Home |
Search |
Today's Posts |
#41
|
|||
|
|||
What is so high end about high end?
"Harry Lavo" wrote in message
"Arny Krueger" wrote in message news:%P33b.274985$Ho3.35716@sccrnsc03... "Dennis Moore" wrote in message news0X2b.273368$uu5.62103@sccrnsc04 "Wylie Williams" wrote in message But if there is a simple method I would like to see instructions. I am considering comparing three amps I own. Normally I would have just substituted and listened to each a few days, comparing opinions with my wife as we went along, but I will consider bias controlled listening. Wylie Williams Which method of testing makes more sense Wylie? Doing DBT's or SBT's which you may never use any of these amps for again. Or hooking them up for a few days. Seeing how you feel about them at the end of it. And picking the one you feel good about to keep listening to long term. ..and then never using any of these amps again but the one selected. Of course, there is nothing that keeps one from "hooking them up for a few days" under DBT conditions. Which method of selection comes closest to matching the long term use of the product? Since there need not be any difference in the timetable or listening circumstances for the evaluations, they are equally close to matching the long term use of the product. However, I think it needs to be said that long term listening has been thoroughly investigated under DBT conditions, and found to be very insensitive when it comes to reliably perceiving small audible differences. How about in perceiving less direct but more important "gestalt" difference? EG, depth of soundstage, focus of instruments, accuracy of timbre over a range from ppp to fff, etc. Long term listening is generally said to involve listening for days or weeks. In contrast, the entire sequence of 9 Beethoven Symphonies can be listened to in something like 6 hours ( I just did it for the Toscanini and Walter versions). Suffice it to say, just about any "gestalt" difference can be revealed in listening tests that are not "long term" as the phrase is commonly used. These are the kinds of things audiophiles can hear in a relaxed, monadic state that they have trouble believing are imaginary when they are put into the typical blind a-b or especially a-b-x test which uses much more of a left -brain approach and based on your own comments here seem mostly sensitive to frequency response and loudness. This would appear to be a collection of assertions that lack proximal factual support. Again, there's nothing about DBT listening that prohibits any kind of long-term listening strategy that one might come up with. Many of these things have been tried and found to provide no joy. OTOH. presentation of the sonic alternatives in quick succession is a well-known and generally-recognized means to optimize listener sensitivity. |
#42
|
|||
|
|||
What is so high end about high end?
I believe Dennis is talking about "long term musical satisfaction' which
is different from pride of ownership. And you must admit, while possible to do so, it is unusual for DBT's to be done with switching done on a scale of days. snip So should we be so surprised that a method developed and found most telling under short term comparisons gives different than expected results when used for long term comparison. DBT's are for detecting subtle audio differences. Long term satisfaction is often not dependent on subtle differences. While I don't want to start another DBT debate here, I strongly disagree with you. You may consider DBTs useful for detecting 'subtle audible differences' but that is exactly what they DON'T reveal. DBTs have been shown to reveal gross differences in frequency response and loudness, but little else - especially with music as a source and with averaged responses of untrained listeners (normal audiophiles). IMHO long term listening satisfaction is based on a variety of subtle audible differences, such as dynamic shadings, soundstage reproduction, imaging, tonal shadings and timbral accuracy. Regards, Mike Yes, Mike got it. I was indeed speaking of long term musical satisfaction. You of course don't know me, but snobbish or even genuine pride of ownership isn't why I own something. If I think it does the music better, and I can afford it then I use it. I have made some useful, but terrible looking things that stayed in my system along time. I have had some beautiful equipment, but the beauty to me was the way it allowed me to enjoy music. I have at times had barely practical arrangements of equipment which allowed me musical enjoyment. Dennis |
#44
|
|||
|
|||
What is so high end about high end?
|
#45
|
|||
|
|||
What is so high end about high end?
(ludovic mirabel) wrote in message news:xDA3b.220180$cF.73109@rwcrnsc53...
(Bob Marcus) wrote in message t.net... (ludovic mirabel) wrote in message news:nsd3b.276882$o%2.127236@sccrnsc02... See initial discussion below: Mr. Marcus responds: If I were a con artist, and wanted to "prove" how acute my hearing was, this is exactly the method I would use. It's really quite clever for that purpose. Because the two sides aren't level-matched, it would be easy to tell when the two sides had been switched. (The image appears in a different place!) All one has to do is declare a "preference" on the first trial. After that, since you'll always be able to tell when the switch has been made, it's hardly a challenge to declare the same "preference" repeatedly. No wonder you've impressed your friends with this trick. Now, why don't you try it level-matched, and double-blind, and then submit an article to JAES (or even S&V!), so we can see what the results really are? bob Mr. Marcus you give me too much credit. Con artist I may be Actually, I don't think you're conning anyone. but I'm not a magician- the two SIDES are NOT switched- only the components are. The room stays where it was. Any differences between its two SIDES stay where they were, the speakers, the cd.player or the turntable stay where they were. The only things switched are the two components-two links between the sound source and the speakers. They are moved from one SIDE to the other SIDE And these components should be as level- comparable as the components were in all the published ABX component comparison tests- all of them with negative outcomes. Ah, so you do level-match them. I thought you said that wasn't necessary. The levels must be close to begin with because TO BEGIN WITH YOU ADJUST THE LEFT AND RIGHT VOLUMES ON YOUR PREAMP TILL THE MONO IMAGE IS DEAD CENTRE. If you had much experience with voltmeter you'd know that very little decibel difference namely 0,5 db. moves the image right or left. Nothing against using a voltmeter- I got no revelations when I did. But you used one, right? Because you absolutely have to, you know. Or rather, your lovely assistant has to, because if YOU adjusted levels after SHE switched the components, it wouldn't even be single-blind, now would it? One needs a particular kind of imagination and /or limited experience to believe that such kindergarten precautions don't occur to others. If all you're using are kindergarten precautions, I'm afraid we have a bias control problem. The monophonic image stays solid with component changes. At least it does in my room- which does not rotate short of three glasses of Cognac. How about yours? Does the earth move when you ABX at home? Do it double-blind with an impartial witness, and show us the data. Mercy- I nearly forgot. Greenhill in his ABX panel cable testing had TWO out of SIX positive results but ONLY when he was comparing a thick and a thin cable with 1,75 db. level difference between them- and ONLY when using pink noise for his signal- not music. These two results are immortalised as "Marcus-Ovchain positive 1,75db. test" because you claimed it as your one and only evidence that ABX can give positive results when comparing components and Ovchain was applauding on the sidelines. That explains why you think that 1,75 db at say 4000hz would be missed by anyone not terminally deaf. For your information- these old ears detect a 0,5 db change in volume when switching my stepped volume controls. I am quite through trying to argue with you about Greenhill's tests. Everything you have written about them has been wrong (save, perhaps, the date of publication), and although you have been corrected many times by many people here, you persist in your misreporting. I can only assume that is willful. I once again invite anyone who actually cares to look up the original article and find out what Greenhill really said. bob |
#46
|
|||
|
|||
What is so high end about high end?
