How Hi-Fi are Stereo Discs? Page 2

This is a case where the techniques of recording are being used to create, almost from the ground up, something whose resemblance to the real thing is, frankly, a fabrication; the real thing does not exist to begin with. The "hall" is not a concert stage, the placement of the performers is not like that at an actual performance, and the balances and timbres of the instrumental sounds are almost entirely under the control of the recording engineers, who may or may not have the same ideas about such matters as the conductor or the composer. If the recording director knows his stuff, and really tries to turn out a natural-sounding recording, the multichannel technique can still yield a very natural-sounding recording. Vanguard uses multichannel recording, and the results have often been very musical and startlingly realistic. But as Mr. Fantel's "technician" implied, natural sound does not seem to be of much concern to the average recording director.

Many recording engineers feel that, since music is composed of instrumental (or vocal) sound, the more instrumental sound they can record, the more music they are bringing to the record buyer. (This explains why so many recordings seem to put every instrument in the orchestra under a sonic microscope, revealing every subtle rasp, scrape, and wheeze that is the inevitable byproduct of music making.) These "raw" instrumental sounds serve as the building blocks from which the recording engineer assembles the total musical sound, adding brilliance here, accentuation there, and reverberation all over, to "enhance the intent of the music." Evidently, the conductor is not felt to be capable of doing this, and the listener is believed incapable of perceiving it, for the "enhancement" is rarely subtle. Triangles clang, violins shriek, and trumpets and woodwinds zoom to the foreground for their moment of glory and then recede into the ranks again.

This "re-forming" of the original musical sounds, the better to "project the musical ideas" (to quote RCA's "Dynagroove" press release), is a perfectly valid approach, in principle. But since its purpose is to point out, as it were, every detail of the musical score, with shadings dictated by the "meaning" of the music, it deprives the listener of his right to interpret the music according to his own tastes, and it does not produce what could be considered a sonic replica of any concert-hall sound. In short, it lifts recordings out of the realm of reproductions and into the category of original creations, whose resemblance to natural concert-hall sound is incidental (footnote 3).

Popular music and that vast area of bland noodlings known as background music have their own special recording requirements. Most of these discs are played on the worst imaginable phonographs, they must of necessity have very restricted dynamic range (so they'll always be loud or soft, depending on where the listener sets his volume control), and since much of their desired effect is achieved through electronic "distortions" of various kinds, it is pointless even to consider them in terms of the original. It is equally futile to apply such idealistic standards to stereo showoff records—the ones Mr. Fantel stated, optimistically, that the public is becoming tired of—because these are intended to startle and to entertain, rather than to convey any illusion of reality.

There are, however, certain kinds of musical material that, having been carefully planned to yield a certain sound in live performance, are best reproduced as realistically as possible. Intrinsically fine voices, good jazz performances, and most classical music fall into this category, and here is where high fidelity finds its real—indeed its only—meaning. Yet material like this has suffered just as much as the fun-and-effects stuff at the hands of the compulsive dial twisters.

John G. McKnight (of Ampex Corporation), writing in the Journal of the Audio Engineering Society (footnote 4), said, "Although we might have difficulty in finding anyone who would admit it, it is not uncommon practice in [cutting] disc masters to use a 70Hz high-pass [bass cutoff] filter, except for organ recordings. It is generally found that the elimination of these very low frequencies gives an improvement in overall sound quality, since the low-frequency noises [in and about the studio] are eliminated, and no significant musical content is removed."

This deserves some comment. First, it is a simple matter for the home listener to kill "low-frequency noises" with a rumble filter or certain types of bass control, but if the extreme bottom isn't on the record at all, there is no possible way of recovering it at the listening end. As for the "improvement in overall quality," and the lack of "significant" musical content below 70Hz, these are matters of opinion. There is a considerable amount of program material below this frequency—one third of the string bass's total range lies below 70Hz—and composers who wrote double-bass passages for this low range presumably would not have bothered to do so had they felt the deepest notes were not "significant." This is only a specific objection to bass restriction, though. Of more fundamental interest to us is the fact that many of the recording companies have taken it upon themselves to dictate to the record buyer what fraction of the total original sound he will get when he buys a "high-fidelity" LP recording.

So, where does this leave the "industry-standard" RIAA equalization curve? The record industry's attitude toward this was evidently pretty well summed up by an executive of one major company who we talked to in connection with this article. "Yes," he said, "we do use the RIAA curve to cut our discs, but it's all a big laugh because the master tapes have their frequency response doctored up all over the place."

