Blind Listening:
Page 5 In retrospect (what a clear view you get through the retrospectoscope!), the drum track on the HFN/RR Test CD was a definite non-starter. Among the 4 available amp comparisons, there were 14 sets of comparisons with this selection in the 8 sessions. Table 5 shows that in only 1 of the 14 was the correctness of the responses any better than by chance alone. As can be seen from Table 4, the other six selections were all equally highly effective in demonstrating the overall ability to distinguish difference rather well but to determine sameness rather poorly.

What does one make of all this? Well, first, the differences between ability to distinguish amp difference and ability to determine sameness were so great (for the statistically oriented, with such low p values and such large ns) that any caviling about test conditions can be promptly rejected (footnote 9). Second, the results reaffirm the general belief that it's easier to tell when there's a difference between things than it is to be sure they're the same. Third, there are some things we'll never know, such as whether or not the ability to distinguish Adcom-first/VTL-second from the reverse was because there were half again as many opportunities to hear it (Adcom/VTL, 1134 chances; VTL/Adcom, 758 chances). This possibility doesn't hold for sameness—Adcom/Adcom and VTL/VTL had equal chances.

Also, be sure of this: these results will have no—repeat, no—impact on the polarized subjectivist/objectivist controversy about whether or not amps that measure similarly same will sound the same when not overdriven. There are too many critics and carpers around for any blind controlled listening test to fully satisfy them all. Some will find fault with this and some with that; this test was no exception.

Why'd we go to all that trouble? JA gave some of his reasons above; I was there for the fun of it and for what I'd learn, but also and more importantly, it seemed that this would be a very useful way to provide a short but cogent educational experience for a lot of audiophiles. Most of our listeners had never had the experience of trying to listen really critically and almost continuously for nearly an hour under well-controlled conditions. These folks developed a very real personal awareness of how much work is involved in careful comparative subjective sound evaluation, and this heightened their respect for those who do a lot of it and do it well. How do I know this? They said so, often and earnestly, after each session, in the halls later, and some even by mail to JA. Why would they lie?

Can these experiences and results lead to something else of value? I sure hope so. When they learned the price differential between two power amps that they themselves couldn't really tell apart, a large number of the listeners indicated that the extra bucks wouldn't be worthwhile to them personally. Often this comment took the form of "I'd buy the cheaper one and a new car." In turn, this train of thought leads on into concern about the cost/sonic-quality relationships of audio components, and thence into how that impacts on enjoyment of music reproduction in the home. Perhaps there will eventuate thoughtful explorations of these matters, either in the pages of Stereophile, or elsewhere, or both.

Final incidentals: Despite JA's warning admonition at the start of each session, some listeners did seem to have tried to identify specific amps. Because there was no standard way to do this on the score sheet, and because we couldn't tell whether these notes were made during the tests or jotted down afterwards as the correct answers were presented, we didn't try to methodically assess how well they did. But it was quite clear that most of those who tried this did rather less well than average on the decision they were supposed to be making. Also, from each session there were lots of adjacent response slips with the same last name but different first names (seemingly one male and one female), or different first initials, quite probably from couples. We specifically avoided comparing each of the pairs and reporting the results—tactfully, we thought.

We enjoyed it. Hope you will, too.—WH

---

Footnote 8: Larry Archibald suggests that the results indicate that the listeners at the show heard a difference with about the same degree of success whether there was a real difference or not. In other words, audiophiles live up to their calling in being committed to hearing differences regardless of reality. I agree that this will tend to make the listeners score less well when the amplifier remained the same, but could it be argued that it will help them score better when there is a difference? I think not, as anyone answering "Different" to every presentation would still have scored no better than chance, ie, 50%, or 3 or 4 out of 7.—JA