Bob Marcus wrote:
(ludovic mirabel) wrote in message news:%wV4b.329114$o%2.151942@sccrnsc02... Mr. Marcus said: If I were a con artist, and wanted to "prove" how acute my hearing was, this is exactly the method I would use. It's really quite clever for that purpose. Mr. Williams asked for a method less complicated than ABX. Very hesitantly, knowing the ways of some in the RAHE I summarised a method that was accepted and published by the "Audio Amateur". Only to be told by an expert called Marcus that it was a "con artist method, devised to just "prove" my acute hearing. I said it COULD be used that way. I also said I didn't think you were conning anyone. snip A simple task in my case : I have volume controls stepped in 0,5 db. steps. And you think 0.5 dB is close enough??? snip some more Eg. The bass is distorted on one side of your room but you still have a statistically significant positive results: Wait--do you have any significant positive results? You haven't posted any. snip yet again There are more things wrong with your left-right method than I can count. Basically, you're asking your brain to sort out sounds in a way it isn't wired to. So, in theory at least, it should not work. I will be perfectly willing to concede that it does work, however, if you will do the following: 1) Level-match to within 0.1 dB (the usual standard, I believe). 2) Perform the tests double-blind, with an independent witness. 3) Report the results in full. Until you do that, all you're offering is a theoretical approach that's contradicted by all existing theory. And given that the existing theory has been developed over decades by people far more expert in the field than you or me, that doesn't give you much ground to stand on. bob Actually I don't think anyone is interested in doing a test to find differences/preferences this way. Example: an important parameter of an audio component is its ability to convey image. This test uses mono signals and only exercises one channel of the device under test. I thought one of the biggest complaints of the original poster was that the DBT test is different than how people listen to their stereos. This L/R test is so different than the way we listen to audio equipment that I am totally surprised that he would even consider it. Not to mention the fact that any dynamic interchannel effects (like crosstalk, power supply effects, imbalance of frequency response in a RIAA circuit, etc.) will be totally undetected. The more I think about the test, the more I am amazed at what it does not do. |
#47
|
|||
|
|||
What is so high end about high end?
"Bob Marcus" wrote in message
A simple task in my case : I have volume controls stepped in 0,5 db. steps. And you think 0.5 dB is close enough??? snip some more Well if there is a .5dB difference he could go up or down to match exactly. With .5 dB steps, the maximum difference between two sources will be .25 dB. And often would be less. At one time the standard was within .2dB which eventually was reduced to .1 dB. At least some of the time you could be testing components that were this close. Would be a good move to measure it and see that it is. But the .5 dB steps don't automatically make the level differences too large. Dennis |
#48
|
|||
|
|||
What is so high end about high end?
(Bob Marcus) wrote in message news:xJo5b.342306$YN5.233613@sccrnsc01...
(ludovic mirabel) wrote in message news:%wV4b.329114$o%2.151942@sccrnsc02... Mr. Marcus said: Snip: There are more things wrong with your left-right method than I can count. Basically, you're asking your brain to sort out sounds in a way it isn't wired to. So, in theory at least, it should not work. Exactly which theory about "how the brain is wired" is against turning your head from left to right to compare sounds. I missed it in the neurophysiology reports I read. A few references and quotes, please. Just to enlighten me. I will be perfectly willing to concede that it does work, however, if you will do the following: 1) Level-match to within 0.1 dB (the usual standard, I believe). 2) Perform the tests double-blind, with an independent witness. 3) Report the results in full. Your " willing to concede" is a real incentive. It will open the gates, no doubt, to respectability amongst electronics' engs, psychometricians, my professional cousins neurologists and neuro physiologists, all the folk who are awaiting your word. Until you do that, all you're offering is a theoretical approach that's contradicted by all existing theory. And given that the existing theory has been developed over decades by people far more expert in the field than you or me, that doesn't give you much ground to stand on. bob Help "us' out with a few points: 1) I said several times that I'm proposing an ad hoc, imperfect, nonscientific, non researched way to avoid bias It only just an elaboration of covering the brand names. It is meant to be used by whoever feels that he'd like to attempt some degree of bias control, very simply, at his home, without using $600:00 switch, and without a DBT method which requires training and, at that, is still controversial (ah, theory!) after 30 years of use in this application. It is not even for everyone. Some, I'm sure will not like it- and that's the end of the matter- as lonng as they have facts not theories to report. I said also that speculative theoretical objections from those without hands-on experience are of little interest to me. But since you insist on taking me there it would help if you quoted a few names and references to YOUR research experts in "existing theory" WITH DIRECT RELEVANCE TO SIMULTANEOUS COMPONENT COMPARISONS WITH RANDOM CHANGES FROM SIDE TO SIDE. So far we have your interpretative report only of what those anonymous experts are supposed to have said but no doubt you have "many" references at your fingertips. 2) Your strict conditions for "conceding" would make sense if, blazing the path to a better alternative, YOU had done all these things- to avoid bias- in YOUR component choices. You must have tested them before purchase with 0,1 db level matching, double blind precautions and independent witnesses- don't forget witnesses. Do you own the ABX switch? It would be interesting to hear what was the otcome of those trials and what system did you collect using these rigid research methods. 3) I missed any further discussion of your theory of the moving centre image as one moves components from side to side. Remember? Surely it wouldn't be that you'd like to divert attention from that "misreporting" (or was it "misunderstanding"?) to a "theoretical" argument about irrelevancies? (see footnote). I wish I could concentrate your mind on this crucial topic: Does the central image move as you move the components from side to side or does it stay steady? 0,1 db level difference or 0,2 or or 0.5 or 1,75- whatever . Ludovic Mirabel Footnote Marcus explained: Because the two sides aren't level-matched, it would be easy to tell when the two sides had been switched. (The image appears in a different place!) No it does not, It stays rock-solid throughout (see below) All one has to do is declare a "preference" on the first trial. After that, since you'll always be able to tell when the switch has been made, .. Yes, you will or not but not because "the image appears in a different place". It does not. The component are Xmoved from side to side but the levels between the two sides do not change. Elementary: you do not use components with different output levels in ABX , you do not use them in left-right. Unless you're deliberately testing for the threshhold of level- perception and use eg. two wires of different diameter If you do, you get the Marcus-Ovchain 1,75db "positive" ABX test. Automatically. As explained to you, twice, very exact level matching is completely unnecessary in the left-right simultaneous comparison. If the better-liked component is moved to the less favoured side of the room and yet is still identified , it REINFORCES the validity of the result. Surely it is not THAT difficult to grasp. But of course there is nothing against exact level matching should you so wish. If you HAVEN'T graspED it YET I despair of explaining it to you. Knowing your scrupulous probity in an argument I'd never suspect that your 0,1db, reliable witnesses and the rest is just a lot of squid-ink. L.M. |
#49
|
|||
|
|||
What is so high end about high end?
