Deeper Meanings Letters part 6

What is good?
Editor: Robert Harley's article relating Zen and the Art of Motorcycle Maintenance to the objectivist/subjectivist controversy was very appropriate. It seems to me, though, that he copped out by not answering the question, "What is good?" in more concrete terms.

One useful definition of "good" in audio equipment would be the ability of a reproduction chain to replicate the sound of the original musical event at the position of the microphones. Only the people who made the recording know what that might sound like, so this definition is of limited value to most of us, but it is a valid definition even though the degree to which a given piece of equipment approaches this ideal must be determined by listening.

Another definition of "good" which I find more useful in evaluating my equipment, and judging different tweaks and modifications that I perform, can be related to two perceivable sonic qualities: resolution and timbre. Some equipment will cause reproduced music to be less harsh, less fatiguing, more pleasant to listen to. All other things being equal, I'd rather listen to equipment that gives instruments pure, rounded tones similar to the live sound of the instruments as opposed to equipment that makes high-frequency overtones stand out like a sore thumb. This is "good." Although the differences between two good preamps may be subtle, they are just as perceivable as the differences between a Mesa Boogie tube guitar amp and a Peavey transistor amp, at least to me.

Resolution, it seems to me, is an objective quality. If I listen to one preamp, maybe I know there are 15 instruments in the mix, but I can hear only eight of them clearly. Then, I may switch to another preamp, and suddenly I can hear not only the eight instruments heard before, but also the other seven instruments, and now I'm able to pick out the melodic lines that each of those previously unheard instruments is playing. I switch back to the first preamp to find I can no longer hear those instruments or their melodic lines—they are obscured. Even though I determined this by listening, it is just as objective as evaluating the ability of two camera lenses to focus by looking at pictures taken with those lenses.

Would objectivists claim that human beings are unable to focus a camera lens (experience) rather than by an auto-focus mechanism (objective measurement)? If audio measurements don't correlate with perceived sonic focus, we're using the wrong measurements. I wouldn't accept out-of-focus photographs simply because some mechanical device told me they were correctly focused. I would say the device is in error and use my own eyes. By the same token, I won't accept out-of-focus sound just because somebody says the THD of a device is unmeasurable.

So, Lipshitz's question, "How do you know what is good?" is like the question, "How do you know that photograph is in focus?" I can see whether a photo is focused by looking at it. I can hear if audio equipment is good by hearing whether it has natural timbre or not, and whether it has good resolution or not (to cite two characteristics of audio quality).

He who has an ear to hear, let him hear! (footnote 3)—Eric Anderson, Ankeny, IA

Reality & analysis
Editor: It is reliably related that Ernest Lawrence (Nobel laureate for the invention of the cyclotron) kept a bottle of a particular brand of cheap scotch whisky in his desk drawer and offered it to visitors, insisting that it was just as good as the best scotch; an assertion that he could prove on the basis that chemical analysis had shown the two to be identical. This idiosyncrasy was tolerated by his colleagues who characterized the cheap scotch as rot-gut. They were amused that he insisted on accepting obviously flawed measurements over the evidence of his senses.

I relate this somewhat apocryphal story as introduction to this letter in response to Robert Harley's "As We See It" article in July. Despite the apparent support it gives to Harley's thesis, those involved with Lawrence clearly believed only that the available measurements were too crude to detect the obvious differences, not that such measurements were inherently impossible. Similarly, today a skilled wine-taster can detect subtle differences well beyond the ability of virtually all available analytical instrumentation.

In two other sensory areas, odor and color, human senses have prevailed until recently. In perfumery the most skilled practitioners still exceed the best analytical capability (but not by much). In color (paints, dyes, pigments, etc.) the instruments have finally been improved to the point that their ability to detect subtleties has exceeded that of human visual ability; it is now possible to achieve color reproducibility in dyes and pigments that was once impossible.

As a scientist I am an objectivist, and so I enter this debate with trepidation, expecting slings and arrows for rebutting Harley. Despite his assertion that he is not attacking the scientific method, in fact he does so in no uncertain terms, as the following quotation makes obvious.

"There are phenomena in audio electronics that defy measurement and are revealed only by listening. To claim that every phenomenon that affects sound quality is measurable and quantifiable is arrogant conceit."

This statement is in direct contradiction to the scientist's notion of physical reality. If this statement is taken at face value, it states that reality contains phenomena that will forever be beyond the ken of the scientist's craft. That is, there are real physical phenomena that can never be measured, quantified, modeled, or predicted. Hence they must be classified as magic.

