SACD & DVD-A: Launch Issues
By John Atkinson   •   October, 1999 Someone once said that if you build a better mousetrap, the world will beat a path to your door. Well, this month, we will see not one but two better mousetraps, in the form of Sony's and Philips' Super Audio CD and the DVD Forum's DVD-Audio. Both are intended to replace the humble CD, now in its seventeenth year; both offer higher-resolution digital audio; and both offer multiple channels. To accompany SACD, Sony's $5000 SCD-1 two-channel player is now on sale (and will be reviewed in the November Stereophile), while Panasonic has announced October sale dates for two DVD-A players, the $1000 Panasonic DVD-A7 and the $1200 Technics DVD-A10.

Hardware is just meaningless metal sculpture without music to play on it. According to information released at HI-FI '99, SACD software will be initially available from Sony, Telarc, DMP, AudioQuest Music, Mobile Fidelity, and Water Lily Acoustics—while DVD-Audio will be supported by releases from Universal Music Group and Warners. Certainly the SACDs exist—I am looking at a stack of them next to my computer monitor as I type these words in late August—but no one yet has seen a true DVD-Audio disc. (The 24/96 discs offered by Classic Records and Chesky are actually DVD-Video discs with the video content reduced to make room for the higher-resolution PCM audio.)

To my surprise, there has been resistance in some quarters as to whether either of these new formats really is a better mousetrap. The argument goes along the lines of "The 16/44.1 'Red Book' CD standard is quite good enough for the man in the street and there aren't enough audiophiles in the world to make SACD or DVD-A economically viable as a medium." Some skeptics even doubt that there can be any audible improvements due to doubling the sample rate or increasing the word length over the CD's 16 bits.

This I find hard to believe, given my own experience of high-resolution digital audio from the recordings I have made. A couple of days ago, I was uploading to the Macintosh hard drives the session tapes for the recording I had made just before HI-FI '99 of Musical Fidelity's Antony Michaelson performing the Brahms and Mozart Clarinet Quintets. Although the 24-bit master tapes had been recorded at an 88.2kHz sample rate, I was downconverting the tapes to 44.1kHz to create the computer files I was going to use to edit the CD master. (My Sonic Solution digital audio workstation has not yet been upgraded to run at the higher rates—it requires a little matter of several thousand dollars.)

The uploading is tedious work, as one hour of four-channel music requires two hours of upload time and another hour of archiving the files to Exabyte tape, but I was able to do other things, such as preparing this issue's "Recommended Components" listing, while I listened to the music.

Once everything had been archived, I switched the dCS 972 converter I was using as a sample-rate converter back to the master tapes' 88.2kHz rate. After three days of being immersed in the 44.1k sound, the 88.2k experience was like a slap in the face: Not only was there "more there there," to use Sam Tellig's immortal phrase, but the relationship between the instruments and the reverberation of Chad Kassem's Salina, Kansas church was far clearer.

This was particularly true of the sound produced by the spaced pair of B&K (DPA) omnis (footnote 1) that I used: the stereo imaging produced by this technique is always unstable and vaguely defined at best. Yet when the sample rate was doubled, the "walls of the hall" were more audible. In addition, the bass was far better defined. I have heard this counter-intuitive effect before—you'd think the high frequencies would be most affected, but it is the lows that become clearer at the higher sample rates.

This seems like a fair test, as the playback chain was always the same, the only difference being the downconversion to 44.1kHz in one case. The dCS 972 was always in-circuit, converting the Nagra-D or Tascam/PrismSound 88.2k 24-bit data outputs from two normal-rate AES/EBU feeds to a single 88.2 AES/EBU datastream, and the D/A converter—the Mark Levinson No.30.6 I review in this issue—was the same both times.

So, as far as I am concerned, DVD-Audio, which will offer at minimum two-channel high-resolution linear PCM audio, is indeed a better sonic mousetrap. And while I have yet to hear Sony's high-resolution DSD format used on Super Audio CD in my home—that will happen next week when I receive the player from Jonathan Scull for measurement—based on my experience hearing it at HI-FI '99, at Sony's New York headquarters, and at AES conventions past, it too is a better mousetrap.

I therefore enthusiastically welcome both new media—and I can't wait for readers to hear my Stereophile recordings at the higher sample rates. However, I do wonder if those who doubt whether better sound quality is alone sufficient to guarantee commercial success have a point. I bought my first stereo system in the 1960s, for example, not because I was dedicated to sound quality in itself but to better appreciate the music that was hiding in the grooves of Sgt. Pepper's Lonely Hearts Club Band, Wheels of Fire, and Electric Ladyland.

Regardless of sound quality, the ultimate success of SACD or DVD will be music-driven. In the same way that it was classical music that drove the first hi-fi boom in the 1950s, and the Woodstock-generation rock that drove the 1970s hi-fi boom, the eventual success of either or both of these new media will depend on the release of recordings that take full advantage of what they have to offer.

---

Footnote 1: I used two pairs of mikes for the recording—the spaced B&K (DPA) omnis and an ORTF pair of Neumann M147 cardioids—to allow some flexibility in producing the final mix. The Mosaic CD was released at Home Entertainment 2002 in New York and can be purchased from the secure "Recordings" page on this website.