chung wrote in message ...
Bob Marcus wrote: (ludovic mirabel) wrote in message news:%wV4b.329114$o%2.151942@sccrnsc02... Snip previous discussion. Chung said: Actually I don't think anyone is interested in doing a test to find differences/preferences this way. Example: an important parameter of an audio component is its ability to convey image. This test uses mono signals and only exercises one channel of the device under test. I thought one of the biggest complaints of the original poster was that the DBT test is different than how people listen to their stereos. This L/R test is so different than the way we listen to audio equipment that I am totally surprised that he would even consider it. Not to mention the fact that any dynamic interchannel effects (like crosstalk, power supply effects, imbalance of frequency response in a RIAA circuit, etc.) will be totally undetected. The more I think about the test, the more I am amazed at what it does not do. I am completely flabbergasted. When you say: "This test uses mono signals and only exercises one channel of the device under test" do you mean that you concluded that I'm using ONE channel at a time? I'm upset because it proves that I'm unable to communicate very obvious and simple ideas to some readers. And it saddens me that you'd attribute such brain-lame ideas to me. Of course BOTH channels are playing at the same time. The channel levels are matched before the first change and never touched after that. Your assistant Xchanges the two components (which of course are electrically comparable) at random from one side to the other. The subject either does or does not consistently recognise the locations . Perhaps you missed the long word in capitals SIMULTANEOUSLY- I'm reposting it for your attention . I'm reposting also the explanation why slight level difference REINFORCES the validity of a correct answerif anything. If that too is not clear please don't hesitate to ask. You reproach me for repetition but it would appear that I have not yet repeated myself enough. Reprint: (regretfully and with apologies) 1) Get a monophonic or near monophonic (eg. centred soprano) signal source. MUSICAL, not an artefact. 2) On the left insert one component, on the right the OTHER ONE- (in the case of interconnects using two of one kind together i.e.source to preamp and preamp to amp on each side will give better contrast.) 3) Listen -write down your preference, get blinded. 4) An assistant now changes AT RANDOM (coin throw) both components from one side to the other or (of course) leaves them where they are keeping the records. 5) This is repeated minimum 15 times- for any length of time and with interval for lunch if you like. EVERY TIME you note your preference The repetition and change are the CRUX. At this point INVARIABLY someone says: No good, room sides differ, levels differ subtly etc. Answer;If there are differences between room sides, speaker volumes etc. and yet you still prefer and locate one of the two component as it moves from "good" to thje "bad"side surely, that REINFORCES the results- yes? no? Eg. The bass is distorted on one side of your room but you still have a statistically significant positive result: "I (...recognise) and prefer the sound of this preamp on EITHER side." The comparison is not just (supposedly) "instantaneous"- it is SIMULTANEOUS. The theoretical objections from people who never tried it are of little interest. Addendum: This is not a research test. No test, not DBT, exists FOR THIS APPLICATION that is validated by properly conducted research. The inferences from other fields (eg. medicine, psychometrics etc) are apple and oranges. Consider just this for one thing. The crux in medical research is proper selection of the subjects to get representative outcome numbers: age, gender, age, race, social backgrounds, dietary and other habits and so on. How do you select a representative panel for component comparison tests. Just untrained audiophiles? You'll get untrained "They all sound the same answer". Trained, middle-ahed chamber music lovers? Well you got a huge selection bias. The truth that our scientists will not face is that NO PROPERLY RESEARCHED TEST , WITH THE VARIETY OF THE AUDIOPHILE SPECIES COMPARING MUSICAL QUALITIES OF COMPONENTS, EXISTS AND THAT NONE IS POSSIBLE IN THE PRESENT STATE OF KNOWLEDGE. (Repetition again). Ludovic Mirabel |
#51
|
|||
|
|||
What is so high end about high end?
(ludovic mirabel) wrote in message news:38U5b.269475$Oz4.71805@rwcrnsc54...
