Blind Tests & Bus Stops Letters, September 2005

Letters in response appeared in September 2005 (Vol.28 No.9):

Amplifiers & measurements

Editor: Although I am one of the few classical musicians I know who actually cares about the quality of reproduced sound, I have never considered my ears to be golden. Still, I can confidentially differentiate realistic-sounding equipment from crap. I attended the debate at Home Entertainment 2005 between John Atkinson and Arnold Krueger (if one can term any verbal interchange with Krueger a debate, or even a discussion). [An MP3 recording of the debate is available.—Ed.] I was fascinated to learn of John's experience with the Quad 405 amplifier because it exactly mirrored mine. Also, like John, I'd previously spoken Peter Walker's name only with hushed reverence, and had owned some of his stuff over the years.

In the late 1970s I was in Waterloo, Ontario, to perform a recital for the Kitchener-Waterloo Chamber Music Society. The Director of that organization, Jan Narveson, is one of Canada's true musical heroes, in terms of his vast and almost single-handed contribution to his community's cultural life. Jan also happens to be a close friend and colleague of Stanley Lipshitz, a very prominent member of the double-blind-testing, measurements-is-everything crowd. Jan had recently acquired a Quad 405, and proudly played it for me through his KEF 105As, which I also owned at the time. I told him that I found it more than a little harsh, and further, that I'd never heard an amplifier whose sound was even in the same ballpark as this one's. His response was, "I guess something is wrong with every other amplifier you've heard, because Stanley checked this one out and said it works perfectly."

Maybe Stanley measured the wrong things.—Robert Silverman, Vancouver, BC

Amplifiers & blind testing

Editor: Thinking there might be something new to be said in the perennial argument over double-blind testing and subjective evaluation, I dutifully listened to the entire debate between John Atkinson and devoted meter-reader Arny Krueger.

Nope. Granting that Mr. Krueger is not a polished speaker, and that the audience's questions were for the most part inaudible, charity does not allow me to concede to Mr. Krueger even one of his points.

Mr. Krueger is an engineer. Now, I don't mean to pick on engineers, but as one member of the debate audience observed, engineers are so intent on getting the correct answer that they often fail to examine the assumptions implicit in the problem. That's a job for a physicist. Consequently, they often lose sight of the purpose of the gear they test: to play music.

If Mr. Krueger had had a more liberal education, he would know that double-blind testing is only one method of conducting an experiment. It is equally valid, and the method of choice in the social sciences, to observe a phenomenon at length, record one's observations, then extrapolate those observations to the world at large. The science, Mr. Krueger, is in comparing one's predictions to what actually happens in the world at large. Double-blind testing is only one means to this end, and not the most important one, at that.

I'm trying hard not to be condescending, but I've heard the same tired, old arguments parroted over and over so often that my patience wears thin. Nevertheless, I will try once more to make my points clear.

First of all, the ABX test is an example of a forced-choice experiment. It assumes that a person can come to a conclusion after one short observation or a series of short observations. That is not always true, or we wouldn't have the term "buyer's remorse." It often occurs that we can go for days, or even years, before noticing a flaw in something.

Second, the process may not measure what you think it is measuring, or the measuring process itself might be contaminating the results. JA explained this in detail during the debate. It is worth mentioning that, at one time, a very prominent magazine tested amplifiers by passing their output through an equalizer and comparing the resulting signal to that from a standard unit. Laughable today, but at the time, it was thought to be the epitome of pure "science."

Third, I have noticed that very few "objectivists" are aware that their reliance on meter readings can be dangerously circular. They assume that their meter readings can detect any inaccuracy. Why is an amplifier accurate? Because the instruments do not show any distortion. Why don't the instruments show any distortion? Because the amplifier is accurate.

I could go on, but I rest my case.

Is there a place for comparative testing? Of course there is. I often use it in designing and modifying my equipment. It is the only practical way to quickly evaluate changes. Still, it can lead you astray, step by step, unless you occasionally compare the equipment with the only real standard, live music.

To his credit, Mr. Krueger listens to live music, a lot of it. His measurements may have some validity. I don't know. His mistake is in assuming that his method is the only real and appropriate test for musical accuracy.—Bob McIntyre, Toledo, OH,

Blind tests & psychoacoustics

Editor: John Atkinson's edifying anecdote concerning the objectivist vs subjectivist debate in the July issue ("As We See It," p.3) suggests a psychoacoustic phenomenon that I hold trust in: the auditory faculties deployed to discern distinctions between hardware are not the same as those deployed when one seeks delightful immersion in the musical experience. As such, while I readily concede my own inability to make correct and consistent judgments during blind tests, I experience no dissonance maintaining a clear preference for one piece of hardware over another after spending ample time engrossed in music that stirs my soul.—Matthew Posillico, Garden City, NY,

Blind testing & disappointment

Editor: I'm very disappointed to discover John Atkinson's position on blind testing ("As We See It," July 2005, p.3). As he is the editor of Stereophile magazine, it is now even more difficult for me to respect anything that I read in the publication. To deny the validity of blind testing is to ignore science. Blind testing is objective science. To ignore its value is equivalent to believing that the world is flat, that there is no truth in evolution, that man didn't land on the moon.

How difficult can it be to set up a blind test for amplifier power cords? Take an amplifier and change out power cords as frequently as the listener chooses without him knowing what cord you changed to, or if you changed it at all. Record not only whether or not he heard a difference, but which sounded better. Do this with a number of listeners, taking whatever time is required, and see if there is any statistical significance. (Does JA know what that means?) Can they actually detect a difference in power cords, and if so, can they determine which sounds better on a consistent basis?

If you can't demonstrate it statistically, then why should I spend $3000 on a power cord? Of course, if you demonstrate that there is no difference, then Stereophile will lose all of its advertisers manufacturing expensive power cords.

JA stated how much more he enjoyed the music after replacing his Quad 405 with an M&A tube amplifier. He did not blind-test the Quad 405 against the M&A, so how does this support his position against blind testing? There are many things that influence what we hear. Ninety-nine percent of people who trade in a $2000 amp for a $5000 amp are going to like the way it sounds better, even if the electronics in each box are identical. This is the way minds work. It costs more, so it has to be better.

In my opinion, Stereophile's editor is doing a real disservice to the magazine's readers by promoting snake oil, such as power cords that cost more than McIntosh amplifiers. When I hear him try to discredit the validity of blind tests, I have to conclude that his education must have been in journalism or marketing rather than in electronics or engineering. Like it or not, audio electronics are about science, scientific methods, and engineering. John Atkinson really needs to spend some time studying psychoacoustics.

Based on the new paradigm, how can anyone feel comfortable buying an audio system? I still read that people think McIntosh amps sound "too warm" and Krell sounds "too cold." These are some of the best amplifiers ever made, and I contend that it is a real stretch to find anything about either to criticize other than the Krell being hot (and perhaps having a shorter life due to the high temperatures). Tube amps? Don't get me started. Thirty years ago I bought the best Mac amp, preamp, tuner, and speakers. I essentially had the best system money could buy. Today you can never be satisfied, because you are being told monthly that there is something better. I read the other day that you need to replace your interconnects and speaker wires every few years. Unbelievable.

For the record, I've been a serious audiophile for 37 years. I continue to be amazed at the garbage I read in these publications, and now I see that it is coming from the top. I'm sick about what has happened to this hobby.—David Sanford,