) that won't achieve this into any normal loudspeaker. Certainly, I've never had a problem even with totally different constructions such as Litz wire and Naim NACA5. Hence, you are just ducking the issue. The question is rather DO ANY cables or amplifiers sound different. The answer is blatantly, obviously, incontrovertibly, YES! Some amplifiers do, but no cable (see the only condition above) sounds different from any other cable. To claim that it does, is utter nonsense. Common sense says they should, physics and electrical engineering predicts that they do, and no one has *ever* demonstrated an ability to hear differences under blind conditions. Irrelevant. Sure it's relevant. Which part of 'no one has *ever* demonstrated an ability to hear differences under blind conditions' is not utterly and totally relevant to your false claim that cables sound different? You are making an extraordinary claim. Prove your claim, or stop all this nonsense. No, YOU are making an extraordinary claim. Simply repeating this nonsense will *never* make it true. -- Stewart Pinkerton | Music is Art - Audio is Engineering |
#448
|
|||
|
|||
In article , Chung
wrote: Jenn wrote: In article , Chung wrote: Mark DeBellis wrote: Chung wrote: Mark DeBellis wrote: I think you are saying the same thing that I have been trying to say, I don't think so. and you put it very well. It may be that there are differences between the presentations, but you can't detect them. You certainly reduce your ability to detect subtle differences if the presentations are far apart in time. That is why we believe that quick switching is the most effective method. It is others who have (apparently) been arguing that if you can't detect a difference, there can't *be* a difference between the presentations. If you cannot dectect a difference using quick switching under blind conditions, while you previously could detect differences under sighted conditions, then it is very likely that you simply cannot detect differences, regardless of whether there is any detectible difference. What is detectible to someone else may not be to you. And, of course, there are differences between presentations that simply are not detectible. For instance, a 0.01 dB difference in level. By a "difference between two presentations" I would understand this to mean a difference between what you perceive in one and what you perceive in the other, not just that there is some physical difference out there. No, I simply mean some physical difference in the sound waves received by your ears. OK, thank you then for the clarification. I agree with pretty much everything you say. I'm just saying something different. What matters, or might well matter to some, is whether the information derived from one source, or its perceptual effect, is the same as that from another source (in blind musical listening). Why should we assume that this reduces to a matter of detecting differences? Because if a difference is not detected by the listener, then the two sources *must* sound the same to the listener. It's really that simple. Actually to be more fully correct, if a difference is not detected by the listener under the given test conditions, then they must sound the same under those conditions. Actually, the qualification you added does not make my statement any "more truly correct". What you were trying to say is that there might be some test conditions where the listener could detect differences. If that is the case, then the listener simply can detect differences under those test conditions, and of course the two will therefore not sound the same to that listener under those test conditions. Rather obvious, is it not? Obvious, yes, but left out of your paragraph. Your paragraph would seem to indicate that if a listener hears no differences under a specific set of conditions, then the listener will hear no differences under any conditions. This, of course, is not a given. |
#449
|
|||
|
|||
In article ,
"Harry Lavo" wrote: "Jenn" wrote in message ... In article , Chung wrote: Mark DeBellis wrote: Chung wrote: Mark DeBellis wrote: I think you are saying the same thing that I have been trying to say, I don't think so. and you put it very well. It may be that there are differences between the presentations, but you can't detect them. You certainly reduce your ability to detect subtle differences if the presentations are far apart in time. That is why we believe that quick switching is the most effective method. It is others who have (apparently) been arguing that if you can't detect a difference, there can't *be* a difference between the presentations. If you cannot dectect a difference using quick switching under blind conditions, while you previously could detect differences under sighted conditions, then it is very likely that you simply cannot detect differences, regardless of whether there is any detectible difference. What is detectible to someone else may not be to you. And, of course, there are differences between presentations that simply are not detectible. For instance, a 0.01 dB difference in level. By a "difference between two presentations" I would understand this to mean a difference between what you perceive in one and what you perceive in the other, not just that there is some physical difference out there. No, I simply mean some physical difference in the sound waves received by your ears. OK, thank you then for the clarification. I agree with pretty much everything you say. I'm just saying something different. What matters, or might well matter to some, is whether the information derived from one source, or its perceptual effect, is the same as that from another source (in blind musical listening). Why should we assume that this reduces to a matter of detecting differences? Because if a difference is not detected by the listener, then the two sources *must* sound the same to the listener. It's really that simple. Actually to be more fully correct, if a difference is not detected by the listener under the given test conditions, then they must sound the same under those conditions. Thank you, Jenn, for clarifying the issue so beautifully. As Mark has been arguing, the "conditions" of a quick, switch, comparative double blind abx test are quite different from extended sequential monadic listening done in a normal listening environment through various listening sessions. Since listening to music is subjective and has a strong emotional component, we cannot be sure we are hearing or measuring the same thing (Mark's main point) and we certainly *can* be sure that "under those conditions" are not the same. Indeed true, in my opinion. Listening to music, for most people, is largely a right-brain based experience. For good discussions of this, see Gardner: Frames of Mind and on a less technical level, Kerman: Listen. At the end of the day, I think that the left brain vs. right brain ways of dealing with the world might well be the basis for the heated arguments on this topic. For example, a person who experiences music in a highly right-brain oriented way, would, I believe, have a great deal of difficulty during the experience of a quick switch test. I hope to experience such a test soon which will help inform this thesis. Thanks for your kind words, Jenn (listening to a beautiful LP of the Netherlands Wind Ensemble doing the Stravinsky Octet. Ahhhh....) |
#450
|
|||
|
|||
wrote:
wrote: wrote: No, it is not an 'EXPLANATION'. Well, it's a heck of a lot more of an explanation than you've got. This is where you're mistaken, and you refuse to admit it. Your "explanation," which you haven't bothered to share with us yet, would require the rewriting of physics and/or psychology textbooks. ???? I am unable to follow you here. If 'X' causes 'Y' in an observer (Q), and observer Q reports hearing 'Y' whenever listening to 'X', and we conclude that 'X' is the source of 'Y', then no further proof is warranted, because we have located the cause of 'Y' as outside of 'Q'. If, however, you argue that 'Q' is the source of 'Y', you have to establish why (in detail) 'Y' always appears when exposed to 'X', and perhaps never with exposure to 'Z' or 'B'. The fact that you have located the source of 'Y' within 'Q' requires MORE explanation that if it is located within 'X'. Do you understand why that is? IOf similar causes produce similar effects, and 'X' is the sourcve of 'Y', then there is no need for further inquiry. The cable is the source of the sound difference, so naturally when the cable is withdrawn and replaced, the sound difference comes and goes with it. It is simple direct causation. If, however, the observer is the source of the sound difference , and nonetheless the sound difference also vanishes when the cable is withdrawn and reappears when it is replaced (and at no other time and with no other product), you have some explaining to do, to account for HOW this might occur. Simply claiming 'expectation bias' won't do, not even for a mere discussion. It's an insufficient explanation. Sigh. I am tired of being taking to task by a man who apparently hasn't the slightest grasp of the standards necessary to produce reliable, repeatable results in the field he is talking about. First of all, you have no data. None. What DATA are you talking about? This is not and cannot be scientific experiment. I am talking about my experiences and the reports of others with similar experiences. Your assertion that you heard something, and then a reviewer reported "the same thing" is meaningless, because you haven't established either: 1) a clear standard for what would constitute "the same thing"; or 2) that you aren't both influenced by, for example, the same manufacturer's claims for the product. What claims? I don't pay any attention to ads. The claims would not, in any case, be translatable to anything that I heard. No ad has EVER claimed that cables or amps sound anything like what I have heard in comparative listening. The differences are complex, rich, and difficult to describe. No 'ad' has ever led me to expect anything like what I heard, or described anything like what I heard. I cannot describe exactly what I heard, to be frank. I CAN, however, tell THAT I heard a difference, and that I preferred one to the other. In fact, in some cases, my experiences were quite contrary to my 'expectations'. The Harman-Kardon power amplifier which I auditioned sounded quite flat and lacking in dynamics, in spite of the fact that the salesman told me that that was its strong point. Several other amps which I auditioned exhibited a more dynamic sound. If 'expectation bias' is your explanation for my experience, you need to account for its failure to act in many instances. You also ignore the problem that it is physically impossible to repeat a sighted listening test. How do you know what I have done? I sit in the dark and listen carefully to a passage with which I am very familiar. I switch the cables or amplifiers and listen to the same passage, and try to note what, if any, differences present themselves. I own several sets of interconnect cable, bought over a period of years. I had settled on a pair of $50 Monster cable interconnects, simply because they were longer than the others I had, and I needed the length. When I brought home the $100 Monster cable interconnect (same length), I listened to several of my other cables as well, just to see what was what. There was no significant difference among several of the other cables, including custom-made ones, and the $50 Monster interconnect. When I put in the $100 ones, I heard a difference. The difference disappeared when I switched them out. I listened to all of the cables in the dark, because I can hear better in the dark. The reason for this is simple: The second time you do the comparison, you have more "information" about the two cables than you had the first time--specifically, you now know the result of the first comparison, and that cannot help but affect your subsequent comparisons. If you can't see why, ask any psychologist. Unprovable, at best. Now please, please go learn something--at least a little--about experimental psychology, it's methods and standards of scientific evidence. Then you'll understand why all your little stories mean nothing. This is not a scientific experiment, sir. It is a product comparison. I have to satisfy NO-ONE but myself. Any comparisons I make are made in such a way that maximizes the authenticity of the comparisons. I do not practice astrology, nor have I have ever been abducted by aliens. I have a degree in philosophy, and I am well-read in the philosophy of science. You are discussing this with someone far more expert in these matters than you are. A typical contemporary work that you may want to look at is: "Philosophical Explanations" by Robert Nozick Saying that I heard what I 'expected' because of 'expectation bias' is a miserable tautology. It proves nothing. It explains nothing. It's just like phlogiston. You need to come up with a rich explanation that explains the phenomena without assigning the cause to the cables or amps AND excludes the possibility that the cables or amps sound different because of something inherent in the cables or amps. EVEN IF 'expectation bias' is operative in some instances, it cannot account for all of the phenomena with which I am acquainted, especially including experiences that are odds with my expectations. 'Expectation bias' is not used in the literature the way you use it. Your argument is therfore supsect right from the start. |
#451
|
|||
|
|||
Jenn wrote:
In article , Chung wrote: Jenn wrote: In article , Chung wrote: Mark DeBellis wrote: Chung wrote: Mark DeBellis wrote: I think you are saying the same thing that I have been trying to say, I don't think so. and you put it very well. It may be that there are differences between the presentations, but you can't detect them. You certainly reduce your ability to detect subtle differences if the presentations are far apart in time. That is why we believe that quick switching is the most effective method. It is others who have (apparently) been arguing that if you can't detect a difference, there can't *be* a difference between the presentations. If you cannot dectect a difference using quick switching under blind conditions, while you previously could detect differences under sighted conditions, then it is very likely that you simply cannot detect differences, regardless of whether there is any detectible difference. What is detectible to someone else may not be to you. And, of course, there are differences between presentations that simply are not detectible. For instance, a 0.01 dB difference in level. By a "difference between two presentations" I would understand this to mean a difference between what you perceive in one and what you perceive in the other, not just that there is some physical difference out there. No, I simply mean some physical difference in the sound waves received by your ears. OK, thank you then for the clarification. I agree with pretty much everything you say. I'm just saying something different. What matters, or might well matter to some, is whether the information derived from one source, or its perceptual effect, is the same as that from another source (in blind musical listening). Why should we assume that this reduces to a matter of detecting differences? Because if a difference is not detected by the listener, then the two sources *must* sound the same to the listener. It's really that simple. Actually to be more fully correct, if a difference is not detected by the listener under the given test conditions, then they must sound the same under those conditions. Actually, the qualification you added does not make my statement any "more truly correct". What you were trying to say is that there might be some test conditions where the listener could detect differences. If that is the case, then the listener simply can detect differences under those test conditions, and of course the two will therefore not sound the same to that listener under those test conditions. Rather obvious, is it not? Obvious, yes, but left out of your paragraph. Your paragraph would seem to indicate that if a listener hears no differences under a specific set of conditions, then the listener will hear no differences under any conditions. This, of course, is not a given. I would then suggest that you are not reading carefully and are drawing erroneous inferences. And, of course, another obvious point is that if you really want to find out whether two pieces of equipment, or two presentations, sound the same, then you should try to detect differences using the most sensitive conditions for detection: quick switching, double-blind, tests using the most revealing source material. |