This is as though someone had crept into your living room in your absence, loosened your preamp's tone control knobs, and re-tightened them in different positions, so that when they said Flat, the response was far from flat. A record's jacket notes never specify what tonal "corrections" were used, and it would be impossible for any home playback system to correct for them anyway. The recording studio's resonant equalizers can produce response-curve shapes that no conventional R/C tone controls can complement, and even if we did use studio equalizers at home, no playback curve could restore the signal to flat response. Different amounts of equalization are used in each individual mike channel, so once the channels are mixed into the final left- and right-hand stereo channels, there is no way of separating them again for individual tonal corrections. And audio perfectionists worry about maintaining preamp equalization to within plus or minus or half a decibel! The recording executive was right; it's all a big laugh, except to those who still believe reproduced music should sound like the real thing.

To the high-fidelity perfectionist, the most infuriating thing about all this is the fact that modern recording and playback equipment could reproduce virtually all the frequency and dynamic range of a live orchestra to within a decibel or so, if given half a chance. A super-disc like this, with !X1dB 30Hz–15kHz response and 50dB dynamic range, would reproduce more cleanly (footnote 5) and more realistically through good, modern playback systems than anything we've ever heard. Yet what we do get for our hi-fi record dollar? Typically, a frequency response of !X6dB from 60Hz to 10kHz, with negligible response below 60Hz and a total dynamic range of less than 20dB! And it's all done intentionally, in the name of better recordings.

The one consistency underlying all this appears to be the prevailing idea that whatever sounds good is high fidelity. Since it is generally conceded that the end result—the sound—is the only criterion by which we can judge the fi a disc, anyone responsible for the sound of a disc feels he has the prerogative, if not the obligation, to do to it whatever he feels is necessary to produce a "good" recording. And this is where all the trouble starts, because "good" does not necessarily mean "high-fidelity."

High-fidelity sound is, literally, highly accurate sound—sound that is very similar to that heard under actual live-performance conditions. Thus, a high-fidelity music recording would be one carrying information which, in playback, is translatable into sounds that are an accurate replica of those that might have been heard in the concert hall.

Fidelity is accuracy, and has little to do with personal preference. "Good" implies a value judgment of liking, as opposed to non-liking. Consequently, any sound that creates a pleasant impression on a listener may be judged by him as "good" sound. The fact that he may not like the actual, live sound of a violin means that he would judge a high-fidelity reproduction of it as being bad, whereas a muffled, low-fi reproduction, erasing the gutty sheen that he dislikes, would be adjudged a "good" sound.

By the same token, a "good" recording may embody other positive virtues that have little to do with its actual sound. A disc that skips grooves on most phonographs, or has inadequate stereo separation, or becomes inaudible some of the time, may be judged a "poor" record, even though the disc itself may have the potential of producing virtually perfect fidelity.

The record manufacturers realized some time ago that, since John Q. Public had learned that hi-fi was desirable, he expected his records to sound good. This posed a dilemma, for the most intrinsically perfect record in the world would not sound as good on the average phono as would a disc with carefully built-in "corrections" to compensate for the phonograph's shortcomings. In fact, our ideal disc might not even track on JQP's console, and it doesn't take a sophisticated listener to know that groove-skipping ain't hi-fi.

Since record making is, first, and last, a business, it is not surprising that the majority of record manufacturers decided to make a few compromises with perfection, in deference to JQP. Low-frequency amplitudes were held to within limits that his pickup could track, stereo spaciousness and separation were juiced up so they'd be audible from his closely spaced loudspeakers, and a few dB of treble boost were added to brighten up the dull top from JQP's console.

Unfortunately, though, once the first of these compromises had been made, the dam was breached. If it was all right to do these things just a little bit, why not do them even more? After all, they did make the records sound more natural on the majority of phonographs, and wasn't it the whole idea of high fidelity to produce natural sound? The answer, clearly, was "Yes" on both counts, but this was the turning point in high fidelity: the realization that carefully calculated deviations from the intrinsic fidelity of a recording could yield truer sound, under average conditions. The recording itself was no longer considered the end result of the record maker. Instead, it succumbed to Organization Thinking and became part of a system, whose ultimate objective was to reproduce music—not the concert-hall sound of music, but the "totality," the meaning, of the music.

Footnote 3: It is rumored that the recording director of one major record company was fired for permitting a piano recording to sound like a real piano, but this has not been confirmed.—J. Gordon Holt

Footnote 4: JAES April 1962, p.107.—J. Gordon Holt

Footnote 5: The "overcutting" that causes distortion from most modern discs is mainly the effect of excessive treble boost, which causes groove accelerations that no stylus has low enough mass to follow.—J. Gordon Holt