Computer Audio Asylum: RE: Better duck under that desk; by Ryelands

Phelonious Ponk wrote:

I'm not saying the high-res people were deluding themselves. That's not the point.

If Meyer and Moran are right to say that it is not possible to distinguish 44.1 KHz from media such as SACD, then the high-res wallahs are deluding themselves. Iâ€™ve yet to hear either SACD or 192/24 recordings but there have been enough reports from competent people on this list and elsewhere that there is a difference to prod me into examining the Meyer & Moran paper with a critical eye.

The point is that Kuchner's study say[s] absolutely nothing on the subject.

Kuchner demonstrated that humans can detect differences in percepts that cannot be resolved by 44.1 KHz recordings. That would affect inter alia the perception of timbre, especially on fast transients. Whether, in practice, on the kit most of us use, this matters (or even whether he is right) is a proper debate but it is nonsense to say his work is irrelevant. If, after all this debate, you canâ€™t see that, we must simply agree to disagree.

I'm not sure where you get the faulty kit, bad sampling procedures, etc. There is nothing in the report that would lead to that conclusion.

Thereâ€™s plenty - if you choose to look. An experienced eye helps but a willingness to scrutinise the text is a good start. The authors, whose integrity is not in question, deserve no less.

Faulty kit :- I got this from, well, Meyer and Moranâ€™s description of faults in their kit. See text.

Bad sampling :- No details are provided of the size and composition of the subject group except that some were students on a recording course, others were â€œinterested partiesâ€ and â€œabout 60â€ were BAS members. The latter group especially was likely to be well aware of the views of Meyer and Moran on the issue of 44.1 KHz and â€œperfect sound foreverâ€.

Competent research must therefore either exclude it from the trials or design (and report on) procedures to evade the problem. Given the circumstances, the first precaution might have been impracticable. The second is therefore critical. There is no report that either was taken so I concluded, not unreasonably, that neither was.

Shoddy procedures :- No details are provided of the procedures used though, as some tests were â€œsightedâ€ and some â€œblindâ€ and as there was no control set, this matters.

If Hi-Res playback precedes RBCD, the subject has prior knowledge of any difference. This will colour his/her perception. All we are given is a picture of an ABX box and â€œA-X-Bâ€ display. How they were used or even what they do, we are not told. There may be reasons for telling subjects which version is under test - but Iâ€™d like to know what they are.

They donâ€™t give any so I concluded, again not unreasonably, that the authors (who report no pertinent experience in their biographies) were simply unaware of these very basic design issues.

The Ã¼ber-howler :- I am embarrassed to own up to this but I have only just noticed that Meyer and Moran obtained their 44.1 KHz signal by digitising the analogue output of their SACD player . (See â€œTestsâ€, para 1 and Fig 1).

They have just assumed that the DACs on the SACD players used are capable of resolving the differences between SACD and 44.1 KHz. Though they suspect the marketing claims of the SACD lobby, they take their kit on trust.

If the SACD players used cannot adequately resolve the signal, Meyer and Moran can run 1,000 subjects through 100 trails on each of 25 SACDs and do so sighted, unsighted and helped by any guide dogs in the BAS. They will never find a difference whether itâ€™s there or not. The trials are not merely â€œflawedâ€, they are worthless.

In summary, itâ€™s no wonder they didnâ€™t find a difference: they used subjects likely to give them the answers they wanted, in all likelihood used procedures that did not prevent that and were, in any case, testing for the wrong thing. Otherwise, itâ€™s fine work.

And you can, of course, choose to accept the "fine work" if you choose.

Thank you. I look forward to your critique of it. The only point Iâ€™d make is that he is a bit vague as to his subject base though he (a) evades bias by design and (b) reports that he adhered to campus selection guidelines so my beef would be pretty muted.

It will, no doubt, be very useful the next time your setting up a system based on misaligned ribbons to play back your favorite square waves. It won't tell you much about the mysteries of music reproduction, though.

No one has suggested I should use mis-aligned ribbons to set up my audio. I donâ€™t use spectrum analysers, distortion meters and the like either but Iâ€™m glad that equipment designers do. Nor do I hire apparatus to determine the limits of my hearing but Iâ€™m glad that scientists do the work and that equipment designers read what they write.

There is much that is mysterious about hearing (and music) but to conclude from that that all of it is forever impenetrable is anti-scientific philistinism posing, in this case, as empirical rigour.

Kuncher demonstrated that the extreme limits of auditory perception were higher than previously thought due to the ear/brainâ€™s ability to perceive the modulation of sounds by suprasonic frequencies (i.e. not themselves directly perceptible). Audiophiles have argued the point for decades, sometimes sensibly, sometimes not and many equipment designers have tacitly acknowledged it for almost as long.

Whether Kuchner proved his point using speaker alignment, low-pass filtering or a poker up his backside is irrelevant to how I set up my audio kit but it could influence the debate on using higher resolutions for distributing recordings.

Audiophiles confuse me. "Trust your ears," they say. "Measurements aren't important."

Audiophiles (and philistines) can say what they like. Iâ€™m writing as an experimental psychologist who would never say that measurements arenâ€™t important, who has some knowledge as to how perception both deceives and informs and who gets angry at the energy with which ill-informed people persist in talking anti-scientific nonsense on both sides of the debate.

I have no personal interest in the outcome of the argument. My kit, which Iâ€™m stuck with for the mid- to long-term future, is aimed at the optimal reproduction of a large RBCD library; I donâ€™t use upsampling and I canâ€™t listen to hi-res recordings. But when a scientist peels back a corner, however small, of the â€œveil of mysteryâ€ that is great music, Iâ€™m intrigued.

Iâ€™m sorry that you and others canâ€™t seem to appreciate what Kuncher has done and genuinely disappointed that you canâ€™t debate the points at a higher level. Thatâ€™s partly what forums like this ought to be all about.

Until, of course, the most obscure and loosely relevant measurements appear to support their pre-conceived notions.

If you find youâ€™re in a hole, just stop digging.

Dave