| Home | Audio mag. | Stereo Review mag. | High Fidelity mag. | AE/AA mag. |
|
By LARRY KLEIN Technical Editor FOR those of us concerned about the sound quality of records, there is good news and bad. The bad news-which for many readers will be old news-is that the oil shortage has resulted in a vinyl shortage, which is forcing record-pressing plants to use increasing amounts of "regrind" in new discs. The not-so-raw material for regrind comes from defective discs and vinyl waste from the pressing line, and discs defective and otherwise returned by distributors. The use of regrind, we are told, produces an inevitable increase in surface noise be cause of the foreign matter that finds its way into the old/new vinyl mix. However, perhaps all is not lost. I recently listened to a record that had the widest dynamic range and lowest noise level of any disc I have ever heard. When the music was loud, it was very, very loud, and during the silent sections one was tempted to check to see if the stylus was still in the groove. The record was produced by the dbx Company, using special encoding on the disc, and it was played through a stereo system that had a dbx decoder patched in. To understand the dbx development, some background in noise reduction is helpful. The compression/expansion (c/e) technique has been used for noise reduction in recording studios and else where for many years. A c/e noise reducer works, in theory, as follows: the audio signal has its dynamic range sharply reduced (compressed) by amplifying the low-level signals and/or by cutting back the high-level signals. This is done before the signal is fed to those devices--such as any kind of transmission line or tape recorder-that contribute unwanted noise. During playback, a corrective expansion is applied by cutting back or amplifying as necessary to restore the original dynamic range. In the process the noise introduced after signal compression is simultaneously reduced. Simple enough in theory, but in practice several complications arise. Obviously, the compression and expansion have to be exactly complementary not only in respect to the operating levels, but throughout the audible range if annoying frequency-response aberrations and noises are to be avoided. In addition, there's the problem of "noise modulation." This term refers to the fact that when the noise and the boosted signal do not overlap sufficiently in frequency for psychoacoustic masking to occur, the c/e action may convert the noise from a steady hiss (which to some degree can be subjectively tuned out) to an obtrusive pumping or swishing sound that varies with the music. To avoid these two problems, engineers have devised various ingenious proprietary techniques, the best known of which is the Dolby sys tem. Most current noise-reduction systems are much more complex than indicated above. Until dbx made their move, no one had ever included the stereo disc itself in the noise-reduction process. It's clear that the more elements of the record/ playback chain you can include in a c/e loop, the better the result is going to be. And since the dbx system does not decode until after the phono preamp, this means that turntable rumble, disc-surface noise (remember regrind?), and even preamp hiss are sharply reduced. The reason for the improvement in dynamic range I heard may not be immediately apparent. Perhaps you have noticed that if your amplifier's volume control is set to provide a really loud signal during heavily recorded passages, the softer passages are frequently smothered by noise. Obviously, if some of the noise could be eliminated, we could start to approach the range of soft-to-loud (the dynamics) heard during a live performance: this would contribute greatly to reproduction realism. The dbx decoder used in the demonstration was a standard unit originally designed as a noise reducer for the professional tape recordist. As a disc decoder, it could be made less expensive (dbx is thinking about a $100 unit with phono preamplifier built in). You would simply plug it in between your record player and an auxiliary input and be all set. But what would you put on the turntable? I hope I have made it clear that the dbx system can do nothing for standard discs. For it to work, we need dbx encoded discs-and these would not be playable without a dbx decoder. Who is going to manufacture such special discs? Will the dealers carry them? What pro gram material will be available? These questions have yet to be answered. A year or so ago Ray Dolby told me that one reason he had not pushed a Dolby system for discs is that, to be effective, it (like the dbx) would have to be non-compatible. He did not want to introduce further confusion into a market already having trouble enough with four-channel recordings in their competing disc and tape formats. Did Dr. Dolby read the signs correctly-or did he fumble the ball at a crucial moment? Several years from now the answer should be very clear. Also see: |
Prev. | Next |