Audio, Etc. (Jun. 1972)

by Edward Tatnall Canby

YES, WE STILL HAVE, or are about to have, two opposing systems for the quadraphonic disc, promoted by you-know-whom, the same old familiar adversaries. It would seem to be a nasty deadlock and a fight to the finish. But it isn't, not by a long shot. We may get an Ultra Disc yet, the Ultimate. Things are moving fast, the corporate blocks unblocking, even with two systems, and other systems still threatening. New ideas keep coming up for the sort of compromising that would allow the big fellows to wriggle out of their present positions, faces more or less saved, images intact. A big order, I know, and you'd hardly think it possible right now. But where there is smoke there is fire, and where reasoned ways out of impasses are made visible and well publicized (we in the press try to help), then sooner or later the results will show.

First, if things continue as is and the RCA discrete disc, some of its thorny incidental problems well solved, appears as scheduled by RCA early this spring, then many of us will go along with the two incompatible disc systems--maybe even 2 1/2. Provided, of course, that we get the joint decoding componentry to play both kinds with ease on our home systems. (The dual decoder, as I suggested in April, preferably with automatic switching: just put on either type of disc and let 'er go.) If a supersonic signal is present (as in stereo FM) the thing switches itself to discrete RCA. If no supersonics, then it plays matrix. I'll buy it. So will you. That is, once again and with emphasis, assuming there is a discrete disc to buy and to play, that the perplexing and very real problems the discrete system brings along with it are in fact worked out, as RCA hopes to work them out. The discrete disc is good, if we take it at its best-as we've already heard it in impressive demonstrations. I'll admit that I do not see how some of its disadvantages can be met. But give the boys time. Let's assume, until we have, a fair trial, that the system 'will go ahead. We will eagerly accept it on a working basis to compare for ourselves with the matrix disc.

But things may go further than you might suppose in the direction of compatibility between these systems. They could go, that is, if the corporations were willing. So, in anticipation and hope, I am projecting here towards the ultimate quadraphonic disc, half with a laugh, half seriously. I've even named it, as you can see. It has already been described in detail and in technical language (though not with my name) before representatives of the relevant companies; it is now public property among engineers, though for the benefit of others of our readers I'll act as if it were a brand-new idea. My fingers are itching to get hold of one of these Ultradiscs! But they haven't been manufactured yet, nor perhaps even put into active experiment. Too bad. Let me lead you, in all logic, straight to the concept of this disc, so you'll see how nicely it might work.

The discrete system adds a new band of supersonic information above the standard audible bands, and FM-modulates a signal into each, cut into each side of the record groove in microscopically tiny undulations. You can hear the high twittering noise if you slow down a discrete disc by hand. The four channels aren't actually discrete--like the FM multiplex broadcast channels, they are matrixed together, share and share, each one partly in the supersonic, partly in the audible signal. Decode the two supersonic signals, de-matrix all four, and as in radio you come out with what you put in. Hence, discrete. When it works it is very accurate.

The matrix systems are variably less accurate, less faithful to the ingoing four signals. And thereby they are subject to all sorts of mainly aesthetic "interpretation," variations in the chosen parameters of playback, according to taste. A complex kettle of fish. It sounds bad, and plenty of people have jumped in to make it sound even worse on this basis. But the matrix system has an absolutely enormous advantage in that it uses only the present standard stereo signal and is indistinguishably playable--or broadcastable--on all present equipment without reservation. Moreover, with certain additions in the way of logic circuitry-more of this in a moment--the latest matrix product is beginning to show real accuracy in playback, approaching that of the ideal discrete arrangement to a practical degree.

Meanwhile, as the news reports say, we should look further at the sources of all these quadraphonic signals. The fact is that they are almost never themselves discrete--that is, wholly different. As in standard stereo, they already share between them a large proportion of similar or identical information. Precise decoding, totally discrete, is aesthetically less important than it might seem for valid playback effect. You can trade one four-way mixture for another mix-down that sounds maybe just as good in the listening. And who hears the "original?" Not you and me.

More indicative of practical thinking behind four-channel sound is the fact that virtually all the quadraphonic material so far published, in any format, is derived from originals not intended for quadraphonic use since much of it is on three tracks or even two! This applies, impartially, to most or all of the present RCA Q8 catalogue and most or all of the Columbia SQ list. But--oh no--that does not invalidate present quadraphonic offerings. If anything, the usefulness of four-channel sound is strengthened by this demonstration of what it can do for all sorts of recordings on any plural number of tracks. (Mono is a different story.) Most quadraphonic recordings are going to be of this type for awhile. It is good to know that they are so effective and to understand that the basic recordings are enhanced right down the line by the new type of home reproduction, whatever the playback system. More of the available information is put to use via four channels than is possible via two channels. That's the story and it's a good one.

This is why the various matrix decoders, though far from identical, are able to turn out such good four-channel sound from standard stereo discs--variably but interestingly. It's why both discrete and matrix discs, sending out more specifically separate four-way information, further improve the quadraphonic effect, as a picture might be given heightened contrast and color, sharper definition. It's all on the plus side, you see. Variable enhancement.