#452
|
|||
|
|||
Chung wrote:
And, of course, another obvious point is that if you really want to find out whether two pieces of equipment, or two presentations, sound the same, then you should try to detect differences using the most sensitive conditions for detection: quick switching, double-blind, tests using the most revealing source material. This is false. It takes me a while to hear all the differences bewteen products. Different program material presents different aspects for evaluation. Quick switching is worthless, in my experience. |
#453
|
|||
|
|||
|
#454
|
|||
|
|||
Harry Lavo wrote:
"Jenn" wrote in message the same to the listener. It's really that simple. Actually to be more fully correct, if a difference is not detected by the listener under the given test conditions, then they must sound the same under those conditions. Thank you, Jenn, for clarifying the issue so beautifully. As Mark has been arguing, the "conditions" of a quick, switch, comparative double blind abx test are quite different from extended sequential monadic listening done in a normal listening environment through various listening sessions. Since listening to music is subjective and has a strong emotional component, we cannot be sure we are hearing or measuring the same thing (Mark's main point) and we certainly *can* be sure that "under those conditions" are not the same. Also ... 1. One current sticking point is whether it makes sense to say that someone is perceiving different things but doesn't perceive a difference. I know some of you think I am confused about this, but I don't think I'm confused. I think there is a genuine distinction to be made here that sometimes gets sloughed over by ordinary language. Why is this relevant to audio? Because the thought (FWIW) is that you could listen to source A, which is a 5-minute song on SACD, then source B, a 5-minute song on CD, and perceive different things, yet at the end of it all not be reliably able to judge *that* they are different (i.e., reliably "perceive a difference"). And if the answer to this is, "DeBellis, you're just confused; it doesn't even make sense to say that you could hear different things, i.e., properties, unless you can tell the sources apart in those circumstances," well, I don't think I'm confused or that it doesn't make sense. And I've given a reason for this in the example of time-distal comparison of sounds of different loudness. 2. OK, suppose the response is then, but you can do short-term time-proximate comparisons that *will* tell us if source A "sounds the same" as source B to the listener. My point is simply that the goal here must be for the tests to tell you whether the information or perceptual effect of A over the 5-minute span is the same as B over the 5-minute span; and whether a short-term time-proximate test tells us that will depend on interesting facts of empirical psychology. It is not just a matter of logic. It's not that, "Well, if they sound the same they sound the same," because we are talking about how things "sound" in different circumstances. If short-term time-proximate tests tell us what they are said to tell us, then, among other things, it can't be the case that small, indiscernible effects accumulate over time in a way that makes for a difference in what things we perceive. Now I happen to think that that's interesting, and sorry if anyone else doesn't. The validity of short-term time-proximate comparisons (to show what they're purported to show) depends on certain empirical conditions being true; it's not just a matter of logic or a tautology about how things "sound." The question then arises as to how we know that those conditions are true, and again this seems to me to be an interesting question. In connection with audio tests, I get the impression, though I could well be wrong, that the main attention has been focused on momentary properties of the signal rather than with properties perceived over longer spans (I suspect that the latter aren't ordinarily even considered), and that the notion of "perceiving different things" has not been separated from the notion of "perceiving a difference." Hence (a) there is not much, if any, evidence that perception over longer spans fails to be predicted by short-term time-proximate tests, and (b) that's not too surprising, because the tests aren't really probing for such evidence. Mark |
#455
|
|||
|
|||
Chung wrote:
And, of course, another obvious point is that if you really want to find out whether two pieces of equipment, or two presentations, sound the same, then you should try to detect differences using the most sensitive conditions for detection: quick switching, double-blind, tests using the most revealing source material. The question is whether presentation A (which may last 5 minutes, say) supplies to a certain listener exactly the same information as presentation B, or has the same perceptual effect as B. The problem of finding out the answer to this is not exactly the same problem as finding out if there are sonic differences between A and B. Maybe there are tons of sonic differences that a given test will pick up; and maybe for this reason it is the most sensitive test overall. Still, there may be a certain class of differences, ones that are in fact psychologically important, that the test does not pick up and was not designed to pick up. So the question has to be, what is the best test for the purpose at hand, a question of psychology, not what is the most sensitive test for telling if two signals are the same or different. Mark |
#456
|
|||
|
|||
Stewart Pinkerton wrote:
No, you can't, and you certainly haven't. If you could, why haven't you collected the $5,000 prize? Say, what is this $5,000 test, anyway? :-) Since there is a fairly high probability that, after enough trials, *someone* will appear to distinguish the cables even though the outcome is just chance -- -- in the same way that, if you flip a coin enough times, eventually you're likely to get 10 heads in a row -- -- why don't we get together, all enter, and agree to split the money? Has anybody figured the odds, anyway? What does it cost to enter? :-) Mark |
#457
|
|||
|
|||
|
#458
|
|||
|
|||
|
#459
|
|||
|
|||
|
#460
|
|||
|
|||
Jenn wrote:
Indeed true, in my opinion. Listening to music, for most people, is largely a right-brain based experience. For good discussions of this, see Gardner: Frames of Mind and on a less technical level, Kerman: Listen. At the end of the day, I think that the left brain vs. right brain ways of dealing with the world might well be the basis for the heated arguments on this topic. For example, a person who experiences music in a highly right-brain oriented way, would, I believe, have a great deal of difficulty during the experience of a quick switch test. I hope to experience such a test soon which will help inform this thesis. Yes Jenn, that is the only way. Maybe you have noticed that much of this mental noise has been created by people who have never experienced a properly conducted DBT, ABX or whatever variant. Even the terminology is interpreted in different(erratic) ways. Thanks for your kind words, Jenn (listening to a beautiful LP of the Netherlands Wind Ensemble doing the Stravinsky Octet. Ahhhh....) This is how I'm listening to a test. I do not concentrate on certain instruments, but try to hear into the space in front and around me. And at the same time understand what the composer wanted to express: his joy or sadness, his love, desperation, celebration, meditation... Somehow also the musician feels the same and adds to the initial expression of the composer his interpretation. And then when I feel like switching, the performance goes on without interruption. The piece continues on the same spot. A little "click" doesn't dusturb me, but otherwise it is important that both streams are in sync and there is no perceivable silence. This is a quick switch test. And what is short is not the music snippet, but the switching action. Compare that with changing the speaker cables or whatever in a long break etc. Even a *one* second break is already disturbing and brings me back into the brain... And you want to know how it sounds when *not* in the brain. This is how from my experience I'm able to detect very miniscule differences between two feeds. -- ciao Ban Bordighera, Italy |