(Bob Marcus) wrote in message news:xJo5b.342306$YN5.233613@sccrnsc01... (ludovic mirabel) wrote in message news:%wV4b.329114$o%2.151942@sccrnsc02... Mr. Marcus said: Snip: There are more things wrong with your left-right method than I can count. Basically, you're asking your brain to sort out sounds in a way it isn't wired to. So, in theory at least, it should not work. Exactly which theory about "how the brain is wired" is against turning your head from left to right to compare sounds. I missed it in the neurophysiology reports I read. A few references and quotes, please. Just to enlighten me. I meant only that the brain is wired to synthesize the sound it hears from the two channels. That's how stereo works, after all. Your method forces you to work against that. I will be perfectly willing to concede that it does work, however, if you will do the following: 1) Level-match to within 0.1 dB (the usual standard, I believe). 2) Perform the tests double-blind, with an independent witness. 3) Report the results in full. Your " willing to concede" is a real incentive. It will open the gates, no doubt, to respectability amongst electronics' engs, psychometricians, my professional cousins neurologists and neuro physiologists, all the folk who are awaiting your word. Until you do that, all you're offering is a theoretical approach that's contradicted by all existing theory. And given that the existing theory has been developed over decades by people far more expert in the field than you or me, that doesn't give you much ground to stand on. bob Help "us' out with a few points: 1) I said several times that I'm proposing an ad hoc, imperfect, nonscientific, non researched way to avoid bias It only just an elaboration of covering the brand names. It is meant to be used by whoever feels that he'd like to attempt some degree of bias control, very simply, at his home, without using $600:00 switch, and without a DBT method which requires training Actually, to be any good, your method would require at least as much training. After all, you're listening for the exact same artifacts. (And, as I noted above, you're listening for them under rather more challenging circumstances.) and, at that, is still controversial (ah, theory!) after 30 years of use in this application. Not among the people who use it regularly to design and test audio components, among other things. It's only "controversial" because some people are uncomfortable with its results and deal with the resulting cognitive dissonance by maintaining a willful ignorance about it. It is not even for everyone. Some, I'm sure will not like it- and that's the end of the matter- as lonng as they have facts not theories to report. But, as I keep pointing out, you have NO facts to report. I said also that speculative theoretical objections from those without hands-on experience are of little interest to me. That is evident. However, those to whom you are recommending this method might be interested to know just how well it has been validated. That is, not at all. But since you insist on taking me there it would help if you quoted a few names and references to YOUR research experts in "existing theory" WITH DIRECT RELEVANCE TO SIMULTANEOUS COMPONENT COMPARISONS WITH RANDOM CHANGES FROM SIDE TO SIDE. So far we have your interpretative report only of what those anonymous experts are supposed to have said but no doubt you have "many" references at your fingertips. 2) Your strict conditions for "conceding" would make sense if, blazing the path to a better alternative, YOU had done all these things- to avoid bias- in YOUR component choices. You must have tested them before purchase with 0,1 db level matching, double blind precautions and independent witnesses- don't forget witnesses. Do you own the ABX switch? It would be interesting to hear what was the otcome of those trials and what system did you collect using these rigid research methods. Ah, but I can point to published results done by people who did. You can't. 3) I missed any further discussion of your theory of the moving centre image as one moves components from side to side. Remember? Surely it wouldn't be that you'd like to divert attention from that "misreporting" (or was it "misunderstanding"?) to a "theoretical" argument about irrelevancies? (see footnote). I wish I could concentrate your mind on this crucial topic: Does the central image move as you move the components from side to side or does it stay steady? 0,1 db level difference or 0,2 or or 0.5 or 1,75- whatever . Well, let's say you have two amps that aren't precisely level-matched. When you switch them side-to-side, the louder amp is now on the other side. Doesn't that shift the image? And doesn't that provide the supposedly blinded test subject with a rather obvious clue as to whether the amps have been switched or not? And since that clue is based entirely on level, rather than on the distinctive sound of the two amps, doesn't that render the test pretty much meaningless? Ludovic Mirabel Footnote Marcus explained: Because the two sides aren't level-matched, it would be easy to tell when the two sides had been switched. (The image appears in a different place!) No it does not, It stays rock-solid throughout (see below) All one has to do is declare a "preference" on the first trial. After that, since you'll always be able to tell when the switch has been made, .. Yes, you will or not but not because "the image appears in a different place". It does not. The component are Xmoved from side to side but the levels between the two sides do not change. Elementary: you do not use components with different output levels in ABX , you do not use them in left-right. Unless you're deliberately testing for the threshhold of level- perception and use eg. two wires of different diameter If you do, you get the Marcus-Ovchain 1,75db "positive" ABX test. Automatically. As explained to you, twice, very exact level matching is completely unnecessary in the left-right simultaneous comparison. If the better-liked component is moved to the less favoured side of the room and yet is still identified , it REINFORCES the validity of the result. Not if the image shift tells the listener that the device favored in the first test is now on the other side. In that case, the two trials are not independent. Surely it is not THAT difficult to grasp. Tsk, tsk. Another condescending remark slip past the moderators. bob p.s.: Just for fun, check out a real expert's take on the problems with simultaneous left-right comparisons: http://groups.google.com/groups?hl=e...cs.cornell.edu |
#52
|
|||
|
|||
What is so high end about high end?
(Stewart Pinkerton) wrote in message ...
On Fri, 05 Sep 2003 05:28:24 GMT, (ludovic mirabel) wrote: I am completely flabbergasted. When you say: "This test uses mono signals and only exercises one channel of the device under test" do you mean that you concluded that I'm using ONE channel at a time? I'm upset because it proves that I'm unable to communicate very obvious and simple ideas to some readers. And it saddens me that you'd attribute such brain-lame ideas to me. I am completely flabbergasted that you'd be surprised by such an attribution..... :-) Of course BOTH channels are playing at the same time. That is exactly the problem. You have at best a 6dB signal to noise ratio between the reference channel and the channel containing the unit under test. This will pretty well guarantee the masking of any real but subtle differences. To claim, as you do, that this works *better* than a time-proximate 'ABX' test, is contrary to everything we know about human hearing - and that is a *lot* more than the 'subjectivists' care to believe. It is now the RAHE time to make lightning and thunder shouting with dramatised indignation: "Quote one sentence of mine claiming that it works better or you'll see my seconds this evening" But I won't. I'm not that anxious to win. No I never said that it works "better" than whatever. If I let slip something like that I recant. It works better for me and for some others. Quite likely not for everyone You see I don't believe that a universally valid "test" for comparing audio components exists. I gave my reasons too many times already but will again if you insist. Your theory why it shouldn't work is very interesting. 6db is good enough for me. I don't mind if you believe me or not, I'm not trademarking anything. I said before that my platoon seargent (in Scotland BTW) used to say "Practical life f.... theory" Try my nontest and then report back. 6db. and all. Ludovic Mirabel |
#53
|
|||
|
|||
What is so high end about high end?
(Bob Marcus) wrote in message ...
