<snip>
You bang on and on about protocol rigor, when the fact is that audible difference -- between cables, Redbook vs hirez, lossy vs nonlossy, and other A's and B's that should NOT be expected to exhibit major audible differences - are typically claimed to be *so* great , by audiophiles and by marketers, that even a less than grade-A ABX protocol should easily reveal them. These claims are the very *currency* of audiophilia.
You were lamenting about a "moving goalpost" but aren´t you presenting such a "moving goalpost" by the statement cited above?
The "nonaudiophile believers" are usually argueing that a diffence can´t be heard (without further specification it has to be taken as "nobody is able to detect a difference") and to fast switch to another assertion - i.e. if not detected in every ABX setup it can´t be of relevance - is sort of a backdoor as the referring to "so great" actually is.
There is simply no common gauge for "great" or "so great" or "remarkable" or whatever description is sometimes used. I´ve written it before that verbalization of a multidimensional listening impression is a difficult task, but of course during in intensive occupation it surely happens that any description is an exaggeration (at least for people not dealing so intensely with it); IIRC Dave Moulton once wrote a nice piece about this occurence.
To be more specific, as we know from the experiments for "inattentional blindness" and the later experiments for "inattentional deafness" it seems to be quite easy to _not_ detect differences that are usually considered to be quite obvious.
Does the "gorilla" or the electro guitar qualify for the term "great/ so great" difference?
Or what about a difference that a listener does not detect at first but never fails to detect after being teached about it? Is it of relevance?
I´d say it depends which means one does not know in advance.
Additionally it depends on the conclusions/statement based on negative "ABX-results" (means ABX and all other test protocolls), the more generalized/categorical the more rigorous the experimental conditions should have been.
And while you , I'm guessing, do 'believe in' the utility of ABX , others who don't, and who really don't care a fig about protocol rigor (heck, they may even be on record as saying DBTs don't work for audio, period), suddenly become zealous about it as it suits them. That's a political tactic, not a scientific debate.
I hope we can agree that this a totally different topic.
And personally i experienced the mentioned difficulties with the specific conditions of the ABX protocol myself and in trying it with two other participants, so dropped it.
But i have to acknowledge that others obviously are doing well with ABX so will not generally dismiss it.
(One alleged reason for the difficulties i´ve cited earlier back from the 1960s)