> The obvious fallacy here is Atkinson's constant implicit
> assertion that all DBTs are equally trustworthy. OK, so
> Atkinson did one tiny little DBT 20 years ago that mislead > him.
I refer you to the recording of our debate, Mr. Krueger, in which I clearly listed the very large number of blind tests in which I have particpated over the past 28 years. This recording is available at http://www.stereophile.com/news/050905debate. I mentioned the 1978 test as being representative, not because it was the only such test in which I have participated.
> We really know nothing about Atkinson's 1978 Quad
They were reported in exhaustive detail in the November 1978 issue of Hi-Fi News. My point is that at that time, I had no reason to doubt the results of the tests. And as I reported in the July Stereophile, I told this anecdote at the debate to make two specific points. First, it demonstrates that my following the then-as-now "objectivist" mantra -- that audiophiles should buy the cheapest amplifier that offers the power and features they need -- had let me down. Second, it pits against one another two core beliefs of the believers in "scientific" testing: 1) that a blind test, merely by being blind, reveals the reality of audible amplifier differences; and 2) that sighted listening is dominated by nonaudio factors, the so-called "Placebo Effect."
To explain my quarter-century-old Damascene experience, you have to accept that either the blind test was flawed -- in which case all the reports that cited that 1978 test as "proving" the amplifiers sounded the same were wrong -- or that the nonaudio factors were irrelevant, in which case the criticisms of sighted listening based on that factor must be wrong.
We still don't know much about the blind test in 1978. I certainly don't have a copy of the article. What was the output impedance of the tube amp? What was the impedance vs. frequency curve of the speakers used? Also, of your own speakers? Or, more simply, what was the frequency response of the amps into the speaker loads?
It also seems that you have set up a false dichotomy, which Mr. Krueger's response has already shown. And of course, to talk of a blind test proving that two amplifiers sound the same is an overstatement, as statistics don't give absolute proof. As to the dichotomy you have set up, we are not simply left with a blind test in which all non-audio factors were eliminated which nonetheless failed, or a declaration that the non-audio factors were irrelevant, thus justifying sighted auditioning.
How do you know that the failure of the DBT to detect significant differences was due to non-audio factors? Mr. Krueger already suggested that it might be that the program material selected was not as good as it might be, that listeners might better detect differences with something else. We don't have the measurements of the FR of both amps into the speaker load so as to see what the differences were.
By what reasoning do you arrive at the conclusion that the failure of a particular DBT says anything about how reliable sighted listening is? I can't make any sense out of your argument here.