NO NOISE! That's what we want today, speaking negatively. Noise is bugging
us. We're spoiled by the digital system, which in both tape and disc brings
us extremely close to TOTAL SILENCE in the absence of signal. That's what
we want! But we can't get it--yet. At least not on discs. We are still living
with the LP in our own consumer systems, as we are living with analog tape.
We will for many a year, maybe.
Ten years? That's Philips' casual estimate. Ten years of LP noise? How can we ever stand it? Well, dbx says it has the answer.
(And we can use it, as never before.) The dbx "noise less" encoded disc! It's here--or rather, it is back and better. This system was launched once before, quite awhile ago, and nobody noticed very much. Wrong moment. This time is better and we will probably notice--we almost have to. For if you can't yet get noise less audio via digital, the very next best way just has to be via electronic-type noise reduction. You reduce the noise without affecting the signal itself.
Or do you affect it? That's the vital question.
Well, this new dbx disc is astonishing. I can say that because I've heard it. Here you have no more than a plain old LP (specially cut and pressed) with the same dbx II circuit for coding that you already know from recent dbx consumer products (and compatible with them)--and it sounds, really, truly, almost like digital. SILENCE! I'm wary of the word "total" in respect to any kind of silence--let's call this audible silence, palpable silence, the kind you almost feel, that makes you think the system is switched off.
You hear--nothing. Except signal, when & if. We get this sort of silence out of the best component and pro equipment (and leave it turned on for days by mistake). We have heard it, unbelievably, in demo after demo of the new prototype digital discs and the digital tape systems. All digital audio has it because silence is inherent in the system itself, which does not read noise. (Well, hardly ever.) But do we get it in our new LP superdiscs, d-to-d or digital-derived? Nope. Not even on the best and/or most expensive. It may have been there in the original signal but the good old phono pickup and the LP surface convert that silence very efficiently into, at best, low noise.
There's only one way to achieve minimal noise on standard LP and that's via a scarce commodity called "TLC." Tender Loving Care. But folks, we really can't depend on TLC for our mini-noise! It's inhuman. It isn't fair. If the natural end state for milk is to be sour, then the natural tendency for the LP record is to be noisy. More Murphy subsets. We aren't in the dairy biz and we'll never, NEVER get predictably quiet LP discs via any sort of regulation, supervision or inspection, or even with TLC. Well, again, hardly ever. But that's not enough.
Surface Silence
So let's examine the dbx offering and see what it might do to help--short of digital. Dbx is ready. The company is now a subsidiary of BSR (USA) Ltd., and it has a dynamic new bossman, Jerome E. Ruzicka, who is far from noise-free. He recently was with the Bose Corp. and now will talk your ear off with enthusiasm for the dbx "noiseless" disc, which is very much his baby. He was so excited at his press conference that he fairly sputtered, and a good many of us went right along with him. When, after his speech, that vaunted SILENCE finally hit our ears, it was indeed extraordinary. This was definitely the first time had heard an LP pressing, played through a standard pickup, that actually compared for the ear with digital.
Especially in com parison with the same recording in an uncoded pressing.
Let's put it into figures, unofficial. I quote Bert Whyte and others to the effect that the limit in S/N for the LP (uncoded) is around 60 dB. Nor mal, smooth LP surfaces average very roughly around 45 dB be low a hypothetical center point, the standard 0 dB level, if I read it correctly. But these last numbers, even with engineering precision, are quite arbitrary because of the huge plus-or-minus factor. In practice, I would suggest, what with the usual swishes, rumbles, ticks, pops, mini-explosions (says my suffering ear), a better everyday average might be -30 dB with occasional violent peaks (say no more!) up to +15.
"Noise louder than signal?" Of course.
You hear it every day.
If I am right, the coded dbx LP disc plays back with an S/N that gets down into the -70 dB range. Extraordinary if true, and it surely accounts for what my ears did not hear. This does indeed compare favorably with digital specs.
Philips says the S/N figure for its compact digital disc is -85 dB, and digital tape goes down astronomically beyond -90. No-not quite totally silent.