#461
|
|||
|
|||
Mark DeBellis wrote:
wrote: 2) We could subject you to a blind test, to determine whether you can indeed distinguish between these cables when you do not know which is which. This could eliminate imagined sonic difference as an explanation (although, admittedly, it cannot confirm it). Not to quibble, and I don't disagree with what you are basically saying in this post, but aren't you entitled to claim more here than you do? If he fails to distinguish the cables, then doesn't that confirm that the difference he heard was imaginary? Alas, statistics doesn't let you draw that conclusion. We set up a null hyothesis--"You cannot hear a difference between these cables"--and set a significance threshold. If he tops the threshold, you've disproved the null hypothesis. But if he doesn't top the threshold, all you've done is failed to disprove the null hypothesis. This has nothing to do with audio, BTW; it's a law of statistics, and affects any statistical test. IOW, one negative test doesn't prove anything. The case against "cable sound" and such is that no one has *ever* succeeded in getting a positive result. After a while, it becomes reasonable to draw the inference that no one ever will. (Reinforced by the lack of any known physical effect that could produce a positive result.) bob |
#462
|
|||
|
|||
Mark DeBellis wrote:
wrote: Saying that I heard what I 'expected' because of 'expectation bias' is a miserable tautology. It proves nothing. It explains nothing. It's just like phlogiston. You need to come up with a rich explanation that explains the phenomena without assigning the cause to the cables or amps AND excludes the possibility that the cables or amps sound different because of something inherent in the cables or amps. EVEN IF 'expectation bias' is operative in some instances, it cannot account for all of the phenomena with which I am acquainted, especially including experiences that are odds with my expectations. 'Expectation bias' is not used in the literature the way you use it. Your argument is therfore supsect right from the start. Let me describe an experience I had awhile back, for what it's worth. I made copies of an audio track onto different brands of CD-R, and at different recording speeds (4x, etc.). And I became convinced that the sound on one was different from the sound on another. One was darker, the other had more "ping." (Hard to put it into words.) Then I listened randomly to one disc or the other, blind. And much of the time (though not necessarily always) I heard the characteristic "darkness" or "ping," but it often turned out that it was the "other" disc that was playing. So I do think that there is a certain tendency to hear differences that are not there, a tendency of the ear or mind to make things sound different when they are not in that respect different. (I wasn't aware of this phenomenon, or at least aware that it could be so definite and vivid, until I encountered it in audio.) Whether this should be called "expectation bias" is a good question, because I have no idea how any particular expectations could have influenced my initial assignment of qualities to one brand of disc (or speed) or another. But once I had established the association, then, yes, I think I had certain expectations as to how it would sound and that this probably helped to cause me to hear the quality I expected to hear. Something I found especially interesting was that there was nothing about the experience of the "darkness" or "ping" to make me think that anything but detection of a quality "out there" was going on. It *felt* just like ordinary perception (whatever that means). Now *of course* I don't know if anything like what went on in the experience I reported is going on in your experience of cable differences. Still, it's a question that could be asked: might it be? Is that a possibility you might want to eliminate, just out of general interest? I don't know what the psychological explanation is for why I heard one disc as "dark" and another as "pingy," but I do know that this sort of illusory phenomenon does occur. You may be right that nobody can offer a full-blown explanation of *why* the differences you hear are somehow caused by expectation, but blind testing could help to show *that* such a phenomenon is, or is not, occurring. Mark Hi Mark, I've been interested in blind testing for a while. I posted about this earlier this year. I think there is a fundamental problem in comparing A to B, in audio: that humans naturally shift their attention to new aspects of the sound when hearing the same bit of music more than once. Suppose I listen to track 1 of some CD, with amplifier A. Then I hook up Amplifier B and listen again. Listening to A, I might have heard things like the quality of the bassline, the smoothness of the midrange, etc. Then I start to listen to B. I have been programmed with certain expectations and a memory of the last details I heard--and, although I've experimented with many styles of blind tests, I have not found a way to escape those expectations. And, listening to a new piece of music twice, I naturally hear different things the second time. Maybe this time I hear the counterpoint. That doesn't necessarily have to do with the amplifier--it is just a normal phenomenon. Much classical music uses repeats: a section of music is played twice. Why isn't this boring? Part of the reason is this normal phenomenon that we shift our attention to new things. Some of the 'objectivists' here advocate quick switching and using very short excerpts. That does seem to get around this problem of shifting attention, but on the other hand, one is no longer listening to the music as music. I suspect that the audible qualities of equipment manifest themselves in the experience of the music--in the way our bodies move to music, in the enjoyment of the music, and so on. So is there any way to conduct a blind test that listens to music as music? Perhaps the test in which one lives with a component for a while, a component that is a "black box" of unknown make. Mike |
#463
|
|||
|
|||
In article ,
