This was odd. Why should a meeting addressing the sound of audio, attended by people whose lives are dedicated to studying the recording and reproduction of music, be considered "unusual" and "out of the mainstream"? And why the hostility toward those who judge an audio component primarily by listening to music through it?
These questions cut to the heart of the underlying conflict between science and human values. Unlike other endeavors, however, where the result of science is more obvious (the measurement of a bridge's strength, for example), audio reproduction is different in that the goal of good audio engineering—the satisfying communication of a musical experience—is an intensely personal event that defies analysis by scientific method. This situation, in which the result of science cannot be quantified by any scientifically acceptable measure, offers an opportunity to explore the relationship between science and human experience. I would like to go beyond the traditional battleground of audio "subjectivists" and "objectivists," "meter readers" and "golden ears," to recognize a common area of understanding that neither thinks exists. Each camp can benefit from the other's knowledge, but summarily rejects the opposing position without realizing that an acceptance of the other's principles does not mean rejecting one's own beliefs.
What I have to say should in no way be considered a condemnation of the valuable contributions made by the AES and its members. Nor should it be thought of as an attack on science and the scientific method. Rather, it should be viewed as an expansion of thought that encompasses the tenets of science without rejecting the very real and important contribution the musical experience can add to the advancement of audio reproduction. Anyone who has read my previous writings in these pages will know that I am most certainly not "anti-technology" or "anti-science." I find beauty in the underlying form of audio technology—how it works—as well as the visceral, immediate result of that technology—musical expression.
Let's first examine the very different ways of looking at the world that are the source of the conflict.
During the conference, there was blatant and widespread antagonism toward the audiophile. In fact, the mere mention of the word "audiophile" (usually in a derisive tone of voice) brought contemptuous laughter from the audience. The general belief among the AES appears to be that audio engineering is sufficiently advanced that virtually no sonic differences exist between competently designed and manufactured products. Furthermore, it is believed that any phenomenon that can be heard can be measured with existing instruments. If a sonic difference cannot be measured, it doesn't exist. Those who hear differences (audiophiles) but cannot prove their existence rationally are grouped with believers in astrology and a flat earth (footnote 1). More important, however, is the AES's exclusion of the listening experience from the study of audio (footnote 2). If all differences can be measured, why listen? And since no significant differences exist between components (not to mention cables), the entire high-end audio business is presumably a fraud perpetrated on an unsuspecting public. Additionally, this belief structure is embraced and promoted by the mainstream audio magazines.
The audiophile contends that each component introduces a unique sonic signature, affecting communication of the musical experience. Everything, from capacitors to cables to the purity of the AC powerline, can affect the perception of music. A component's musicality is evaluated by listening, not measurement. The conveyance of the musical experience, not the numbers generated by "objective" tests, is the true measure of a component's quality. The language used to express this musicality (or lack of it), terms like palpability, soundstage, grain, involving, and bloom, are attacked by engineers as poetic nonsense. This lexicon is in sharp contrast to the language of science used to describe an audio component: 0.002% THD, 0.001% IMD, 20-20kHz ±1dB, etc.
Thus the lines of division are drawn.
During a post-conference discussion with Stanley Lipshitz (footnote 3), a man who epitomizes the rational, scientific approach to audio, he asked a question that not only shocked me, but sparked the inspiration for this essay. He was expressing dismay at the idea put forth in print by both Martin Colloms and me: rather than measurements providing an indication of a component's quality, there was an inverse relationship between measurements and sound.
I expressed this view in my technical assessment of the Wadia 2000 D/A processor in Vol.13 No.1. To Dr. Lipshitz, this position not only negated the value of measurement, but was diametrically opposed to the very essence of science. Scientific method, with measurement as its primary tool, is supposed to move us away from the unknown and toward a greater understanding of nature. To state that a component with poor measured specifications may indicate better sonic performance was an outrageous blasphemy that threatened the very advancement of audio science (footnote 4).
At this point in the discussion, John Atkinson began to contend that the Wadia 2000 has measured performance far worse than the cheapest Japanese CD player, yet is widely regarded to have superb musicality. He got only as far as "The Wadia 2000 sounds good, but..." when he was cut off by Dr. Lipshitz.
"Ah, but how do you know what is good?" the professor interjected.
This question opens an entire area of philosophic inquiry. Before getting around to answering the professor's question, allow me to digress.
About 14 years ago I read a book called Zen and the Art of Motorcycle Maintenance (footnote 5) that has had a profound influence on my thinking ever since. After my experience at the conference, and especially Dr. Lipshitz's question, I decided to reread Zen and consider its wisdom in relation to the whole subjective/objective audio question.
Footnote 1: Audio's David Clark, a leading proponent of the idea that all amplifiers sound the same, actually drew a comparison between audiophiles and believers in a flat earth during the conference. So closely is David Clark associated with the "all amplifiers sound the same" school of thought that at the AES reception before the conference began, he was greeted by Lucasfilm's Tomlinson Holman:"Hi Dave all-amplifiers-sound-the-same Clark." This is the same David Clark who, when asked what source material he uses to evaluate speakers, responded: "I use the same things as my test instruments do. Pink noise is very revealing and Techron sine sweeps, though they give you a headache, are very revealing. I've produced a pulse generator. Pulses tell me a lot about the sound of things." (The comparison between audiophiles and astrology was made during a question-and-answer session by a member of the audience who also happened to be the program leader of the audio testing division of Consumer Reports.)
Footnote 2: I should make the distinction between the very important listening tests conducted on loudspeakers, where obvious colorations exist, and the rejection of the listening experience in evaluating electronics.
Footnote 3: Stanley Lipshitz is Professor of Applied Mathematics and Physics at the University of Waterloo in Waterloo, Ontario. He is a past president of the AES and has published many papers of original research in audio. Dr. Lipshitz is one of the most brilliant minds in audio today. I do not intend to disparage him—I have the greatest admiration for his intellect—but I use him in this discussion because his position exemplifies the prevailing objectivist philosophy.
Footnote 4: The following is a quote from Dr. Lipshitz's paper presented at the conference: "If one reads some audio reviewers, one finds a strong negative correlation between measured performance and rated sound quality. This is particularly worrying as it implies one of two things—either many standard measurements, which are supposed to correlate with better sound, in fact do the opposite (which I do not believe), or else the component's aberrations are actually being praised as representing 'better' sound, whereas in actual fact they are errors of performance. This latter is what I see occurring more frequently now, and represents a disservice to the audio industry and the goal for which we are striving. I personally have met a number of well-known audio pioneers who have become very disillusioned at what they see as the triumph of the antithesis of all they have stood for at the hands of charlatans who trample on their ideals." (Presumably, I am a "charlatan" who "tramples" on the ideals of science because I think the Wadia 2000 sounds better than a $200 CD player whose performance, by any measurable, "objective," and rational standard, is far superior to the Wadia's.)
Footnote 5: Zen and the Art of Motorcycle Maintenance: an Inquiry into Values, by Robert M. Pirsig, is published by William Morrow and Company.