But might as well be. If you stick your ear right into the speaker on silent pas--sages (at your own risk), you just might hear something but you probably wouldn't. So in terms of achieved S/N via the LP record, dbx provides us with exactly what we want in the most negative abundance. This might very well hold us until the digital disc gets around, maybe within that 10 years or so. That could be important.
Pay the Piper?
If dbx is this good, there must be a catch, a price to pay, whatever its value. That price is in the compression-expansion cycle, a treatment of the whole signal through some pretty fancy circuit work. And right there I see a lot of leery faces and eyes looking askance. That sort of thing? "'Jo-no" We hate noise but we want our audio pure and lovely, untouched by human hands. Or extra electronic circuits, thank you. We will even listen to all that NOISE (gritting our teeth) in order to preserve signal pristinity." Or will we? It all depends.
The dbx system uses the long-familiar two-step type of noise reduction, via a pair of mirror-image circuits that code the signal before recording and then decode it in the playing, back to where it started, at the same time taking any subsequently added noise (uncoded)--say LP surfaces or tape hiss--right down along with it to far below "normal." Ideal result: Signal unchanged, noise reduced. It's a nice concept once you've got it straight in the mind (which isn't always easy, I've found). And there are many ways to design the double circuitry. The idea is, of course, best known today in the popular "Dolbyized" cassette as well as in Dolby professional equipment.
Indeed the two systems, Dolby and dbx, fill very similar rival niches in audio usage, both professional and con sumer. As we know, dbx even de signed a plug-in replacement for Dolby A pro equipment-they are that close in type of function. Either one does its noise reducing very well when the two mirror-image circuits work as intended. But, of course, one system will not decode the other. There are profound differences.
Those differences are technical and do not belong here in detail. Yet the two systems' basic principles are important if we are to evaluate the "fi" of any sort of noise-reduced LP disc.
History into Future
There could be a Dolby disc--there may be?--parallel to the familiar Dolby cassette. That would involve "Dolbyizing" the disc itself and playing it back through conventional Dolby B decode. Now there is, in fact, a dbx disc (let's not speak of "dbx-ing") and it works the same. The record is dbx II encoded in the production pro cess and decoded in the playback through reverse dbx circuitry, added to your system in one form or another ac cording to need. As simple as the cassette. Right here the Dolby/dbx similarity ends. And the differences begin.
I think I can see why there could be a dbx disc in our future and not a Dolby disc, though both are possible. The most profound difference between these systems is historical, which ex plains a lot.
Dbx is newer, goes further (-30 dB to Dolby's approximate -10 dB) and takes bigger risks, at least in theory, by boldly treating the whole signal. Dbx depends on the incredible accuracy of modern electronic design. Dolby, de signed in an earlier time, most ingeniously avoids that risk, and that is why Dolby is still with us. A very neat and interesting comparison.
Dolby's circuit came out of the early 1960s in its original launching, which is a very long while ago in audio terms.
To my mind Ray Dolby's thinking at that time had the real quality of genius, safety-pin type. It was simple, unerring, and to the point. Nobody else thought of it. Dolby A, the original, and Dolby B, for consumers, each remain the same today in their operating parameters as they always have been--and interchangeable right down the line, old and new. The Dolby system is hellishly clever--it "solved" the problem of signal purity by simply leaving the signal untouched. All but a tiny fraction in the very low-level area. The ear's masking ability, ignoring noise in the presence of louder signal, had been used before but this was an ingenious application. For most of the signal, Dolby was a non-circuit. And it gave us that 10 dB. Not surprising that the pros went along in force! The idea still holds-don't tamper with my signal.
Not any more than you have to. So Dolby lives on.
But dbx can give us much more noise reduction via the two-stage approach-IF we accept treatment of the whole signal, complete. That's the rub. Heads still shake at the thought.
Signal purity! We like it better than ever. And dbx boldly grabs your entire audio signal, not only compresses it 2:1 across the board, 100-dB wide in the dynamics, but even adds pre-emphasis up at the top and de-emphasis in the playback. That's tampering, if you feel that way. But look at the results. Not merely all that noise reduction. A grandly widened range in the replaying, a full 100 dB instead of the safe average 50 dB or so of the usual LP.
