The Truth About High End
I was irked by Alan Lofft's article on perfectionist audio ("High-End Stereo: Sense and Nonsense," October 1982), but probably not for the same reasons some other undergrounders will be.
Actually, I thought the article was excellent, and do not take issue with most of what Mr. Lofft said. It was, sadly, all too true. What I do resent is Stereophile magazine being tarred with the same brush which blackened the names of some of the more deserving members of our kind.
Among the subscriber-supported "undergrounders," Stereophile is almost alone in having resisted the urge (compulsion?) to be carried away by a preoccupation with trivia. We do not rhapsodize over the smallest increment of improved definition, imaging, transient response, and subsonics, and for not so doing we have earned the scorn of most of the other "undergrounders."
Most of the minutiae by which the lunatic fringe compares competing products are in fact less significant than differences we often observe from one sample to another of a given product, which is why we are less inclined to pick nits than are some other undergrounders. We have always contended that accuracy through the middle range, encompassing 90% of the music, is the first criterion for a music-reproducing system, and that all other considerations are secondary. If a system can't cut the midrange mustard, we don't much care how well it does anything else.
I am still not convinced, one way or the other, of the audio purist's claim that amplifiers which measure the same can sound different, because the measurements referred to are often too rudimentary to reveal many of the measurable differences betweeen components. But observation has convinced me that amplifiers which measure identically on simple THD and frequency-response tests do indeed sound different. When we describe the ways in which two competing amplifiers differ in sound, we feel we are performing a much more valuable service for our readers (who are advised to mate components with ones that complement their sound) than we could by merely quoting test results and saying they were "splendid."
Throughout our entire 20 years of publication, we have endeavored to maintain as high a level of technical accuracy as we can, and to eschew any taint of mysticism, occultism or spiritualism. And although we feel that Stereo Review is being too reactionary by denying the validity of subjective testing, we also feel that much of what is reported by the underground press to be audible and significant is in fact nothing more than self-deception (autohype). We are ashamed of what subjective testing has become and what it has done to audio testing, and our frequent potshots at the lunatic fringe is a major reason why our circulation is as small as it is and as sane as it is. But we still insist that subjective testing has its place in audio reporting. Yes, it has yielded some bizarre flights of obfuscatory fancy, but it is possible to describe reproduced sound in terms understandable to all.
We were the first magazine to attempt to do this, because we felt it made more sense to listen to and try to describe a component's sound than to try to interpret its measurements in terms of their probable effect on the sound.
Thus, a loudspeaker might be described as murky, aggressive, retiring, sizzly or squawky, and 90% of the readers will be able to visual ize what is meant. Middle range colorations can often be likened to human vowel sounds—oo, ah, ih—which objective measurements will show to have the same mouth-cavity resonant frequencies as the loudspeaker peaks which cause the colorations. As imperfect as subjectivity may be, we feel it is more meaningful to a prospective buyer than any number of measurements.
Actually, distortion measurements can tell us much about a component's sound, if carried beyond the simplest total harmonic distortion presentation. Second-harmonic content determines how "alive" and close-up the sound seems, while high-order odd harmonics (5th, 7th, 9th) add harshness and artificial "detail" to the high end. It is the relative distribution of harmonic distortion components which makes tubed components more agreeable to many listeners than solid- state components, despite the higher total harmonic content of the tubed equipment. (THD does not show that tubes produce little high-order distortion, transistors a lot more of it.)
These are the kinds of objectively demonstrable differences Stereophile is concerned with, and are why our equipment reports are relevant to music listeners who relate more to live music sound than to objective measurements. They are also why we resent being classified as a crackpot cultist magazine, along with the ones which obviously are.