Audio, Etc. (Jan. 1978)

Home | Audio Magazine | Stereo Review magazine | Good Sound | Troubleshooting



I turn to the following expression with considerable compansion. And if you find those words odd, they merely reflect the gist of my thought which is, can we compand or use our home audio system to good advantage as per a burgeoning of recent new equipment, or will things be OK if we leave our music as it is, just as it comes from the audio package-tape, disc, or broadcast? The answer is yes ... both ways. Some people would put things the other way around: should we compress and expand; but it's all the same to me since expansion is but the mirror twin of compression, two processes which, when all goes well, are one and add up to zero. Which is exactly what the whole thing is about.

The villain of this compransion, is the left-handed twin (gauche, sinister), compression. Compression of signal is a necessary evil in many areas of audio signal processing to make things fit into the audio frame.

Nobody does it for fun & games (though they'll do it for background). On the right hand (dexterous, adroit, righteous), we have expansion, the saving grace, the redeeming feature and, if I many say so, also inclined to take off on its own, minus twin. It's fun to expand.

When it comes to transmitting music through the interface between acoustic and electronic there is no room for anything but careful compromise, which means compression, limiting, and other negative-type circuitry. Various equalizations are part of the picture, but compression is the main thing, at least as of my present concern, and it isn't simple.

Even live orchestras, I have to remind you, must sometimes acoustically compress their own dynamic range, repress their natural expression, even before the microphones get involved; that's one way to do the necessary job. And a nice sub-question is, should we expand them electronically in our homes, to make them play the way they might have played, if--? Well, I can tell you right now that that particular expansion will never work. The only way to make musicians sound as if they are playing LOUDER is to have them play louder ... overtones, physical effort and all that. You cannot make a soft trumpet sound like a loud one by turning up the volume.

But if we merely bring them back to normal, after electronic compression? That is another story and it should work just fine--if we can do it. That's where the dynamic range expander comes in. A mirror-type circuit, its expansion of the volume range matching the compression built into the audio signal. If it works we should have zero, the same thing the players played.

I recently read with pleasure our Equipment Profile on the dbx Model 128 (Audio, Nov. 1977, pg. 106), a piece of equipment which is highly involved in this sort of mirror operation. If there is anybody in the biz who knows the ins and outs of compransion, it must be dbx, and this model is doubly involved, once for the traditional dbx two-stage noise reduction (via mirror compression/expansion circuits) and again via a range expander, hopefully to compensate for the compressions that undeniably are there in the long chain of electronics that leads from living music to the living room. Compransion is a versatile principle with many uses.

Now it happens that I had another dynamic range expander right in my own home circuit, built into a piece of equipment loaned to me for another purpose. It will be nameless because it is not currently in production--but do you think I didn't try it? I try everything. I listened to normal recorded music with this thing in my circuit, to see what would happen, this time.

In all the years I have been writing I had not yet heard one of these expander devices that, however noble the intentions, sounded right.

Instead, they have left me thoroughly disturbed, for they seemed always to destroy that easy sense of "presence" that is the very basis for good listening; they intruded, they called attention to themselves, like a amiable puppy that comes in and won't leave you alone.

The idea is splendid but it never seemed to work out in practice. I just found that I like the plain music better, compressed or not.

So I tried this one, now some years old-and it was the same. I used the helpful controls, I tried to fool myself by leaving the thing on and hoping to forget it was there next time--no go.

The best setting was zero. Out of the circuit. But--? Hope never dies! Even so, the idea is good, it ought to work.

In a way, I was disappointed. Because I rejoice when somebody, at last, manages to do the heretofore impossible. It happens right along, it can happen again. Reason says that this sort of expansion just has to work, if it is done right, especially now, when we can do so much more that we useta could.

The problem, of course, is that here we are dealing not with one fixed, known compression, to be precisely matched by a fixed and known expansion. Far from it. There are almost as many compression parameters as there are record companies, broadcasters, tape machine builders, and disc cutters; not only that, one compression may be superimposed upon another, in a different place and time, and indeed probably is. These gents don't even know each other to speak to, let alone compare notes on compressions. All in all, I find it astonishing that their final products get through to us with any sense in them at all, what with so many operating circuits at so many places in the audio chain. In fact, I have a healthy respect for our record producers (and even some of our broadcasters) for their practical know how.

If compress they must, then they have learned to do it gracefully.

