Steve, Tom Nousaine published a very interesting article in Sound & Vision detailing the results of blind tests of speaker cables. He had audiophile listeners compare 16 AWG zip cord to their own chosen brand/model of speaker cable, in their own homes, on their own systems, with their own chosen playback source and media. He gave the listeners a choice between ABX switching and cable-swap (scored as same-different) methods (except the last listener; for her, he used both ABX and the cable-swap method because the others had chosen ABX). No time limits; in fact, he let the listeners warm up and practice so they would become as comfortable as possible with the tests. A minimum of 10 trials per listener. None of the listeners, despite their high opinions of their own hearing prowess, could correctly identify the cables they were listening to a statistically significant number of times: in comparing the zip cord to a set of T1 bi-wires, one listener guessed correctly 3 out of 10 times, which is within the likely range of results that would arise from just guessing or flipping a coin. Another listener's results were 4 out of 10; again, within the range of strictly chance.
In short, the listeners were asked to prove that they could indeed hear audible differences between the cables, and none succeeded.
Regarding ABX: Some may allege that rapid A-B switching "hides" real differences. That's one vague assertion! My opinion is to the contrary: it allows the closest thing to actual side-by-side comparison possible in audio. Obviously we can't listen to two things simultaneously and judge between them; the best we can do is put the same audio through both DUTs, match the levels, and freely switch between them as needed to make the comparison.
