THE END OF THE BEGINNING
When "Digital Domain" first appeared, in April 1984, I wanted to start my stint at Audio with a splash. So I shocked readers by declaring that analog media had been a mistake. I said that an analog signal is indistinguishable from the noise and distortion inherent in the medium and declared that this "spells failure for an audio information-storage system."
I expected a few letters, and received many. A lot of people were really steamed. They couldn't understand how the high art of analog technology could be so easily dismissed. They argued that analog, whether acoustic or not, was the only true carrier of audio, and that digital could only form a numeric approximation of the original event.
But I also got a lot of letters from readers expressing complete agreement. They immediately accepted the notion that digital technology inherently promised far greater returns. It was that unexpected response which actually made a believer of me. Unsolicited enthusiasm for the new technology made it clear to me that its success was no longer in question. I got D/A converter, because only one-bit words are output.
Speaking of D/A converters, we saw players with one, two, or even four converters. And their word lengths varied from 14 to 16 to 18. Upstream, we saw all kinds of error-correction algorithms, including players that diagnosed the down to business, trying to keep track of digital's rapid progress.
It was a tough job because things happened fast and furious, particularly in the pioneering digital audio system--the Compact Disc. Whereas some observers originally predicted that the advent of digital technology would reduce all products to the sameness indigenous to ones and zeros, the exact opposite came true. A real binary renaissance occurred. Technology sprouted in all directions.
We saw laser pickups with one- and three-beam designs. There were players with brickwall analog output filters, and with two-, four-, eight-, and 16-times digital oversampling filters. Compact Disc co-inventor Philips has even devised a chip set with a 256-times oversampling filter; this innovative design doesn't even have a traditional kind of error they had encountered and selected the appropriate correction method. As the circuitry got more sophisticated, the parts count grew smaller. Early players had masses of chips in their primary sets; that new Philips chip set has two.
Although we heard more than a few tall tales about digital audio, we also encountered unexpected yet bonafide twists and turns in the technology.
Data recovery, it turns out, isn't entirely immune to many banes of analog storage, such as vibration. Suddenly CD players appeared with elaborate anti-vibration measures, including isolation feet, floating disc drives, sandwiched cabinet metals, and even fairly exotic transport parts made of composite ceramic and fiberglass materials.
It was soon discovered that all digital data is dirty (due to all that high-frequency clock information spraying around), and countermeasures were introduced. Some players used opto -isolators to block digital noise and to prevent that noise from entering sensitive analog output stages. To prevent spurious beat frequencies, some Compact Disc players used a single master clock to synchronize multiple circuits.
CD players themselves branched out into diverse forms. Single-disc home players were soon joined by portables and changers.
Car players were introduced, including trunk-mounted changers. As competition heated up, all kinds of features appeared. Some players flashed a light when there was a disc error, some randomly selected tracks, some had pitch control, and most were programmable and came with remotes. You could buy a CD player for $90 or $9,000, and hear claims that machines at either end of the price spectrum would sound the same.
CD players had subcode jacks for a while, then those faded away. Apparently, the jacks will begin to appear again on some new players, in the guise of graphics outputs. Many players sported digital audio output jacks. These jacks tapped data off before the analog conversion section, to deliver the error-corrected bit stream to an outboard D/A converter (perhaps in a preamplifier). On some players, that digital interface appeared in two forms-electrical coax and optical fiber.
The Compact Disc itself underwent several dramatic transformations. No one could decide on its maximum playing time. At first, as pressing plants ventured into that tricky submicroscopic business, we were lucky to get 60 minutes. Then disc length crept up toward 70 minutes, and reached the announced limit of 74.
Now, an 80-minute disc has appeared. And slowly, disc prices have started to relent a bit.
At whatever length, that shiny 12-cm platter was the exclusive property (like it or not) of music lovers. But its incredible storage capacity soon attracted other corners. CD-ROM was introduced for data-storage applications.
CD-V was announced as a combination audio/video storage medium. CD-I holds great promise as an interactive, multimedia format. The CD also shrunk to an 8-cm size, called CD-3. Now that graphics appears destined to be reintroduced, there is talk of CD-G. Just to make sure no one gets confused, the original CD has officially been renamed CD-DA, which stands for Compact Disc-Digital Audio.
Hardly a day goes by without yet another flavor of CD. The most recent is THOR-CD, a recordable/erasable disc announced by Tandy Corporation. (THOR stands for Tandy High -intensity Optical Recording.) As "Digital Domain" readers know, recordable CD technology has been around for some time, and includes the MOR (magneto-optical recording) system we examined some time ago.
About the only thing that has remained the same is the sampling rate of 44.1 kHz. That good old number has never varied. (One question we haven't gotten around to is exactly where that decidedly odd amount came from.) Of course, CD was only part of the fast-paced digital story. R-DAT was announced, but it quickly met opposition from record labels. (These companies apparently were unwilling to suffer the same fate as the software manufacturers that banked zillions of dollars in profits from another recordable medium, videotape.) Still, R-DAT is here.
You can walk into your Ford dealer today and buy an R-DAT player and premium sound system for only $1,540. However, the 1988 Lincoln Continental which carries it costs slightly more.
With all the excitement of CD and the notoriety of R-DAT, many people have come to equate digital audio with its storage media. This is only the opening salvo. Just as digital technology has upgraded that weak link in the audio chain, it will upgrade the other links.
Many preamplifiers already contain D/A conversion circuitry to permit a digital-to-digital transfer from a digital storage medium. Soon, the control circuitry itself will go digital as well, with digital equalization, gain control, balance, switching, etc. Similarly, the day isn't far off when digital power amplifiers will be introdu9ed. That should really cause a stir in vacuum-tube enclaves! Most of audio's diehard notions are in flux. In the analog past, the storage medium was all important. The storage accuracy (or lack of it) largely determined the sonic quality of the end result. With digital, storage is relatively unimportant. As long as enough of the data is there to permit the audio signal's reconstruction, the medium is good enough. And why bother with separate components? All you really need is a minicomputer with storage and user interface.
In short, as many readers confidently expected, digital audio is redefining the art and science of audio technology. Moreover, this is just the beginning. Digital technology has taken the first small steps required to upgrade storage. But new kinds of digital signal processing, and even newer kinds of storage, are in the wings. Add entirely new concepts in hardware and software, such as hypermedia-a human, idiosyncratic way of integrating information-and it's hard telling in which direction things might go.
One thing is sure. Digital audio is off to a great start. Most audiophiles are strong supporters (voting with their pocketbooks) of digital audio technology, and look forward to any future advances made possible by the creativity it engenders.
In the first "Digital Domain," the stated objective was to "achieve an understanding of digital audio, in terms both of bits and bytes and of the philosophical nature of this important technological development." Now that we're entirely comfortable discussing things like sampling rate and hypermedia, our goal has been met.
Thus, I hope Audio readers have benefited from, or at least enjoyed, this column. Since my goal of introducing digital audio technology has been served, it is time for this author to say goodbye and move on to other professional responsibilities. Thanks for including me in your reading itinerary, and thanks for all those letters.
(adapted from Audio magazine, Aug. 1988)
= = = =