(ludovic mirabel) wrote in message news:38U5b.269475$Oz4.71805@rwcrnsc54... (Bob Marcus) wrote in message news:xJo5b.342306$YN5.233613@sccrnsc01... (ludovic mirabel) wrote in message news:%wV4b.329114$o%2.151942@sccrnsc02... Big snip. I said: Surely it is not THAT difficult to grasp. Tsk, tsk. Another condescending remark slip past the moderators. bob p.s.: Just for fun, check out a real expert's take on the problems with simultaneous left-right comparisons: http://groups.google.com/groups?hl=e...cs.cornell.edu I'll get to the snipped portion of this posting, with Mr. Marcus latest revision of his "brain- wiring" theory, sometime later- after a lunch and a snooze. But here for the 2nd. time in 24 hrs Mr. Marcus sees fit to start special pleading for censorship- first in regard to Mr. Lang's text in the "Audiophile Press" thread, and now in regard to mine. And once againhe does it in the main forum not in the "discuss" subforum created specifically for discussion of moderating problems. And these his remarks "slip past the moderators" as he puts it, confirming my opinion in the "Audiophile Press" thread that the new guidelines are not only emasculating but also very, very difficult to administer fairly. And his delicate sense of propriety (he doesn't like "condescension") does not extend to insinuation like "REAL expert's take" against mine obviously very "UNREAL" expertise ( the honour by the way I never claimed). So I had a look at Sean Olive's text. It seems that someone posted a letter in the "Stereophile" proposing comparing components placed one on the left the other one on the right side. Period. No RANDOM CHANGES from side to side which are the CRUX of the left-right method. (I repeated it so many times that by now it nearly sends me to sleep). S. Olive as befits a real scientist spends many lines proving that a stupid method is indeed a stupid method. But where is our Mr. Marcus in all this? Why does he grab onto this 100% irrelevancy, pins it on me, clouds the issue and sends readers on a 100% irrelevant wild goose chase? It wouldn't be because he's hard-pressed for a rational argument, would it? Well, we've been there before. He did the same in regard to his let's call it... "misreporting" of Greenhill's article. There too he made , let's say... "mistaken" claims that suited his case and when faced with Greenhill's own tables and conclusions sent the readers "to see for themselves". When I quoted ALL the relevant portions of Greenhill's conclusions verbatim and challenged him to give one single quote contradicting it, he seeming somewhat upset and irritable, said that he'll no longer discuss Greenhill in RAHE. What now Mr. Marcus? Will you still discuss S. Olive's opinions or refuse to, forcing me to quote all of it here? Ludovic Mirabel |
#54
|
|||
|
|||
What is so high end about high end?
|
#55
|
|||
|
|||
What is so high end about high end?
ludovic mirabel wrote:
chung wrote in message ... Bob Marcus wrote: (ludovic mirabel) wrote in message news:%wV4b.329114$o%2.151942@sccrnsc02... Snip previous discussion. Chung said: Actually I don't think anyone is interested in doing a test to find differences/preferences this way. Example: an important parameter of an audio component is its ability to convey image. This test uses mono signals and only exercises one channel of the device under test. I thought one of the biggest complaints of the original poster was that the DBT test is different than how people listen to their stereos. This L/R test is so different than the way we listen to audio equipment that I am totally surprised that he would even consider it. Not to mention the fact that any dynamic interchannel effects (like crosstalk, power supply effects, imbalance of frequency response in a RIAA circuit, etc.) will be totally undetected. The more I think about the test, the more I am amazed at what it does not do. I am completely flabbergasted. When you say: "This test uses mono signals and only exercises one channel of the device under test" do you mean that you concluded that I'm using ONE channel at a time? No. You were saying that you compared *one* channel of one product vs. *one* channel of another product. You were listening to both L and R channels, but those channels were processed by different pieces of equipment (for example, L channel came from one channel of amp A, and right channel came from one channel of amp B). And you said you were using mono material. I'm upset because it proves that I'm unable to communicate very obvious and simple ideas to some readers. And it saddens me that you'd attribute such brain-lame ideas to me. You said it, not me . Please answer this: how do you compare the stereo image ability of one product vs another using your method? Of course BOTH channels are playing at the same time. The channel levels are matched before the first change and never touched after that. Your assistant Xchanges the two components (which of course are electrically comparable) at random from one side to the other. The subject either does or does not consistently recognise the locations . Perhaps you missed the long word in capitals SIMULTANEOUSLY- I'm reposting it for your attention . I'm reposting also the explanation why slight level difference REINFORCES the validity of a correct answerif anything. If that too is not clear please don't hesitate to ask. You reproach me for repetition but it would appear that I have not yet repeated myself enough. Reprint: (regretfully and with apologies) 1) Get a monophonic or near monophonic (eg. centred soprano) signal source. MUSICAL, not an artefact. 2) On the left insert one component, on the right the OTHER ONE- (in the case of interconnects using two of one kind together i.e.source to preamp and preamp to amp on each side will give better contrast.) 3) Listen -write down your preference, get blinded. 4) An assistant now changes AT RANDOM (coin throw) both components from one side to the other or (of course) leaves them where they are keeping the records. 5) This is repeated minimum 15 times- for any length of time and with interval for lunch if you like. EVERY TIME you note your preference The repetition and change are the CRUX. So far, it seems like I understand your method. You are trying to find a preference of one product over another, by comparing one channel of one product vs one channel of the other product. As I said before, this is very different than the way you listen to music, which is using both channels of the same product(s) at the same time. The rest of your post snipped, due to repetition of previously posted material. Ludovic Mirabel |
#56
|
|||
|
|||
What is so high end about high end?
chung wrote in message ...