"Mark DeBellis" wrote: Harry Lavo wrote: "Jenn" wrote in message the same to the listener. It's really that simple. Actually to be more fully correct, if a difference is not detected by the listener under the given test conditions, then they must sound the same under those conditions. Thank you, Jenn, for clarifying the issue so beautifully. As Mark has been arguing, the "conditions" of a quick, switch, comparative double blind abx test are quite different from extended sequential monadic listening done in a normal listening environment through various listening sessions. Since listening to music is subjective and has a strong emotional component, we cannot be sure we are hearing or measuring the same thing (Mark's main point) and we certainly *can* be sure that "under those conditions" are not the same. Also ... 1. One current sticking point is whether it makes sense to say that someone is perceiving different things but doesn't perceive a difference. I know some of you think I am confused about this, but I don't think I'm confused. I think there is a genuine distinction to be made here that sometimes gets sloughed over by ordinary language. Why is this relevant to audio? Because the thought (FWIW) is that you could listen to source A, which is a 5-minute song on SACD, then source B, a 5-minute song on CD, and perceive different things, yet at the end of it all not be reliably able to judge *that* they are different (i.e., reliably "perceive a difference"). And if the answer to this is, "DeBellis, you're just confused; it doesn't even make sense to say that you could hear different things, i.e., properties, unless you can tell the sources apart in those circumstances," well, I don't think I'm confused or that it doesn't make sense. And I've given a reason for this in the example of time-distal comparison of sounds of different loudness. 2. OK, suppose the response is then, but you can do short-term time-proximate comparisons that *will* tell us if source A "sounds the same" as source B to the listener. My point is simply that the goal here must be for the tests to tell you whether the information or perceptual effect of A over the 5-minute span is the same as B over the 5-minute span; and whether a short-term time-proximate test tells us that will depend on interesting facts of empirical psychology. It is not just a matter of logic. It's not that, "Well, if they sound the same they sound the same," because we are talking about how things "sound" in different circumstances. If short-term time-proximate tests tell us what they are said to tell us, then, among other things, it can't be the case that small, indiscernible effects accumulate over time in a way that makes for a difference in what things we perceive. Now I happen to think that that's interesting, and sorry if anyone else doesn't. The validity of short-term time-proximate comparisons (to show what they're purported to show) depends on certain empirical conditions being true; it's not just a matter of logic or a tautology about how things "sound." The question then arises as to how we know that those conditions are true, and again this seems to me to be an interesting question. In connection with audio tests, I get the impression, though I could well be wrong, that the main attention has been focused on momentary properties of the signal rather than with properties perceived over longer spans (I suspect that the latter aren't ordinarily even considered), and that the notion of "perceiving different things" has not been separated from the notion of "perceiving a difference." Hence (a) there is not much, if any, evidence that perception over longer spans fails to be predicted by short-term time-proximate tests, and (b) that's not too surprising, because the tests aren't really probing for such evidence. Mark Mark, what you write is very interesting. I'm current doing some reading in the current and back issues of a scholarly journal called "Music Perception" (University of California Press), and some of the pieces there are applicable to our questions about audio listening, I believe. I'll report here about what I learn, as I learn it. |
#464
|
|||
|
|||
In article , Chung
wrote: Jenn wrote: In article , Chung wrote: Jenn wrote: In article , Chung wrote: Mark DeBellis wrote: Chung wrote: Mark DeBellis wrote: I think you are saying the same thing that I have been trying to say, I don't think so. and you put it very well. It may be that there are differences between the presentations, but you can't detect them. You certainly reduce your ability to detect subtle differences if the presentations are far apart in time. That is why we believe that quick switching is the most effective method. It is others who have (apparently) been arguing that if you can't detect a difference, there can't *be* a difference between the presentations. If you cannot dectect a difference using quick switching under blind conditions, while you previously could detect differences under sighted conditions, then it is very likely that you simply cannot detect differences, regardless of whether there is any detectible difference. What is detectible to someone else may not be to you. And, of course, there are differences between presentations that simply are not detectible. For instance, a 0.01 dB difference in level. By a "difference between two presentations" I would understand this to mean a difference between what you perceive in one and what you perceive in the other, not just that there is some physical difference out there. No, I simply mean some physical difference in the sound waves received by your ears. OK, thank you then for the clarification. I agree with pretty much everything you say. I'm just saying something different. What matters, or might well matter to some, is whether the information derived from one source, or its perceptual effect, is the same as that from another source (in blind musical listening). Why should we assume that this reduces to a matter of detecting differences? Because if a difference is not detected by the listener, then the two sources *must* sound the same to the listener. It's really that simple. Actually to be more fully correct, if a difference is not detected by the listener under the given test conditions, then they must sound the same under those conditions. Actually, the qualification you added does not make my statement any "more truly correct". What you were trying to say is that there might be some test conditions where the listener could detect differences. If that is the case, then the listener simply can detect differences under those test conditions, and of course the two will therefore not sound the same to that listener under those test conditions. Rather obvious, is it not? Obvious, yes, but left out of your paragraph. Your paragraph would seem to indicate that if a listener hears no differences under a specific set of conditions, then the listener will hear no differences under any conditions. This, of course, is not a given. I would then suggest that you are not reading carefully and are drawing erroneous inferences. Untrue. Your statement suggests that if a listener doesn't hear a difference in a test, that for that listener, there is no difference. That thought doesn't at all take into account, for example, test validity or testing conditions. And, of course, another obvious point is that if you really want to find out whether two pieces of equipment, or two presentations, sound the same, then you should try to detect differences using the most sensitive conditions for detection: quick switching, double-blind, tests using the most revealing source material. Perhaps. I'm researching that at present, along with research concerning how we perceive music, particularly from a point of view of a trained musician. |
#465
|
|||
|
|||
"Ban" wrote in message ...
Jenn wrote: Thanks for your kind words, Jenn (listening to a beautiful LP of the Netherlands Wind Ensemble doing the Stravinsky Octet. Ahhhh....) The Netherlands Wind Ensemble is perhaps the world's best. I'll never forget hearing their recording of Mozart's Serenade in c minor, K.388. The slow movement was unbelievably luscious. Norm Strong |
#466
|
|||
|
|||
|
#467
|
|||
|
|||
In article ,
wrote: "Ban" wrote in message ... Jenn wrote: Thanks for your kind words, Jenn (listening to a beautiful LP of the Netherlands Wind Ensemble doing the Stravinsky Octet. Ahhhh....) The Netherlands Wind Ensemble is perhaps the world's best. I'll never forget hearing their recording of Mozart's Serenade in c minor, K.388. The slow movement was unbelievably luscious. Norm Strong Agreed on all counts. I heard them live in '88 at a conference. Just wonderful. Their present recordings (only on CD) are concentrating on very contemporary works. Also great. The sound on those old Philips LPs is among the most natural sound recorded, IMO.... Mozart (virtually all of the wind music), Richard Strauss (the best recording of the beautiful Op. 7), Gounod, etc. Fantastic in all ways. |
#468
|
|||
|
|||
"Ban" wrote in message ...
