THE CASE OF THE IMPROVING MEDIUM

I have been reviewing Compact Discs since their introduction, and by now I have listened to thousands of them. The most memorable and remarkable aspect of this listening experience has been the well-nigh incredible playback reliability of the medium.

Of the thousands of CDs I have listened to, I have encountered fewer than a dozen truly defective recordings. These defects were in the form of odd repetitive noises, intermittent mis tracking which caused "skips" and loss of signal, and sections where the tracking system would become "stuck" and endlessly repeat several notes of music. I also found a number of CDs which mistracked in minor ways: quite often these anomalies could be corrected simply by cleaning the disc. One might possibly attribute the remarkably low percentage of defective CDs to sheer luck, but I know quite a few people with substantial CD collections who also attest to their basic reliability.

Reliability is not a matter of luck, but is an inherent characteristic of the medium, stemming from CDs' clever and efficient error-correction system. The error-correction system is really inter dependent with the laser-tracking and focus servo systems. When these servos function properly, they can compensate for disc eccentricity, surface irregularities and undulations, variations in disc reflectivity, and even changes in the intensity of the laser beam. When the laser encounters dust, dirt, fingerprints, scratches, or any kind of oily film on the surface of the disc that can diffract the beam and de-focus it from the information pits, the CD's error-correction code comes to the rescue. This is really a combination of CIRC (Cross Interleave Reed-Solomon Code) and CRCC (Cyclic Redundancy Check Code), and one can certainly say that without it the Com pact Disc would never have become a viable music-playback system.

This code corrects many of the errors that disturb or interrupt the data stream. If each error can't be corrected outright, the circuitry will analyze accurate signals surrounding the unread able error signal, predict what data would be needed to mask the error, and then actually supply data to conceal it! Of course, when the errors in the disc are too massive, even the large redundancy built into the CRCC is overwhelmed and the result is an audible glitch or a complete loss of signal.

All of the foregoing will be rather basic information to many people, while others may find it enlightening. My purpose in discussing it, however, is to preface the point that while all CD players use the basic code as originally specified by Philips and Sony, some manufacturers incorporate engineering changes in the error-correction circuitry to enhance its efficiency. Thus, error-correction and concealment performance can-and does-vary among different brands of CD players.

Not all the errors that these systems are called on to correct are due to these infrequently encountered major anomalies in the discs. Some are due to external influences, and not enough attention is paid to the isolation of CD players from external vibrations and mechanical and acoustic feedback.

Error-correction systems treat vibratory phenomena as errors and use up so much of their correction capacity in dealing with them that even the best are at times overwhelmed and cause various distortions and dropouts.

However, while major disc anomalies are rare, they do exist. During a recent listening session I encountered a CD that could only be described as a sonic disaster area. It grossly distorted, skipped, intermittently lost signal, and emitted all manner of rude repetitive noises. Now, I was listening to this disc on one of the most highly regarded and expensive CD players, and I have access to many machines, including most of those that are considered state-of-the-art. Some of them use the Sony 16-bit linear decoding sys tem: some use double oversampling, and others are prime examples of the Philips quadruple-oversampling technique. Well, friends, I played that of fending CD on all these machines, and although some players' error-correction systems valiantly tried to restore order, all failed--save one. On that particular CD player, the manufacturer had gone to considerable lengths to improve the efficiency of the error-correction system. The company calls it "variable error correction," and while they do not offer detailed technical in formation on this proprietary system, they are in no way bashful about extolling its virtues. And well they might. for on their player. this most flawed of CDs was perfectly reconstructed and pro vided lovely music with not so much as a polite hiccup! Surely one would think the CD player that accomplished this feat must be the latest and greatest chromium-plated. turbo-charged example of the digital art. Would you believe it was the quiet, unassuming, but well-dressed ADS CD3? This middle-priced CD player is a nicely integrated design. a well-made. user-friendly unit that com bines functionality with an uncluttered control panel. The machine provides superior performance in all the important CD parameters, and deserved the quite favorable review in the June 1985 issue of Audio. I have not tried my nasty CD on every CD player, and there may well be some other machines which can cope with its anomalies, but the CD3 is an exemplar for the outstanding qualities of its special "variable error-correction" system.

I am not saying that the performance of error-correction systems is the single and absolute criterion for judging the quality of a CD player. Certainly there are many other important considerations. However, there is little doubt that good error correction plays a major role in successful CD playback.

To continue the saga of my errant CD: The manufacturer of the expensive, high-quality player I was using earlier when I discovered that CD asked if they could borrow the disc.

They were quite chagrined, and I expect a bit chastened. to find that their really quite splendid CD player did in deed have problems playing through my super-defective disc. At their plant, the CD was examined under a high-power microscope. which revealed that in the passage with the most anomalies, the information pits were oddly shaped and a number of tracks were skewed beyond the tolerances normally allowed in the mastering.

This by-now notorious CD is at present on its way to Philips in Holland, where it will be examined under a scanning electron microscope. Under the tremendous magnification possible with the SEM, the nature of its problems will quite likely be clarified.

Incidentally, I have a duplicate CD of the one which had caused so much trouble, and it plays back flawlessly on all the CD machines on which I have tried it, no matter what their pedigrees.

Could this CD have been processed, somehow, from a different master of the same program? I found my involvement with the anomalies of CDs--and the complexities of error correction--quite interesting and stimulating. It certainly impressed me anew with what a techno logical tour de force the Compact Disc system is, and made me marvel that such a sophisticated product is so widely available and so reasonably priced.

(adapted from Audio magazine, Jul. 1986; Bert Whyte)

= = = =