ludovic mirabel wrote: chung wrote in message ... Bob Marcus wrote: (ludovic mirabel) wrote in message news:%wV4b.329114$o%2.151942@sccrnsc02... Snip previous discussion. Chung said: Actually I don't think anyone is interested in doing a test to find differences/preferences this way. Example: an important parameter of an audio component is its ability to convey image. This test uses mono signals and only exercises one channel of the device under test. I thought one of the biggest complaints of the original poster was that the DBT test is different than how people listen to their stereos. This L/R test is so different than the way we listen to audio equipment that I am totally surprised that he would even consider it. Not to mention the fact that any dynamic interchannel effects (like crosstalk, power supply effects, imbalance of frequency response in a RIAA circuit, etc.) will be totally undetected. The more I think about the test, the more I am amazed at what it does not do. I am completely flabbergasted. When you say: "This test uses mono signals and only exercises one channel of the device under test" do you mean that you concluded that I'm using ONE channel at a time? No. You were saying that you compared *one* channel of one product vs. *one* channel of another product. You were listening to both L and R channels, but those channels were processed by different pieces of equipment (for example, L channel came from one channel of amp A, and right channel came from one channel of amp B). And you said you were using mono material. I'm upset because it proves that I'm unable to communicate very obvious and simple ideas to some readers. And it saddens me that you'd attribute such brain-lame ideas to me. You said it, not me . A great quip. How do you think of such funny things to say just like that on the spur of the moment.? I said it, not you Please answer this: how do you compare the stereo image ability of one product vs another using your method? I don't. It is not perfect. Just a compromise for easy use at home. Shall I say what else is not perfect in other tests for comparing components? Of course BOTH channels are playing at the same time. The channel levels are matched before the first change and never touched after that. Your assistant Xchanges the two components (which of course are electrically comparable) at random from one side to the other. The subject either does or does not consistently recognise the locations . Perhaps you missed the long word in capitals SIMULTANEOUSLY- I'm reposting it for your attention . I'm reposting also the explanation why slight level difference REINFORCES the validity of a correct answerif anything. If that too is not clear please don't hesitate to ask. You reproach me for repetition but it would appear that I have not yet repeated myself enough. Reprint: (regretfully and with apologies) 1) Get a monophonic or near monophonic (eg. centred soprano) signal source. MUSICAL, not an artefact. 2) On the left insert one component, on the right the OTHER ONE- (in the case of interconnects using two of one kind together i.e.source to preamp and preamp to amp on each side will give better contrast.) 3) Listen -write down your preference, get blinded. 4) An assistant now changes AT RANDOM (coin throw) both components from one side to the other or (of course) leaves them where they are keeping the records. 5) This is repeated minimum 15 times- for any length of time and with interval for lunch if you like. EVERY TIME you note your preference The repetition and change are the CRUX. So far, it seems like I understand your method. You are trying to find a preference of one product over another, by comparing one channel of one product vs one channel of the other product. As I said before, this is very different than the way you listen to music, which is using both channels of the same product(s) at the same time. It is. It is not perfect. It does not pretend to be scientific. Shall I say what other test method is not like listening to music? Especially if you're using pink noise that some of your colleagues (perhaps not you, in which case I apologise) claim is the best testing signal for that other well-publicised "test". The rest of your post snipped, due to repetition of previously posted material. On the whole if this is all you can criticise a reader might take it as a recommendation. By the way- did you try it? Ludovic Mirabel |
#57
|
|||
|
|||
What is so high end about high end?
ludovic mirabel wrote:
chung wrote in message ... ludovic mirabel wrote: chung wrote in message ... Bob Marcus wrote: (ludovic mirabel) wrote in message news:%wV4b.329114$o%2.151942@sccrnsc02... Snip previous discussion. Chung said: Actually I don't think anyone is interested in doing a test to find differences/preferences this way. Example: an important parameter of an audio component is its ability to convey image. This test uses mono signals and only exercises one channel of the device under test. I thought one of the biggest complaints of the original poster was that the DBT test is different than how people listen to their stereos. This L/R test is so different than the way we listen to audio equipment that I am totally surprised that he would even consider it. Not to mention the fact that any dynamic interchannel effects (like crosstalk, power supply effects, imbalance of frequency response in a RIAA circuit, etc.) will be totally undetected. The more I think about the test, the more I am amazed at what it does not do. I am completely flabbergasted. When you say: "This test uses mono signals and only exercises one channel of the device under test" do you mean that you concluded that I'm using ONE channel at a time? No. You were saying that you compared *one* channel of one product vs. *one* channel of another product. You were listening to both L and R channels, but those channels were processed by different pieces of equipment (for example, L channel came from one channel of amp A, and right channel came from one channel of amp B). And you said you were using mono material. I'm upset because it proves that I'm unable to communicate very obvious and simple ideas to some readers. And it saddens me that you'd attribute such brain-lame ideas to me. You said it, not me . A great quip. How do you think of such funny things to say just like that on the spur of the moment.? I said it, not you It's funny because if I had made that comment, it would not have passed moderation. Please answer this: how do you compare the stereo image ability of one product vs another using your method? I don't. It is not perfect. Just a compromise for easy use at home. Shall I say what else is not perfect in other tests for comparing components? Not perfect? How about close to useless? If you read audio reviews, stereo imaging is one of the biggest measures of how well a piece of gear is perceived. When you read about differences between amps, wires, CD players, etc., inevitably the ability to produce pin-point imaging is a necessary property of a desirable audio product, and it's what differentiates products. How can you set up a preference without observing how well the imaging is? You have other major problems, too. If you only test one channel of a product, you are essentially ignoring all the (a) interchannel effects like crosstalk, matching, etc., as well as (b) not exercising a product under actual-use conditions like driving *both* channels. What if the power supply cannot handle driving two channels to the loudness you want? Does anyone really listen to equipment the way you do your test? (snip) The repetition and change are the CRUX. No, the imperfections that you tried to glance over are the crux of why yours is a poor method. So far, it seems like I understand your method. You are trying to find a preference of one product over another, by comparing one channel of one product vs one channel of the other product. As I said before, this is very different than the way you listen to music, which is using both channels of the same product(s) at the same time. It is. It is not perfect. It does not pretend to be scientific. That part is definitely clear. Another big problem is that the test is very poorly controlled. As others have pointed out, the left and right speakers do not sound the same even if they are the same model, due to placement and room acoustics. When you tried to equalize level, using a mono or "mostly mono" source, you are balancing left and right at a small subset of audible frequencies (and 0.5 dB steps are not fine enough BTW, and have you confirmed that they are actually 0.5dB steps?). Those speakers will not be giving identical response over the complete audio range. Therefore, in your listening test, if you use "mostly mono" or stereo material, there will be imaging cues. For instance, if you use mostly mono source the 1st violin may sound a little bit left of center in one setting, but if left and right were swapped, it will sound a little bit right of center. That tells you L-R has been swapped. That makes the test poorly controlled since the subject knows something has changed. Shall I say what other test method is not like listening to music? With DBT, you can listen to music the way you normally audition equipment at a store, i.e., listen for differences in normal reproduction. Plus, DBT exercises the equipment the way they are meant to be used. Huge improvements over yours. Especially if you're using pink noise that some of your colleagues (perhaps not you, in which case I apologise) claim is the best testing signal for that other well-publicised "test". The rest of your post snipped, due to repetition of previously posted material. On the whole if this is all you can criticise a reader might take it as a recommendation. Huh? That is all I can criticize? Have you read carefully? Have you skipped the other "concerns" posted here? In fact, I have not found one thing about your method that is an improvement over DBT. By the way- did you try it? No, and certainly no plans to, given these huge deficiencies of your method. Ludovic Mirabel |
#58
|
|||
|
|||
What is so high end about high end?
chung wrote in message . net...