So our expansion mirror circuit is tackling the impossible, which is to match exactly all of these alterations in the sum total as represented in the signal which reaches your home. The mirror isn't flat. It's full of sonic wriggles and wavers and bumps. That's what I have been hearing--so far. The compressions themselves, again, are generally smooth and unnoticeable, even if they are there. And yet--? I am beginning to see, after all these years, how it could be done. And I am beginning to suspect that the more recent manufacturers have figured it out.

Maybe the impossible is no longer so impossible?? There are new ways to go at the problem. But look a bit further into home listening.

The Future Force

Yes, ideally, compransion--compression/expansion--is a serious basis for the consumer dynamic range expander. But it is not the only basis. If I know the normal hi fi owner, there is another aspect and you can guess what it is-the sheer pleasure of a versatile new way to play around with your fi, to suit your taste and show off your equipment, and to heck with studious mirror images. You can build up your loud music even louder, cut back the soft parts to a smooth whisper, increase the total impact by a whole new order, and all without lifting a finger.

Nothing new in this sort of pleasure! After all, your choice of volume level has always been free and variable all over the spectrum, except for practicalities like neighbors. Some like it loud, some prefer the discreet (don't I know). Myself, I've learned to like it loud and you'd be amazed how often I am asked ... please turn it down. You think I'm just a musician? Same thing for tone controls; mostly we do with them what sounds good. And how about space expansion, reverb, Audio Pulse? All these are adjustable creatively to your choice and not necessarily according to any acoustic music original. So why not dynamic range expansion too? This is a legitimate way to use a clever automatic circuit and I see no reason to be doubtful about it, if that is your taste. A lot of people are going to like the newest expanders on just this basis, a single stage operation that enhances the signal for new creative effects. That, if you wish, is the musical wave of the future and a bigger factor in home hi-fi than most of us care to admit. I am all for it, with only one reservation: these sounds, these effects, are not for everyone.

For some of us, the two-stage expander remains a more important objective, factual compensation, as nearly as possible, for the alterations built into the musical signal. A flatter mirror is all we want. And decidedly, this is a tougher task, though I think it is the one dearest to the hearts of those who design and build expanders.

Music, acoustic music, the sound of music out of the original acoustic instrument, is still a major basis for audio reproduction and those of us who are concerned with this music are understandably leery of creative sonic extremes of any sort as the music comes forth in the home. We are not literal minded. No symphony in the home can ever sound like a symphony in the hall, nor should it. But, there must always be a relationship for us between these two extremes, one that allows us to hear the sense of our music freely "through" the audio medium with that ever-necessary illusion that it is there, right out in front of us.

To go back to my start, yes, we can do OK without dynamic range expansion, for the most part, because the altered audio product, such as it is, has been very cleverly tailored to fit the living room. And yet--? Can we do more? We always can.

No, we will never be able individually to match all those compressions with exact expansions, because we don't know which ones are there and how they work. That approach is hopeless. Not by a vast choice of different curves, either-a setting for every record company, broadcast network, disc cutting studio! That would be like the old equalizations for records, before standardization. Do I remember Old Columbia, New Columbia, NAB, RIAA, RCA, and dozen more choices; we were in Alphabet Smith's famous soup all over again. No, not that.

Computer Compransion Instead, I begin to see the way. By a modern statistical approach, maybe computer aided, of the sort we now do all the time. An expander circuit that operates--my first thought-like a Gallup Poll. Not a literal compensation for all those compressions, but a statistically derived correction, accurate plus-or-minus, almost a prediction, taking sophisticated computer account of a vast amount of known data-we do know the possible compressions. An expander circuit that would turn out to be right within acceptable limits, with only a few careful controls. Phew! One falls into jargon in these things. But this is how it could be done, and maybe has been done.

We have the two essentials. First the computer-type design techniques, the sort that have made incredible strides in the last decade. We can turn out "blueprints" of a predictable accuracy unthinkable a few years ago in the handwork, hit-or-miss, and intuition/guess era. Or worse, the systematic try-everything era. Multiply the old slow calculations by a thousandfold and you have what we have.

Tools for design. Second, we can build the circuitry to match, also maybe a thousand times more sophisticated, yet practical for home equipment.

Solid state, ICs, and all the rest.

So I am convinced, brainwise, that we CAN produce and probably HAVE produced a dynamic range expander to satisfy all of us, even including me.

And all the music lovers. It figures. It has to be! All I have to do now is to hear it.

(Source: Audio magazine, Jan. 1978, Edward Tatnall Canby)

= = = =

Prev. | Next

Top of Page    Home

Updated: Thursday, 2017-01-19 9:05 PST