But it is important to keep your chosen four signals as discrete as possible from master tape through to home playback. Too much "play," like a worn steering gear or a loose drive shaft, blurs the accuracy of a control system and hence its predictability. On this point, obviously, any system that gives tight control is desirable, whereas the variable, unpredictable effects of assorted matrixing can give our recording engineers hideous headaches.

So without a doubt the discrete disc from RCA would win hands down right now, if it weren't for the aforementioned problems that RCA is now out to solve, as far as possible, before its major launch of the new discs. But these unresolved qualifications, the need for a new-type pickup and stylus, the problem of wear, a presently shortened playing time and lower volume level, the very great difficulty of cutting (and playing) grooves from 30 to 60 KHz, the incompatibility with FM broadcast, are the sort that give the matrix disc a relative boost--for it has none of these problems.

A much more serious competition from the matrix camp, however, is provided by that crucial tightener-up of matrix accuracy, the logic circuit. The key to the matrix disc, without the slightest doubt, is in the perfection of these ingenious circuits, which immensely increase the accuracy, the faithfulness of the decoding. The logic, you might say, picks up hints from its matrix and converts them into certainties. Fantastic.

Without logic circuits, matrix-decoded sound is, shall I say, pleasantly diffuse. It can be lovely, but it is imprecise and unstable. Move a foot or two or change the volume levels, and everything moves. Somehow I keep thinking of a mild inebriation--that's how unassisted matrix circuits work. "Composer" circuits, indeed! An excellent name for them. By themselves, even when matched to their own discs, they tend to treat the original four channels with a sort of drunken fervor. The effects are delightfully there, but they're awfully imaginative. Speaking more soberly--a left rear signal, say, is indeed preponderantly in the left rear speaker but it is also spuriously much too evident in other speakers, if at a lower volume. (What if you are near one of them, or move a bit?) No matrix is in itself a precision decoder, so far as I know. Just a good general indicator.

But keep in mind that few quadraphonic signals are ever one-channel, intended for one speaker only. Like the stereo signal, most are intended to be heard in all the speakers to some degree. Discrete reproduction merely duplicates the original balance and phasing as intended. If things are off a bit, no great aesthetic harm is done. The matrix without logic is very good to hear and do not doubt it. It just isn't precise. With logic added, we have something else again.

Logic circuitry is new and tricky, at least in this field, and we do not yet have final answers in commercial form. Manufacture of matrix units in present circuit-board constructions is obviously well ahead and practical at no great expense. Lafayette, for example, splurges with not only an SQ decoder in its new moderate-cost componentry but also a different "composer" matrix and is still able to add switching for a four-input position and a two-sided hookup that joins both speakers on each side for still another kind of listening. But to add a logic circuit to these first-generation models would have been too costly and/or not yet satisfactory. (Sony's SQ has a very good partial logic.) Not, at least, in conventional circuit board form. So mostly we do not yet have logic in our decoders-and we need it, in all equipment, if the matrix disc is finally to get off the ground.

Too expensive and complex? Just wait. Wait awhile, until the logic chip makes its appearance. That'll be almost any day. I've held a small box of the first production models of such a unit in the palm of my hand. Tiny things the size of the end of my finger, the chip itself almost invisible among the surrounding contacts. When units like these, incorporating the usual dozens of semi-conductor elements, get into production to go with the basic matrix, the matrix-type disc is going to blossom into its own. Soon. About the same time, in fact, that the discrete disc is slated to hit the market in force. So they tell us. Next autumn, they all keep saying.

The battle will quicken then, and there'll be motion at last! Fair fight coming up, out in the open where we can all join in with our own economic clout; and may both systems win hands down. Perfected discrete disc vs. matrix with-logic. That's how they tell it.

What about the Ultra Disc? Well, since the secret (if it ever was one) is already out, I'll let you in on it. Came as a foxy suggestion from the Columbia camp, intended for propaganda leverage. But, .by gum, it might work! So simple.

The idea is, why doesn't RCA modulate its discrete discs with an SQ signal, so they'll play both ways? (Or it could be any other matrix.) That'd be two discs in one, two systems in a single groove. Superb! You see-it is quite possible. You'd play the thing with an SQ decoder and the matrix would decode the lower band into four channels, ignoring the supersonics. Play it on an RCA discrete system and the SQ would be ignored, but the RCA decoder would properly combine the super-tweet-tweets with the lower audio signals for a different set of four channels. Wow! A built-in AB! I dig.

Alternatively, Columbia might produce the complementary opposite, an SQ disc with an RCA supersonic overlay. A + B = B + A. 'Course there are a few technical problems, but no matter. The idea is so sharp that it just might persuade these two giant corporations to ease off their present high horses a bit-if we make enough hilarious noise. Actually, since RCA has four signal areas in its two bandwidths its disc is open to any four-way solution of the basic quadraphonic concept. And so, if it wants to, RCA can technically swallow up all the other systems ever devised, including matrix. (Well, maybe.) But if this company were merely to follow this one suggestion and, quietly, unobtrusively, add SQ signals to its discrete discs, you can guess where we'd be in no time flat. One universal quadraphonic disc--no less than my Ultradisc itself. What else? All we'd need would be a better name. Maybe we'll get it yet, if you'll all keep laughing.

(Audio magazine, Jun. 1972; Edward Tatnall Canby)

= = = =