ludovic mirabel wrote: chung wrote in message ... ludovic mirabel wrote: chung wrote in message ... Bob Marcus wrote: (ludovic mirabel) wrote in message news:%wV4b.329114$o%2.151942@sccrnsc02... See discussion below: Mirabel: Shall I say what other test method is not like listening to music? Chung: With DBT, you can listen to music the way you normally audition equipment at a store, i.e., listen for differences in normal reproduction. Plus, DBT exercises the equipment the way they are meant to be used. Huge improvements over yours. At least on paper. So using left-right I prefer one component even though I can not judge imaging, crosstalk, minute residual speaker imbalances, (inaudible after careful equalisation with my two digital equalisers) etc. (Of course you have also telepathic vision concerning my preamp's custom level controls deficiences). If I tell you that the center image does not move you're free to believe me or not. If I tell you that next I put my preferred component in both channels and return for refund if horrible things happen to imaging, crosstalk, power what not you may say it is not research. No it is not. It is just a modest home-practical way of avoiding the nightmare of sighted bias that I read about in RAHE every day. That's all. It does not replace experienced listening to a variety of good recordings and it adds nothing to it. (I discover nothing "subtle" that I do not hear otherwise) Now let's see what you have on offer for a $ 600:00 switch. It ought to be a huge improvement with all the the loving speculative theory you lavish on it. I must come now to the boring part. You heard it so many times before- in fact it gives you headache. You may have even answered it and I missed it but just this one time repeat it- if only for such as All Ears waiting for the Voice of the Science to speak out. It so happens that ALL the available published audiophile panel ABX component comparisons had one outcome and one only: "They all sound the same". All: cables, preamps, amps, cdplayers, dacs. Now Mr. All Ears has 3 choices: 1) Everything in audio really does sound the same (unless it measures grossly diferently eg. Wattage etc). Of course many think that if you believe this nonsense you'll believe anything. Or: 2) There is something the matter with the "great improvement" which makes it useless for a cross- section of audiophiles component comparison test.(Research is another matter entirely) 3) Buy the switch and sign up for PC courses. But what if you're one of the hopeless ABXers (ask the researchers about them?) How will you know if you are hopeless or if everything really sounds the same? Ludovic Mirabel P.S. Next, if the "great improvement" does not pan out tell us about concentrating on the speakers- but how do you know that if you ABXed good full range speakers with a listener panel they wouldn't sound the same to most of them too? Room improvement? If you had my room you'd know that selling the house would be the only remedy. I tried wall fibreglass pads, bookshelves located strategically, bags in the corner and all that. I heard much, much less difference than between the silver cable a la Kimber and the zipcord. Previous text: (if you lasted this far) Especially if you're using pink noise that some of your colleagues (perhaps not you, in which case I apologise) claim is the best testing signal for that other well-publicised "test". The rest of your post snipped, due to repetition of previously posted material. On the whole if this is all you can criticise a reader might take it as a recommendation. Huh? That is all I can criticize? Have you read carefully? Have you skipped the other "concerns" posted here? In fact, I have not found one thing about your method that is an improvement over DBT. By the way- did you try it? No, and certainly no plans to, given these huge deficiencies of your method. Ludovic Mirabel |
#59
|
|||
|
|||
What is so high end about high end?
Hi Ludovic
I too have tried some of the modes of testing you suggest. Having found them being suggested in a different forum. I used it on some speaker cables in purely subjective tests, using no meters scopes or gauges of any sort. As we all know by now (having been told SO MANY TIMES) that copper is copper and all differences are at least 140db below 10V or 10Amps ,non linear non proportional. When using different wires for Left and Right speakers in mono mode I of course did NOT hear and tonal difference or change in the noise floor, and in stereo i did NOT hear the sound image wander about the sound stage, the depth of the stereo image was NOT grossly distorted, and these effect were NOT highly frequency(note) dependant. It is NOT a method for determining any differences in any part on the signal chain, most definitely not revealing any difference in amplifiers. You canNOT easily isolate or nullify the variations of the room effects / speaker placing by swapping the left and right wires. But as these tests were done on AB basis but weren't done DBT all of the above results are of course NOT valid. "ludovic mirabel" wrote in message news:7dn8b.425040$o%2.192669@sccrnsc02... chung wrote in message . net... ludovic mirabel wrote: chung wrote in message ... ludovic mirabel wrote: chung wrote in message ... Bob Marcus wrote: (ludovic mirabel) wrote in message news:%wV4b.329114$o%2.151942@sccrnsc02... See discussion below: Mirabel: Shall I say what other test method is not like listening to music? Chung: With DBT, you can listen to music the way you normally audition equipment at a store, i.e., listen for differences in normal reproduction. Plus, DBT exercises the equipment the way they are meant to be used. Huge improvements over yours. At least on paper. So using left-right I prefer one component even though I can not judge imaging, crosstalk, minute residual speaker imbalances, (inaudible after careful equalisation with my two digital equalisers) etc. (Of course you have also telepathic vision concerning my preamp's custom level controls deficiences). If I tell you that the center image does not move you're free to believe me or not. If I tell you that next I put my preferred component in both channels and return for refund if horrible things happen to imaging, crosstalk, power what not you may say it is not research. No it is not. It is just a modest home-practical way of avoiding the nightmare of sighted bias that I read about in RAHE every day. That's all. It does not replace experienced listening to a variety of good recordings and it adds nothing to it. (I discover nothing "subtle" that I do not hear otherwise) Now let's see what you have on offer for a $ 600:00 switch. It ought to be a huge improvement with all the the loving speculative theory you lavish on it. I must come now to the boring part. You heard it so many times before- in fact it gives you headache. You may have even answered it and I missed it but just this one time repeat it- if only for such as All Ears waiting for the Voice of the Science to speak out. It so happens that ALL the available published audiophile panel ABX component comparisons had one outcome and one only: "They all sound the same". All: cables, preamps, amps, cdplayers, dacs. Now Mr. All Ears has 3 choices: 1) Everything in audio really does sound the same (unless it measures grossly diferently eg. Wattage etc). Of course many think that if you believe this nonsense you'll believe anything. Or: 2) There is something the matter with the "great improvement" which makes it useless for a cross- section of audiophiles component comparison test.(Research is another matter entirely) 3) Buy the switch and sign up for PC courses. But what if you're one of the hopeless ABXers (ask the researchers about them?) How will you know if you are hopeless or if everything really sounds the same? Ludovic Mirabel P.S. Next, if the "great improvement" does not pan out tell us about concentrating on the speakers- but how do you know that if you ABXed good full range speakers with a listener panel they wouldn't sound the same to most of them too? Room improvement? If you had my room you'd know that selling the house would be the only remedy. I tried wall fibreglass pads, bookshelves located strategically, bags in the corner and all that. I heard much, much less difference than between the silver cable a la Kimber and the zipcord. Previous text: (if you lasted this far) Especially if you're using pink noise that some of your colleagues (perhaps not you, in which case I apologise) claim is the best testing signal for that other well-publicised "test". The rest of your post snipped, due to repetition of previously posted material. On the whole if this is all you can criticise a reader might take it as a recommendation. Huh? That is all I can criticize? Have you read carefully? Have you skipped the other "concerns" posted here? In fact, I have not found one thing about your method that is an improvement over DBT. By the way- did you try it? No, and certainly no plans to, given these huge deficiencies of your method. Ludovic Mirabel |
#60
|
|||
|
|||
What is so high end about high end?
