Perfect Sound Forever?

When some unknown copywriter coined that immortal phrase to promote the worldwide launch of Compact Disc in late 1982, little did he or she foresee how quickly it would become a term of ridicule. Yes, early CDs and players offered low background noise, a flat spectral balance, and freedom from wow and flutter. But all too often, the music encoded in the "perfect-sounding" pits seemed to have taken a vacation, leading the renowned recording engineer John Eargle to offer, in the medium's defense, that if you were to hear just one CD that sounded good, digital technology would be proved to be okay.

While we waited for that exception to truly prove the rule (footnote 1), we found that neither was the word "forever" a guarantee. I remember one radio station proudly proclaiming its allegiance to the future by promising to play CDs exclusively, only to have its very first broadcast marred by a skipping CD. (That this can still be a problem can be seen from the discussion of Stereophile's Concert CD set in this issue's "Letters.")

But even as "perfect sound" and "forever" were being recognized as marketing hype, the problems with both the CD medium and with digital recording were starting to be addressed. It was known even before the CD launch in 1983 that professional digital recorders needed word lengths longer than 16 bits if there was any chance that the data on the CD would have anything approaching true 16-bit resolution. Dreadful digital editors were replaced by transparent hard-disk systems. The performance of analog/digital converters took a quantum leap forward in late 1988 with the introduction of the Robert Adams-designed 20-bit UltraAnalog ADC module. The realization that datastream jitter could significantly reduce signal resolution took rather longer to develop; but coupled with the introduction of quieter, more-linear DAC chips and high-resolution digital filters, this enabled even inexpensive playback systems to achieve true 16-bit resolution in the mid-'90s.

However, even as that was happening, engineers started to realize that significantly better sound quality could be achieved with digital word lengths longer than 16 (see the January 1994 "As We See It"). Mastering techniques and processors such as Sony's Super Bit Mapping, Apogee's UV-22, DG's 4D, and Meridian's 618—the latter used on Concert—all preserve a significant amount of 20-bit sound quality on the 16-bit CD. And now we have the commercial availability of High Definition Compatible Digital CD (see our interview with its inventors in this issue), which uses a buried subchannel to take 16-bit replay to a new level of sonic performance—at least in the opinions of Robert Harley and myself.

Yet so far, the overall reaction to HDCD has been mixed. Putting aside commercial questions regarding Pacific Microsonics' licensing policy and the fact that only one record company, Reference Recordings, is currently releasing HDCD-encoded CDs, not all who have heard decoded HDCD think it offers significantly improved sound quality. This puzzles me, as I feel HDCD's opening up of the reproduced soundstage and its clearer presentation of recorded detail are reminiscent of what I have heard from true 20-bit digital media. But some denizens of cyberspace have found decoded HDCD recordings to sound merely different rather than better, leading me to muse that if we actually heard perfect sound quality, would we even recognize it for what it was?

The answer to that question will become clearer once more record companies release HDCD-encoded CDs and as more D/A-processor manufacturers incorporate the Pacific Microsonics digital filter chip. I wonder, however, if a more serious criticism of HDCD is that it might be too late. As you will have read in April's "Industry Update" (pp.331-37), the 16-bit CD itself could be superseded. Both Toshiba and Sony/Philips have announced CD-sized media—two different "Digital Video Discs"—that can carry between four and twelve times as much data as a CD. Although this new technology is aimed at the video market, many see its potential for carrying a music signal encoded with up to 24-bit data words or with a higher sampling rate—either of which might render moot the need for HDCD.

Don't throw your CD and laserdisc players in the trash just yet. HDCD offers a real jump forward in sound quality right now, and I don't think a high-density CD medium—whichever of the two wins out—will be a commercial reality much before early 1997 (footnote 2).

But I am excited about the potential for the true coming-together of video and audio offered by DVD. I envision a common-carrier future in which a header on every 5", high-density disc tells a universal decoder what it is: an MPEG-2-encoded movie with a Dolby AC-3-encoded, 5.1-channel soundtrack; or a movie with a DTS Zeta-encoded six-channel soundtrack; or a movie with three different-language, Dolby Pro Logic stereo soundtracks; a 24-bit stereo music recording; a 16-bit, 44.1kHz-sampled, HDCD-encoded, six-channel music recording; a 20-bit, 88kHz-sampled, Ambisonics-encoded surround-sound recording; and so on. All the signal processing would be handled by a versatile DSP engine, the player automatically adjusting the playback decoding algorithm—video and/or audio—to whatever was appropriate.

And if a new recording technology was introduced, the owner would simply upgrade her player's software. Or, more likely, the disc itself would contain the new instructions for the decoding DSP—much as Dolby Stereo Digital films include the AC-3 decoder software in the data area before the movie starts. (The cinema's system automatically updates itself if it detects a more recent software version on the film.)

As with all future-gazing, I'm probably more wrong than right. But one high-density medium for many high-quality formats—that's an idea that excites me even more than the introduction of HDCD.—John Atkinson

Footnote 1: The aphorism "The exception proves the rule" is almost always bandied about incorrectly. If you do find an exception, far from the rule being "proved" to be true, the exception means that the rule has been tested and has been found to be wrong. In this case, however, the exception would indeed prove the validity of CD's "perfect sound."—John Atkinson

Footnote 2: Future watchers tend to be too pessimistic about the timing of their predictions. With 2002-vintage hindsight, my "1997" for the introduction of hi-rez audio media based on the DVD-V platform was wildly optimistic.—John Atkinson

Share | |
Site Map / Direct Links