Jenn wrote: Indeed true, in my opinion. Listening to music, for most people, is largely a right-brain based experience. For good discussions of this, see Gardner: Frames of Mind and on a less technical level, Kerman: Listen. At the end of the day, I think that the left brain vs. right brain ways of dealing with the world might well be the basis for the heated arguments on this topic. For example, a person who experiences music in a highly right-brain oriented way, would, I believe, have a great deal of difficulty during the experience of a quick switch test. I hope to experience such a test soon which will help inform this thesis. Yes Jenn, that is the only way. Maybe you have noticed that much of this mental noise has been created by people who have never experienced a properly conducted DBT, ABX or whatever variant. Even the terminology is interpreted in different(erratic) ways. Your assumption, at least in my case, is wrong, Ban. I have done such tests. And what part of quick-switch, short-snippet, comparative testing (which is what I object to) is hard to interpret? It is the combination of the three that is the problem, not quick-switching per se. In rank order I rate a) short-snippet, b) comparative, and c)quick-switching as factors that potentially can contribute to flawed or invalid results. Thanks for your kind words, Jenn (listening to a beautiful LP of the Netherlands Wind Ensemble doing the Stravinsky Octet. Ahhhh....) This is how I'm listening to a test. I do not concentrate on certain instruments, but try to hear into the space in front and around me. And at the same time understand what the composer wanted to express: his joy or sadness, his love, desperation, celebration, meditation... Somehow also the musician feels the same and adds to the initial expression of the composer his interpretation. And then when I feel like switching, the performance goes on without interruption. The piece continues on the same spot. A little "click" doesn't dusturb me, but otherwise it is important that both streams are in sync and there is no perceivable silence. This is a quick switch test. And what is short is not the music snippet, but the switching action. Compare that with changing the speaker cables or whatever in a long break etc. Even a *one* second break is already disturbing and brings me back into the brain... And you want to know how it sounds when *not* in the brain. This is how from my experience I'm able to detect very miniscule differences between two feeds. I also listen this way when critically evaluating, Ban. But this is *not* blind, abx style or ab style testing using short segments of sound repeated a dozen and a half times...which is how Tom Nousaine, Arny, Stewart, and other vociferous proponents of the test do it. In fact in any test that requires dozens of "trials" this approach is simply too cumbersome because each trial takes too long and ear/brain fatigue sets in. And stretching it over a dozen or so days is impractical. Also note, as you describe it, you are listening to a "performance", where you can place sounds in context. ABX proponents disdain this, thinking that castanets, crickets, or tympani strokes taken out of any other musical context are enough. A lot of the confusion in trying to use these tests to do open-ended evaluation of audio components, in my opinion, comes from the fact that an adequate musical context is not established first as a basis for the evaluation. For example, if evaluating tympani, the brain naturally wants to take into account the acoustic environment it is in, and that in turn will be generated from the sounds of the full orchestra. A tympani can sound quite different depending on the acoustics of the auditorium and the distance in the mincing. If listening to the full piece and concentrating on the piece itself, one might come to the realization "I think something doesn't sound right with the tympani's" using this amp...I haven't heard it sound that way before on this recording, and it sounds less natural than it does in the concert hall". Simply hearing a tympani out of context is not likely to generate the same perception. Moreover, longer term listening has an emotional component that takes time to register (see Oohashi, et. al.). Quick switching even in a musical context will dramatize some differences, but emotional response longer term can also be affected by the gear doing the reproduction and this will take a few minutes to sink in and may even be a much longer term factor.. It is unlikely to be a quickly conscious factor...it is probably better expressed via an unconscious "feeling" preference (right brain) than a conscious choice (left brain). This is the one area that I think quick-switching fails...the longer term effects simply don't always manifest themselves right away. |
#469
|
|||
|
|||
"Mark DeBellis" wrote in message
... Stewart Pinkerton wrote: No, you can't, and you certainly haven't. If you could, why haven't you collected the $5,000 prize? Say, what is this $5,000 test, anyway? :-) Since there is a fairly high probability that, after enough trials, *someone* will appear to distinguish the cables even though the outcome is just chance -- -- in the same way that, if you flip a coin enough times, eventually you're likely to get 10 heads in a row -- -- why don't we get together, all enter, and agree to split the money? Has anybody figured the odds, anyway? What does it cost to enter? :-) Mark It doesn't cost anything to enter, but then it also doesn't cost anything to "contribute". A problem. There is absolutely no assurance that if somebody did win there would actually be money contributed. And that's what separates this little charade from reality. When they have all sent legally binding promissory notes to an escrow holder, then we will know they are serious.. |
#470
|
|||
|
|||
Mark DeBellis wrote:
wrote: 2) We could subject you to a blind test, to determine whether you can indeed distinguish between these cables when you do not know which is which. This could eliminate imagined sonic difference as an explanation (although, admittedly, it cannot confirm it). Not to quibble, and I don't disagree with what you are basically saying in this post, but aren't you entitled to claim more here than you do? If he fails to distinguish the cables, then doesn't that confirm that the difference he heard was imaginary? No, it does not. That's the fallacy of the excluded middle. 'Expectation bias' could indeed be operative, while I still am able to distinguish the cables because of something in the cables. (On the other hand, if he does distinguish the cables, it doesn't mean that there wasn't *any* imagined sonic difference; it just means the apparent difference wasn't *all* imaginary.) However, it's late, and I could very well not be reading you correctly. Mark |
#471
|
|||
|
|||
wrote:
wrote: What DATA are you talking about? This is not and cannot be scientific experiment. I am talking about my experiences and the reports of others with similar experiences. Excuse me? You made the very testable claim that you can hear the difference between cables. Fine, let's test it. No, it's not testable. That's the point. According to current scientific theory (and as a student of the philosophy of science, I presume you will not come back with the old "but that's just a theory" line), there are basically three possible reasons (assuming no mechanical failures, like bad connectors) why you might have such an experience: 1) The electrical characteristics of those cables are such that one attenuates the signal substantially more than the other, resulting in an audible decrease in the volume emerging from the speakers. 2) The electrical characteristics of those cables are such that one attenuates certain frequency ranges of the signal substantially more than the other, resulting in an audible difference in the frequency response emerging from the speakers. 3) You imagined that you heard a difference, based not on the sound produced but on other things you knew or believed about those cables. That's it. So far as I know, physics has discovered no other possible explanations for audible differences between cables beyond #1 and 2 above. As for #3, the propensity of humans to hear differences where none exist is well-established. I'm not sure what electrical differences are possible. The differences I heard in the cables' sound are hard to describe, but I'll try. High frequency transients (e.g., brushed cymbals) seemed more realistic and detailed, with more distinctness between events. Depth of image was somewhat greater. Voices were more palpable. Bass lines were stronger. The 'image' was overall more vivid, more realistic. Not only that, but these traits were consistent between auditions and recurred over several days. (Unsupportable stuff snipped) |