"chris" wrote in message ...
Hi Ludovic I too have tried some of the modes of testing you suggest. Having found them being suggested in a different forum. I used it on some speaker cables in purely subjective tests, using no meters scopes or gauges of any sort. As we all know by now (having been told SO MANY TIMES) that copper is copper and all differences are at least 140db below 10V or 10Amps ,non linear non proportional. When using different wires for Left and Right speakers in mono mode I of course did NOT hear and tonal difference or change in the noise floor, and in stereo i did NOT hear the sound image wander about the sound stage, the depth of the stereo image was NOT grossly distorted, and these effect were NOT highly frequency(note) dependant. It is NOT a method for determining any differences in any part on the signal chain, most definitely not revealing any difference in amplifiers. You canNOT easily isolate or nullify the variations of the room effects / speaker placing by swapping the left and right wires. You do "nullify and isolate" the room/speaker effects VERY EASILY if you random- swap your component be it the amp, OR the interconnects from one side of the system to the other. If you still prefer, blind, one amp or one type of interconnects whichever side of the room they are on you "nullified and isolated" the room/speaker effects. This seems to me so obvious that I suspect that you did not follow the swapping protocol as outlined below. If you don't swap you're wasting your time. But as these tests were done on AB basis but weren't done DBT all of the above results are of course NOT valid. The "double blind" technique was a necessary forward step in medical drug research so that the researcher did not impart his enthusiasm to the patient. If your helper is your wife who couldn't care less, single blind is good enough. There is no need to slavishly imitate methods designed for completely different tasks just to get the reflected glory of the prestigious name. And using the name does NOT automatically confer validity. Thanks for your interest. Ludovic Mirabel "ludovic mirabel" wrote in message news:7dn8b.425040$o%2.192669@sccrnsc02... chung wrote in message . net... ludovic mirabel wrote: chung wrote in message ... ludovic mirabel wrote: chung wrote in message ... Bob Marcus wrote: (ludovic mirabel) wrote in message news:%wV4b.329114$o%2.151942@sccrnsc02... See discussion below: Mirabel: Shall I say what other test method is not like listening to music? Chung: With DBT, you can listen to music the way you normally audition equipment at a store, i.e., listen for differences in normal reproduction. Plus, DBT exercises the equipment the way they are meant to be used. Huge improvements over yours. At least on paper. So using left-right I prefer one component even though I can not judge imaging, crosstalk, minute residual speaker imbalances, (inaudible after careful equalisation with my two digital equalisers) etc. (Of course you have also telepathic vision concerning my preamp's custom level controls deficiences). If I tell you that the center image does not move you're free to believe me or not. If I tell you that next I put my preferred component in both channels and return for refund if horrible things happen to imaging, crosstalk, power what not you may say it is not research. No it is not. It is just a modest home-practical way of avoiding the nightmare of sighted bias that I read about in RAHE every day. That's all. It does not replace experienced listening to a variety of good recordings and it adds nothing to it. (I discover nothing "subtle" that I do not hear otherwise) Now let's see what you have on offer for a $ 600:00 switch. It ought to be a huge improvement with all the the loving speculative theory you lavish on it. I must come now to the boring part. You heard it so many times before- in fact it gives you headache. You may have even answered it and I missed it but just this one time repeat it- if only for such as All Ears waiting for the Voice of the Science to speak out. It so happens that ALL the available published audiophile panel ABX component comparisons had one outcome and one only: "They all sound the same". All: cables, preamps, amps, cdplayers, dacs. Now Mr. All Ears has 3 choices: 1) Everything in audio really does sound the same (unless it measures grossly diferently eg. Wattage etc). Of course many think that if you believe this nonsense you'll believe anything. Or: 2) There is something the matter with the "great improvement" which makes it useless for a cross- section of audiophiles component comparison test.(Research is another matter entirely) 3) Buy the switch and sign up for PC courses. But what if you're one of the hopeless ABXers (ask the researchers about them?) How will you know if you are hopeless or if everything really sounds the same? Ludovic Mirabel P.S. Next, if the "great improvement" does not pan out tell us about concentrating on the speakers- but how do you know that if you ABXed good full range speakers with a listener panel they wouldn't sound the same to most of them too? Room improvement? If you had my room you'd know that selling the house would be the only remedy. I tried wall fibreglass pads, bookshelves located strategically, bags in the corner and all that. I heard much, much less difference than between the silver cable a la Kimber and the zipcord. Previous text: (if you lasted this far) Especially if you're using pink noise that some of your colleagues (perhaps not you, in which case I apologise) claim is the best testing signal for that other well-publicised "test". The rest of your post snipped, due to repetition of previously posted material. On the whole if this is all you can criticise a reader might take it as a recommendation. Huh? That is all I can criticize? Have you read carefully? Have you skipped the other "concerns" posted here? In fact, I have not found one thing about your method that is an improvement over DBT. By the way- did you try it? No, and certainly no plans to, given these huge deficiencies of your method. Ludovic Mirabel |
Reply |
|
Thread Tools | |
Display Modes | |
|
|
Similar Threads | ||||
Thread | Forum | |||
Help -- connecting an amplifier using a high input | Car Audio | |||
High Pass Filtering - How Audible? | Audio Opinions | |||
Direct Connect Hub With Only High Quality MP3s? | Audio Opinions | |||
"High Power" Head Unit to Power Separates? | Car Audio | |||
High End Hyperbole | Audio Opinions |