DVD: A Reality Check

The announcement in October 1995 of the Digital Versatile Disc (DVD) set the hearts of audiophiles and music lovers pounding. Although primarily a digital video and computer-data storage format, DVD's massive capacity could be applied to a "super CD" audio-only disc. Finally we would be liberated from the musical limitations of the CD's 16-bit word length, 44.1kHz sampling rate, and two-channel format. We were tantalized by reports of 96kHz sampling rate, 24-bit word length, and multichannel playback. Digital redemption appeared to be just around the corner.

This belief was fostered by audio magazines eager to report on the latest technology. In addition, the much-heralded September 1996 release date for DVD players implied that the technology was fully realized. Consequently, some audiophiles are putting off buying CD transports and digital processors in fear of them becoming obsolete in a few months (footnote 1). But the widespread assumption in the audiophile community that a DVD-based high-quality audio disc is on the verge of commercial introduction constitutes a strong misconception.

The unfortunate reality is that any new audio disc is years away from introduction—if, indeed, it comes to pass at all. Many factors contribute to this delay. First, although the Japanese Conference on Advanced Digital Audio (ADA), held April 15, published a specification "wish list" regarding the standard for DVD to be used as the next-generation audio carrier (see "Industry Update," July 1996, p.29, as well as the Acoustic Renaissance for Audio proposal in August 1995, Vol.18 No.8), no standard has been agreed upon. Setting the sampling rate, word length, playing time, and number of channels is more than a technical issue; it is an exercise in what Meridian's Bob Stuart calls "geo-technical politics," with all the time-consuming wrangling that term implies.

More important, the joint format agreed to by the Sony/Philips camp and the formerly competing Toshiba/Warner group obviates the high-quality audio disc's backward compatibility with existing players. The original Sony/Philips proposal was for a 1.2mm-thick, single-sided, dual-layer disc that could store conventional "Red Book" audio (16-bit, 44.1kHz) on one layer and a high-density signal on a second layer. A conventional CD player would be able to read the Red Book layer; a high-density player would read both this and the high-density layer. The laser would simply focus to slightly different depths in the 1.2mm-thick polycarbonate substrate. Sony demonstrated this dual-layer disc at the World CD Conference in San Francisco in March 1995.

The advantages of this format are twofold: backward compatibility and single inventory. The Sony/Philips disc would not only play on high-resolution DVD-based machines, but also on the huge installed base of conventional CD players—a consumer wanting better sound need only buy a high-density CD player to access the second layer. From a marketing perspective, a single-inventory scheme is essential to a recording format's acceptance; retailers simply will not tolerate stocking another version of the same title. Without backward compatibility and single inventory, the high-quality CD will not get off the ground.

The problem arose when Sony/Philips accepted the Toshiba/Warner idea of a double-sided disc bonded together. Each half of the disc was now 0.6mm thick, not the 1.2mm of a standard CD. This modification to the original proposal introduced technical problems in putting a Red Book layer and a high-density layer on a single-sided disc. A conventional CD player designed to focus through a 1.2mm disc won't read a pit layer only 0.6mm away from the disc surface. Proposed solutions include molding a riser ring around the disc's inner diameter to position it farther away from the playback mechanism's objective lens.

Assuming that this technical hurdle can be overcome, does a disc storing 24 bits of data at a 96kHz sampling rate automatically confer high-quality sound? Not in my opinion. Although the much higher data rate has the potential to produce better sound, the implementation of the new analog-to-digital converters, mastering practices, and recording techniques are large variables in sound quality. Just because a storage format may have the capacity of 24-bit words doesn't mean that's the signal resolution delivered to your loudspeakers. And the necessary A/D hardware and infrastructure are in limited supply and almost non-existent, respectively (footnote 2).

Unless a high-quality audio disc can be produced with a minimum of additional cost, be backward-compatible with existing CD players, and provide a single inventory to retailers, I think it unlikely that the high-quality audio CD will ever become a reality. Making a better-sounding CD is absolutely the lowest priority for hardware manufacturers and record companies. After all, the general public isn't complaining about bad CD sound. Why should they invest in a new format the masses don't feel is necessary?

So don't expect a new CD format for at least three to five years. Throw in another two years for a substantial number of titles to become widely available, and you're talking about the high-resolution CD becoming a commercial reality in the year 2002. Your existing CD playback components will probably wear out before they become obsolete.—Robert Harley

Footnote 1: Although you may be able to buy a DVD player later this year, few movies will be available during the first year (see "Industry Update" in this issue). The software and hardware companies have yet to agree on licensing arrangements, regional encoding (to prevent a film's US DVD release from appearing in other parts of the world before the film is shown in foreign theaters), and other issues.—Robert Harley

Footnote 2: However, the Pacific Microsonics HDCD encoder can output 20-bit data encoded at 88.2kHz. And at HI-FI '96 in New York at the end of May, recorder manufacturer Nagra and converter manufacturer dcs were comparing 24-bit, 96kHz-sampled recordings with conventional 16-bit, 44.1kHz DATs made from the same microphone feed (full details in next month's report). The sonic difference was a) in the right direction and b) not trivial. And auditioning the 20-bit master tapes for the last three Stereophile recordings leaves me dissatisfied with the noise-shaped 16-bit versions on the commercial CDs.—John Atkinson.