#472
|
|||
|
|||
wrote:
wrote: wrote: What DATA are you talking about? This is not and cannot be scientific experiment. I am talking about my experiences and the reports of others with similar experiences. Excuse me? You made the very testable claim that you can hear the difference between cables. Fine, let's test it. No, it's not testable. That's the point. More denial of reality. Do you really think that whole fields of science do not exist simply because you don't know about them? According to current scientific theory (and as a student of the philosophy of science, I presume you will not come back with the old "but that's just a theory" line), there are basically three possible reasons (assuming no mechanical failures, like bad connectors) why you might have such an experience: 1) The electrical characteristics of those cables are such that one attenuates the signal substantially more than the other, resulting in an audible decrease in the volume emerging from the speakers. 2) The electrical characteristics of those cables are such that one attenuates certain frequency ranges of the signal substantially more than the other, resulting in an audible difference in the frequency response emerging from the speakers. 3) You imagined that you heard a difference, based not on the sound produced but on other things you knew or believed about those cables. That's it. So far as I know, physics has discovered no other possible explanations for audible differences between cables beyond #1 and 2 above. As for #3, the propensity of humans to hear differences where none exist is well-established. I'm not sure what electrical differences are possible. Well, knock me over with a feather. The differences I heard in the cables' sound are hard to describe, but I'll try. High frequency transients (e.g., brushed cymbals) seemed more realistic and detailed, with more distinctness between events. Depth of image was somewhat greater. Voices were more palpable. Bass lines were stronger. The 'image' was overall more vivid, more realistic. Not only that, but these traits were consistent between auditions and recurred over several days. All of which, if a response to a real physical condition, must be the result of the effect that said cables have on the electrical signal passing through them. And since some folks DO know what electrical differences are possible, we have the finite set of possibilities listed above. We also have ways of testing and eliminating some of those possibilities. (Unsupportable stuff snipped) More denial of reality. bob |
#473
|
|||
|
|||
Jenn wrote:
In article , Chung wrote: Jenn wrote: In article , Chung wrote: Jenn wrote: In article , Chung wrote: What matters, or might well matter to some, is whether the information derived from one source, or its perceptual effect, is the same as that from another source (in blind musical listening). Why should we assume that this reduces to a matter of detecting differences? Because if a difference is not detected by the listener, then the two sources *must* sound the same to the listener. It's really that simple. Actually to be more fully correct, if a difference is not detected by the listener under the given test conditions, then they must sound the same under those conditions. Actually, the qualification you added does not make my statement any "more truly correct". What you were trying to say is that there might be some test conditions where the listener could detect differences. If that is the case, then the listener simply can detect differences under those test conditions, and of course the two will therefore not sound the same to that listener under those test conditions. Rather obvious, is it not? Obvious, yes, but left out of your paragraph. Your paragraph would seem to indicate that if a listener hears no differences under a specific set of conditions, then the listener will hear no differences under any conditions. This, of course, is not a given. I would then suggest that you are not reading carefully and are drawing erroneous inferences. Untrue. Your statement suggests that if a listener doesn't hear a difference in a test, that for that listener, there is no difference. That thought doesn't at all take into account, for example, test validity or testing conditions. Well, Jenn, try to read carefully now. Here's what I said: "Because if a difference is not detected by the listener, then the two sources *must* sound the same to the listener. It's really that simple." That means that if a listener does not detect any difference, in (a) a test, or (b) in any test, or (c) in all tests, then the sources must sound the same in (a) that test, or (b) in any test, or (c) in all tests, *respectively*. I did not say that if a listener does not hear any difference in one test, then the two sources must sound the same under different test conditions. That's what you believed I suggested, and that was an erroneous inference from what I wrote. Note that I carefully did not even say that there were no differences between the two sources. What was important was that the two sources sound the same if the listener could not detect any difference. Almost by definition. |
#474
|
|||
|
|||
Harry Lavo wrote:
"Mark DeBellis" wrote in message ... Stewart Pinkerton wrote: No, you can't, and you certainly haven't. If you could, why haven't you collected the $5,000 prize? Say, what is this $5,000 test, anyway? :-) Since there is a fairly high probability that, after enough trials, *someone* will appear to distinguish the cables even though the outcome is just chance -- -- in the same way that, if you flip a coin enough times, eventually you're likely to get 10 heads in a row -- -- why don't we get together, all enter, and agree to split the money? Has anybody figured the odds, anyway? What does it cost to enter? :-) Mark It doesn't cost anything to enter, but then it also doesn't cost anything to "contribute". A problem. There is absolutely no assurance that if somebody did win there would actually be money contributed. And that's what separates this little charade from reality. When they have all sent legally binding promissory notes to an escrow holder, then we will know they are serious.. I am one of the contributors. I believe I said that I would contribute $200. If "BEAR" were to enter to try to tell "silver wires" from normal copper wires, I will contribute $500. If someone wants to set up an escrow, I will gladly write the check. Of course, Harry might have been right in an unintentional sort of way. We contribute because we know that it probably will not cost us anything, because (a) BEAR or anyone else (such as Harry) will likely not take the test, and (b) if someone takes the test, he is most likely to fail the test. |
#475
|
|||
|
|||
Mark DeBellis wrote:
Buster Mudd wrote: Mark DeBellis wrote: Buster Mudd wrote: I think you keep harping on this notion that our perceptions of these two very different phenomena should somehow be similar. The phenomena (musical memory & audio memory) are similar only in that they are initially triggered by an auditory stimulus; beyond that they are so dissimilar, both in their own content & in the methods by which we perceive them, that to refer to them both as "information we pick up in this way" is to incorrectly characterize them. We *don't* pick up both types of information in the same way once the auditory stimulus has gotten past the ear. I have no idea if our perceptions of them should be similar or different, but given that we have tests for one kind of information, how are we entitled to rely on those tests to tell us that there are no differences in the other kind of information? We're not; but has anyone claimed otherwise? I don't recall anyone in rec.audio.high-end ever asserting that tests...be they ABX, DBT, quick switch, slow switch, monadic, any kind of tests at all...could or couldn't determine differences in our perceptions of *musical content*. Every reference to using statistical testing to identify perceived or imagined differences I've come across here has been pertaining to sonic attributes of audio components. It was the example of SACD vs. CD that I started with, so let me return to that. I thought the idea was that certain tests could tell us that SACD and CD are perceptually equivalent (if they are). For clarity, I trust you won't mind if I rephrase that supposition as "certain tests could tell us whether or not SACD and CD are perceptually equivalent"? Because that is what you're saying, right? If the tests don't tell us that there are no differences in the kind of information that has to do with the longer-term kind of memory, how do they show perceptual equivalence? And by "the kind of information that has to do with the longer-term kind of memory" are you (once again) talking about Musical Content? Because if so, that's *NOT* the "kind of information" that would allow one to discern a perceptual equivalence between an SACD player and a CD player...or between an SACD recording and a CD recording, for that matter. |