Dbx puts no more than 50 dB on the record, cuts narrower grooves closer together (thanks to much lower maxi mum excursion in the lows), and over all cuts out nearer the edge for cleaner inner-groove sound in the same playing time. That makes for new flexibility. The mechanical strains on the LP-groove system itself are neatly transferred to the more capable electronic area.
Dbx even goes back to the master tapes (or wide-dynamics d-to-d cut ting) to find its 100 dB of dynamic range, often reduced for the standard LP cutting. Remastering. It's a whole new version of the recording, maximized to take advantage of every favor able break provided by this code-de code procedure. Pretty good, eh? Solid thinking. But how about the signal quality? I'll give no specs, but the argument is straightforward. Since the mid-Six ties there have been enormous advances in audio electronics. We have learned to "treat" our audio signals, in whatever way, in many ways, with in credible precision and lack of distortion even in systems of great complexity. Our signals go in clean, travel through a million or so modifiers--and come out clean. Just compare amplifier specs of 20 years back and today.
Look inside our audio gadgetry, al ready in wide hi-fi use. Study Peter Scheiber. It's a new electronic world.
And thanks to transistors and then circuit boards and ICs, we can build enormously sophisticated circuits that still manage to be practical and reasonably cheap to produce. We know how to make them work not only with low distortion but with uncanny precision. Enough said.
New-Era Cleanliness
So now, maybe, we can indeed grab the whole of a signal, put it through compression and re-expansion and more, and come out clean? Maybe not quite 100 percent. But close. Without any specs at all, that is the argument.
Dbx, you understand, comes straight out of this new era and has had the time to become highly knowledgeable in noise-related compression/expansion techniques. If I am right, then, the dbx II circuitry is a very sophisticated product of recent advanced thinking.
It has to be. I think maybe I'd trust my best audio to it. Would you? You'll have to decide.
Of course, there are a few minor ancillary problems before this dbx disc can get through to us in the millions and save our noiseless souls. No new recording system will succeed without a very substantial software immediately available. By itself, the system is nothing. We know what Philips has in store for us in its digital baby disc-the best of Europe and the world. Do you remember the stacks of ready LPs that came from Columbia right at the beginning? What can a smallish outfit like dbx, not a record company at all, do on its own? Well, quite a lot, it seems. Bossman Ruzicka understands the crucial aspect of available software and he is going to strenuous lengths. Dbx isn't about to buy out RCA or Columbia, but they're not going to toss you only a few tiny labels, take it or leave it.
Instead, they are setting up an emergency ad hoc processing and distribution department at dbx that will do everything, almost anything, that might help a record label to get started and avoid the ominous problem of wasteful double inventory--a separate disc for each type, dbx and standard.
There will be no double inventory.
dbx takes on the coded version, does the vital remastering from early-generation originals; dbx in effect buys the records from the label and distributes them. A big strain on the resources, but that seems to be the way they'll do it, for the present. It is a brilliant, brave, and positive approach. It may well work. Anything, anything that gets this product out on the market in quantity is good.
Needless to say, the decoding end of the process is right in dbx's alley.
Most of the recent dbx components al ready can decode the new record, and there is a new Model 21, just for disc.
Other decoding configurations, built into phono equipment, should follow according to need as with the Dolby cassette.
Now don't laugh. As of the first press conference some months ago, dbx had exactly 19 releases on its list of coded LPs and a few more under he hat. Not exactly a Schwann catalogful. But that very day they announced that Vanguard was on board and they have Vox-Turnabout, two important labels for a good start. Also smaller labels, plus their own audiophile specialty line, beginning with the work of Mark Levinson. The big companies are another story-they will act reluctant, you can bet. Even so, we should look carefully and listen with care to the dbx sound. It just might rescue us from LP noise for some years to come. Lovely idea.
(Adapted from: Audio magazine, Dec. 1979; Edward Tatnall Canby )
= = = =