Blind Listening Letters
A "Golden Ear" or a "Lucky Coin"?
Editor: I've been called many things in my job as an equipment and music reviewer, but "lucky coin" ("Letters," May 1989) is by far the most aggravating. I confronted David "all amplifiers sound the same" Clark at the June 1988 CES and told him that I could hear differences among amplifiers and, furthermore, that anyone who couldn't ought not be reviewing them. He countered that unless I could demonstrate my ability in a double-blind test, my assertion was groundless.
When he called a few months later asking if I'd organize a double-blind test at the AES, I jumped at the opportunity. I worked long and hard, with help from many people in the audio community, to set up a test that would satisfy the measurement freaks, and I believe we did. I took my own test just once (like every other participant) with David Clark in the room, and scored five out of five correct identifications. Not only did I correctly identify "same or different," I volunteered which amp was which and got that right four out of five times as well.
Good enough? No. Statistically insignificant, I was told by the dominant Dr. Stanley Lipshitz wing of the AES. "Lucky coin," I'm called by reader Dayton. There's no satisfying those lacking discerning ears. They call it science. I call it jealousy.
I will never again take the trouble to organize or participate in a double-blind test under the aegis of the AES, because I have no doubt that close-minded collectivists parading as scientists will once again find a way to discredit the glory of individual achievement.
Incidentally, Clark attempted to insert equalizer networks in the amplifier circuits to flatten their frequency responses into the difficult load presented by the Infinity speakers. While the Threshold was essentially flat, the Crown amp was down 2dB at 20kHz. The VTL 500s had more significant deviations from flat response.
The Crown was restored to flat response, but David Manley (correctly, in my view) refused to allow Clark to flatten the VTLs, claiming that would be a test of equalizers, not amplifiers. (I'm simplifying his arguments to save space.) Clark and his cohorts claim that deviations from flat frequency response account for all audible differences among amplifiers.
That AES convention attendees—supposedly trained engineers—couldn't hear the obvious sonic difference between the VTLs and the two flat-response transistor amps bears sad witness to the low state of their art. But then, so does listening to most contemporary recordings.
While I've got the bully pulpit: I read lots of mainstream automobile magazines. Why is it that their readers never complain about reviews of sexy, expensive BMWs, Ferraris, Lamborghinis, and the like—cars most of them will never drive, much less own—while readers of your specialized audio magazine keep bitching and moaning about your coverage of exotic, expensive, high-performance audio equipment?—Michael Fremer, Senior Music Editor, The Abso!ute Sound, Sea Cliff, NY
Golden ears, or just trained?
Editor: It was a pleasure participating in John Atkinson's blind amplifier A/B test at the Stereophile High End show (footnote 1). Hearing his comments, along with those of others who took the test, brought several thoughts to the surface that I have been mulling over for some time.
1) An A/B blind listening test will prove whether or not the people being tested can hear differences, but it does not prove that there are differences.
2) Perhaps a person's ability to hear differences between two or more components is largely a learned skill. Some people learn faster and ultimately have more abilities than others. This seems true with hearing differences between components, just as some people are better at astrophysics than others, better at pole-vaulting than others, better Indy-car drivers, etc.
A helpful A/B blind component test would therefore have two parts:
First, a learning phase: Qualified (!?) test subjects would become acquainted with the room, equipment, and software being used, as well as training in what to listen for. This learning process would also include being tested under non-blind conditions (learning how to take the test). For some people, training might take hours; for others it might take several days.
Second, the testing: Once proper training has taken place, better testing results will be obtained, and we can finally get closer to supporting with statistics that amplifiers, interconnects, CD players, etc., do have differences ranging from subtle to extreme. People should still be allowed to judge for themselves, however, which differences they prefer.
As an owner of Audio Ecstasy in San Luis Obispo, California, I have spent the last several years critically listening to audio components in a variety of situations. I felt the Stereophile A/B test was fairly straightforward and was able to score six correct identifications out of seven comparisons. The people seated near me in the test did not fare so well, yet claimed that when told what they were listening to, they could easily hear differences.
Perhaps it should have been noted on your test sheets what the occupation of the subject was. The only other person in the room who scored as well as I did had a manufacturer's badge on. I wonder, did audio-industry people score better in general than non-industry people at the show?—Jon Iverson, San Luis Obispo, CA
PS: No, I do not manufacture Eagle power amps!
Footnote 1: I half expected to see you using a Carver M1.0t amplifier and the "mystery monoblocks" for your A/B test!