#476
|
|||
|
|||
"Chung" wrote in message
... wrote: Chung wrote: And, of course, another obvious point is that if you really want to find out whether two pieces of equipment, or two presentations, sound the same, then you should try to detect differences using the most sensitive conditions for detection: quick switching, double-blind, tests using the most revealing source material. This is false. It takes me a while to hear all the differences bewteen products. Different program material presents different aspects for evaluation. Quick switching is worthless, in my experience. You simply do not understand what quick switching means. Quick switching does not mean that you have a short time to listen to each product. You can listen to those for minutes, hours, or days if you want to. Quick switching means that there is no long delay between the two presentations such that you still have a very accurate memory of what the other product sounded like. Now once you have identified differences, it will be helpful to use short material that brings out those differences. For example, there may be a certain phrase in a vocal recording, or a particular passage in a precussion recording, that reveals such differences clearly. It will be good then to use short snippets, conisisting of those phrases or passsages, to help you identify differences in your (hopefully) blind test. This way you give yourself the best chance to detect differences. IMO both are useless. As the LP believers often report, a day of listening to CD puts them on edge, causing headache and fatigue. Perhaps a day of listening to SACD is less problematic than is CD. I feel that members of the LP camp prefer that format because the cross-talk "softens" the edges in each channel. One hears a considerable amount of left channel information coming from the right speaker and v.v. |
#477
|
|||
|
|||
|
#478
|
|||
|
|||
wrote:
Mark DeBellis wrote: wrote: 2) We could subject you to a blind test, to determine whether you can indeed distinguish between these cables when you do not know which is which. This could eliminate imagined sonic difference as an explanation (although, admittedly, it cannot confirm it). Not to quibble, and I don't disagree with what you are basically saying in this post, but aren't you entitled to claim more here than you do? If he fails to distinguish the cables, then doesn't that confirm that the difference he heard was imaginary? No, it does not. That's the fallacy of the excluded middle. 'Expectation bias' could indeed be operative, while I still am able to distinguish the cables because of something in the cables. That's what I go on to say below, but I'm not sure how it relates to the above, which is a different point. There I'm saying that if someone *can't* distinguish the cables blind, it means the difference they seemed to hear was imaginary. (On the other hand, if he does distinguish the cables, it doesn't mean that there wasn't *any* imagined sonic difference; it just means the apparent difference wasn't *all* imaginary.) Mark |
#479
|
|||
|
|||
Ban wrote:
Jenn wrote: Indeed true, in my opinion. Listening to music, for most people, is largely a right-brain based experience. For good discussions of this, see Gardner: Frames of Mind and on a less technical level, Kerman: Listen. At the end of the day, I think that the left brain vs. right brain ways of dealing with the world might well be the basis for the heated arguments on this topic. For example, a person who experiences music in a highly right-brain oriented way, would, I believe, have a great deal of difficulty during the experience of a quick switch test. I hope to experience such a test soon which will help inform this thesis. Yes Jenn, that is the only way. Maybe you have noticed that much of this mental noise Cheap shot. If I understand the point you go on to make, quick-switch tests do not always involve the comparison of short snippets. Sometimes they involve, rather: listening to a stretch of A in order to establish a context, and then listening to a short portion of B, and evaluating how well B fits in that context. Yes? Thanks, that's a helpful clarification. But I think what I've said applies to this sort of test. has been created by people who have never experienced a properly conducted DBT, ABX or whatever variant. Even the terminology is interpreted in different(erratic) ways. Thanks for your kind words, Jenn (listening to a beautiful LP of the Netherlands Wind Ensemble doing the Stravinsky Octet. Ahhhh....) This is how I'm listening to a test. I do not concentrate on certain instruments, but try to hear into the space in front and around me. And at the same time understand what the composer wanted to express: his joy or sadness, his love, desperation, celebration, meditation... Somehow also the musician feels the same and adds to the initial expression of the composer his interpretation. If we are asking what such tests establish and why, then it is important to understand what the tests involve. But I think basically I do: it's discrimination. I'm not sure why you think it's essential to actually *experience* such a test from the point of view of a subject, in order to understand why it shows what it shows (any more than it is necessary to build a switchbox). You're not saying that when a person experiences the test, he or she somehow *sees* that the test is valid for the relevant purposes? That would be like saying, a person won't understand why sighted comparison is valid unless they experience the sound of the cables for themselves. It's not as if the experience of the test establishes its own validity. If in order to do a quick-switch test properly, one would have to understand and feel at one with the composer's joy, then the whole idea of making this a repeatable, objective matter would be in big trouble. And then when I feel like switching, the performance goes on without interruption. The piece continues on the same spot. A little "click" doesn't dusturb me, but otherwise it is important that both streams are in sync and there is no perceivable silence. This is a quick switch test. And what is short is not the music snippet, but the switching action. Compare that with changing the speaker cables or whatever in a long break etc. Even a *one* second break is already disturbing and brings me back into the brain... And you want to know how it sounds when *not* in the brain. This is how from my experience I'm able to detect very miniscule differences between two feeds. I doubt that your experience tells you, by itself, that if you don't detect a difference in that experience then there is no difference over long-term listening. That's a question of empirical psychology that's not revealed in any immediate experience. Mark |
#480
|
|||
|
|||
wrote:
Mark DeBellis wrote: wrote: 2) We could subject you to a blind test, to determine whether you can indeed distinguish between these cables when you do not know which is which. This could eliminate imagined sonic difference as an explanation (although, admittedly, it cannot confirm it). Not to quibble, and I don't disagree with what you are basically saying in this post, but aren't you entitled to claim more here than you do? If he fails to distinguish the cables, then doesn't that confirm that the difference he heard was imaginary? Alas, statistics doesn't let you draw that conclusion. We set up a null hyothesis--"You cannot hear a difference between these cables"--and set a significance threshold. If he tops the threshold, you've disproved the null hypothesis. But if he doesn't top the threshold, all you've done is failed to disprove the null hypothesis. This has nothing to do with audio, BTW; it's a law of statistics, and affects any statistical test. Well, hang on a sec though. Suppose it seems to someone that they hear a difference between two cables. If the difference is not (entirely) imaginary, then the person is detecting a difference. And if he is detecting a difference, then we can predict from this that he is very likely, in a blind test, to distinguish the cables significantly better than chance. So if he does *not* so distinguish the cables in such a test, that is evidence that the apparent difference was imaginary. I don't see what's wrong with that reasoning, or why statistics blocks it. Mark |
Reply |
Thread Tools | |
Display Modes | |
|
|
Similar Threads | ||||
Thread | Forum | |||
common mode rejection vs. crosstalk | Pro Audio | |||
Topic Police | Pro Audio |