A Matter of Dimensions Page 2

If you think that my selection has been biased in favor of the subjectivist viewpoint, and that Heyser must himself have been a crazed "golden ears," neither is the case. If you were an admirer, as I was, of Richard Heyser's loudspeaker reviews in Audio (a full index of them is included in TDS), you will remember that he was one of the few reviewers who attempted a complete technical analysis of the product under test. Read again two of his last reports, on the Quad ELS-63 (the latest version of which is put through its paces by Sam Tellig and myself in this issue) and the Thiel CS3, in the June and November 1985 issues of Audio respectively, for about as thorough a technical exposition as it is possible to find. (His brief comments on the listening in his reviews, however, always gave me the feeling that there was more he wanted to say but, for whatever reason, felt it inappropriate to do so.)

Heyser appeared to believe passionately that the experience induced by inserting a new component into a system could be explained by measurement, hence his invention of TDS. But any one measurement, or even any limited set of measurements, could not describe what the listener would hear. His writings convince this reader that when it comes to assessing the ultimate worth of a component, it is the whole—ie, the listening—that matters, not the sum of any of the routinely measured parts.

The quotations succintly illustrate, I feel, the general point that emerges from this collection of Richard Heyser's writings: that, indeed, the conflict between measurement and listening is due to protagonists for both sides using different frames of reference. There can't be communication unless a common language is agreed upon; this was exactly Stuart Chase's thesis.

Can the subjectivist writer expect to be taken seriously by the engineers responsible for designing the components he is listening to when all he will talk about is the increased ability of a component to present the emotional content within the music? Does he even expect his loyal readers to take his writing seriously when he says something along the lines of "it increases the specificity of image—location of instruments and voices, exact locations of boundaries between sound source and surrounding air—by about 200%..." (footnote 3). (Sounds like a very unique observation to me.)

On the other hand, can the test-bench reviewer seriously expect anyone to make use of his evaluations to make a buying decision when all that his review consists of is an assurance that the component conforms to a set of numbers that are apparently arbitrarily chosen to indicate "goodness"?

"Arbitrarily?" Yet what other word can be used to describe the list of specifications that are quoted everywhere, from catalogs to even the review headings in Stereophile? The last thing any rational reader would expect is that the specifications could "describe" the sound of a component. How, then, can such specifications be expected to yield a presumable "goodness" quotient? And if they are not relevant, then why bother with them at all, apart from that fact that some—loudspeaker sensitivity, preamplifier noise levels, and amplifier output power into a range of loads—can indicate system-matching problems?

The problem is that traditional measurements developed along directions not suggested by their usefulness but by their ability to be performed at all. As Heyser points out in TDS, for example, measurements of acoustic phase were extremely difficult until the advent of modern TDS and FFT techniques. They were thus omitted from the traditional measurement cocktail and considered unimportant. Even the set of measurements we routinely perform for Stereophile's reviews are dictated by what is achievable; we can only attempt to project from them to explain the "why" behind the "what" is heard.

Of course, that is the correct way 'round, in my opinion: Note what is heard, then try to explain. In the early days of sound reproduction, all evaluation was done by listening; test equipment did not exist. As test techniques evolved, each involved a reduction of the complex mix of parameters representing the listening experience to just one that could be measured and plotted against another: voltage or pressure plotted against time to give the traditional frequency response, for example, or Total Harmonic Distortion against output level into a fixed resistive load.

But, I am sure, to those engineers who developed such tests 50 years or more ago, it was still the whole experience that mattered, the measurement being used merely to aid diagnosis than to determine the quality of the experience. Now, however, in the latter years of the 20th century, a whole culture has arisen that insists that the measurements are the experience. If ever I came across a worse case of mistaking the messenger for the message, I have forgotten what it was.

As Heyser repeated again and again in his essays and papers, the reproduction of music is a multidimensional event. Music consists of an instantaneous sound-pressure level which changes according to the logical demands of two things which have no physical reality: the way in which music is structured in time and pitch, and how that structure is ordered by the composer/musician. The real aspects of the concrete framework that supports those two abstractions are, according to Heyser, at least five-dimensional in that not one dimension is causally related to any other (footnote 4).

Footnote 3: John W. Cooledge, The Abso!ute Sound, September/October 1988, p.63, as quoted by Euphonic Technology's Michael Goldfield in this issue's "Manufacturers' Comments."

Footnote 4: Best described in "A View Through Different Windows," reprinted in TDS from Audio, February 1979.