Harley's example of the high-end audio designer also denies the validity of the engineering process for designing superior products and substitutes for it an arcane procedure that is more akin to sorcery than electronics (footnote 4). He claims that the specs are essentially irrelevant and that only listening will reveal the value of a design. There are a myriad of different types of capacitors, resistors, wire, cable, etc., all of which are said to affect the musical quality of an amplifier. To make the correct choice by cut-and-try methods alone, even with all the present and past designers working feverishly to judge by listening tests all the possible permutations in order to find the few superior possibilities, would be utterly impractical. Thus, apparently, the high-end designer uses some method that is beyond the ability of science to fathom in order to make these critical choices (footnote 5).

But if this a real and valid capability, should it not be applied to the design of nuclear power reactors, the space shuttle, and commercial aircraft—they could all benefit from improvements in the equivalent of the "musical experience." The last such design tour de force was done by Kelly Johnson, who designed the SR-71 with a slide-rule. He admits that he would have done better with the aid of a Cray computer.

The final noteworthy point concerns Harley's criticism of the blind test. Blind tests have often been shown to be invalid, but invariably because the controls were not strict enough and bias entered into the results. There may be procedural difficulties in carrying out a blind test for audio equipment evaluation, such as listening to too short a passage, but these can be corrected. To assert that blind testing is inherently flawed is to invalidate the enormous effort that has led to almost all medical advances of the past 100 years. In that period the double-blind test has become the accepted standard for testing the efficacy of a new therapeutic drug or procedure. It was devised in order to assure that new drugs and procedures were effective. Where it has not been applied, ineffective, unnecessary, or ill-advised treatment has occurred. Radical mastectomy is a good example. Numerous drugs for cancer treatment have been promoted without the benefit of adequate blind testing. They resulted in quite a few tragedies before being exposed as frauds (deliberate or not).

Fortunately, audio equipment is merely pocketbook-threatening. But it should be subject to the same rigorous, unbiased evaluation that I expect for my washing machine and television set. In the case of televisions, I am under the impression that vision is considerably more complex than hearing—the volume of brain devoted to the former is considerably larger than that devoted to the latter. Yet I am unaware of a similar argument regarding the existence of undefinable phenomena that affect the quality of the visual experience in movies or television. Is this situation peculiar to audio?

I do not know how to carry out a double-blind test of audio equipment in any practical way, but if it were possible, that is what should be done. (In a double-blind test the individuals involved in the test would not know whether or not they were evaluating equipment differences; some, the control group, might listen only to live music.) Harley argues that blind tests obscure differences—truly a non sequitur. Blind tests are a basic tool in measuring subtle differences in phenomena not easily subjected to instrumental evaluation. Blind tasting is the accepted modality in wine evaluation, and is capable of defining very subtle differences that are well correlated among tasters. It works well for odor and in other areas.

Harley speculates that (some, the objectivist) scientists and engineers perceive music differently than he and others do. Perhaps, but it seems unlikely. I do not know a significant percentage of all scientists and engineers, but of those whom I have known reasonably well (probably less than a hundred), nearly all have had a well-developed appreciation of music, often have played an instrument, frequently read musical scores fluently, and occasionally were regular members of amateur performing groups. Furthermore, it should be noted that there are several mathematicians among them. Mathematicians are right-brain dominant, like most artists and musicians, and routinely function with both halves of the brain active. They experience Bach in a musical and analytical fashion that I find astonishing, but they apparently cannot hear the differences that Harley insists are present.

Neither can the musicians with whom I have discussed the problem. My brother-in-law is head of the music department of Oklahoma University and our visits often revolve around music live and recorded, frequently in the company of other members of the faculty and graduate students. I must admit, however, that their interest in this controversy is negligible, and they are disinterested in spending time resolving it to my satisfaction.

Footnote 3: I can't believe people are still arguing about these things.

Footnote 4: This does not mean that there is not plenty of room for intuition based on experience in electronic design. My own bible for electronics design is The Art of Electronics, by Horowitz and Hill.

Footnote 5: How does a designer go about creating a component with inferior technical specifications and superior musical sonics such as are claimed for the Wadia 2000? Is the skill teachable? Why do other components with equally inferior specs sound inferior? Are there two completely different processes going on: one in which objectivist scientists and engineers create and develop the technology, and a second wherein individuals with unique and undefinable skills apply it in ways that are incomprehensible to the developers?