Technical Articles / Reports
The 2001 CES Report by E. Brad Meyer

High-Bit Follies

For many years audio, once the center of attention, seemed to be the abandoned child of CES. The exhibits of many equipment manufacturers were relegated to distant sites, leaving a few big companies at the main convention hall. A group of high-end companies, dissatisfied with their venue, the Alexis Park Hotel, opened their own show (called T.H.E. Show) next door at the St. Tropez, a physically similar facility. This year, the competition between the Sony/Philips Super Audio CD and everybody's else's high-bit format, DVD Audio, moved the center of attention southward as Philips and Sony took space outside the CES umbrella at the St. Tropez to promote and demonstrate SACD. Most of the DVD-Audio demos were in large booths at the convention center. One of the official CES news stories pointed out that there were five conference sessions at the show devoted to audio topics, with two special sessions on mobile audio and digital, and that there were audio exhibits in all four main show facilities.

At first, the war looked like a battle between two doomed camps. I've written before that since the compact disc is the most successful, durable and universal audio carrier ever, it is an exercise in futility to try to introduce a new audio medium that can't be played on 800,000,000 existing CD players worldwide. As of last year, DVD Audio, which from the start has made use of the extra channels that seem to be the main reason for changing over, had been delayed over and over by problems with copy protection and standards; SACD for its part was introduced solely as a two-channel medium for extremely rich audiophiles who believed salvation lay in more bits; and neither group seemed inclined to use what I thought was the most important part of both specifications: the possibility that the high-bit audio would be encoded on a semi-transparent layer overlying an aluminized base carrying a conventional "red-book" two-channel CD bitstream.

The picture has altered in several ways, many of them relevant to these arguments. The first big change is the phenomenal growth of DVD video hardware. DVD is the most successful format introduction in electronics history, having grown in three years to a total installed base of 14.5 million players in the U.S. alone. This growth curve is steepening, too; more players were sold in December, 2000 than in all of 1999, and the total may reach 30 million by the end of 2001. Here's the big payoff for the audio people: Virtually all the discs in both new formats (all of them in SACD) contain, in addition to two or more tracks of high-bit audio, a DVD video track, sometimes with a simple still picture in the video segment, that can be played on existing home theater setups. (For the technically curious, the video is inside the high-bit area, which occupies a band about an inch wide around the outer part of the disc.) This is crucial, since home theaters comprise the huge majority of existing multichannel systems. Suddenly, record companies have an alternative to bringing discs out into a tiny market of early adopters. Everyone with a DVD player can at least put the disc in and hear something. Whatever you may say about sound quality or bits, the biggest difference still lies between music and silence.

Multichannel SACD -- The Press Conference

The SACD camp seems to have realized that the high-bit two-channel productions that so thrilled the audiophile market will not sound so impressive coming out of the front two channels of a home theater system, less so in fact than multichannel DVD Audio encoded in Dolby AC-3 (or DTS). They have turned their attention to multichannel, and one of their two demo rooms was devoted to the "new" format. (It's new partly because those people who shelled out $5,000 for the first SACD player are already stuck with obsolete equipment that is not upgradeable.)

A Philips representative, having boasted to reporters during their SACD press conference about DVD Video compatibility, was asked what would happen if a new multichannel disc was inserted in a two-channel SACD player. The answer: nothing. (The biggest difference is between music and silence.) It is tempting to say that if this format fails it could be largely due to this blow seemingly aimed at early adopters -- rather like the move Sony made in the late '70s to Beta II speed that left the original Beta I buyers in the lurch -- but in fact there aren't enough of these people to influence the market in any meaningful way, and they'll probably all be so busy enjoying their newer and much cheaper multichannel players that they won't complain. The truism remains: The pioneers are the ones with all the arrows sticking out of them.

Hardware prices for the formerly stratospheric SACD, meanwhile, are falling fast. At the show Philips introduced the SACD 1000, a multi-channel player with a $2,000 MSRP. And Sony says they will introduce this summer (Philips promised their equivalent by next year) a five-disc changer that plays DVD video, CD audio and SACD for about $400 -- a huge leap into a much broader market. One hint about why this is a marketing decision, not a technical one, came when the Philips representative boasted that many SACD-capable DVD video players would become available soon, because it was so inexpensive to add SACD to an existing design. Asked how much, he said softly but unmistakably that the answer was "in the single U.S. dollars". There are DVD players selling for $125 retail now, so what with the extra jacks and wiring and more pages in the manual and so on, you might have to charge $200 for the SACD version -- if anyone decides to aim that low.

In another move that could still prove decisive in this battle, the SACD camp has pursued dual-layer technology to make their discs playable on those other eight hundred million machines. There is a picture in pressday2.htm of an SACD jewel box with a little round sticker on it, indicating that it bears a conventional CD audio layer. The SACD press conference room was lined on two of its walls with SACD titles, no two the same, and while the man from Philips shared his company's marketing plans with us I counted the stickers on the front wall. There were 40 titles there, 15 with CD Audio.

The SACD wall in the Philips room at the St. Tropez. There were 140 or so titles on two walls, of which approximately 20% contained a second layer of conventional CD Audio pits.

My view of the side wall was blocked but those discs appeared to have significantly fewer CD layers among them. There are currently 180 titles available, according to Philips, of which roughly 140 were on display in the room.

If available titles are an indication of which format is ahead in a race, the SACD camp has taken the lead, and they know it. When asked if they had plans to include DVD Audio capability to their players, since these chips too might be very inexpensive, Philips gave a reply that translated loosely into movie vernacular would be, We don't need no steenkin' DVD-Audio.


Multichannel SACD -- The Demo

The Philips demo began with a statement that hardware is useless without music, which is what drives everything -- a gentle reference to their growing catalog of titles. The first selection was a Vanguard recording from the '60s of a female voice with a small chamber orchestra. There was a noticeable peakiness at around 1 kHz in the miking of the voice, emphasized somewhat, I suspect, by the monitors, but the music-making was lovely and the selection was very effective. We heard multichannel presentations of various sources, including some choral music recorded in three versions on the same disc -- a CD audio layer and two-channel and six-channel SACD versions. Representing larger forces was a piece of movie music recorded by the London Symphony at Abbey Road, which was convincing in its overall presentation despite some rather unpleasant string sound; the usually more difficult brass and percussion were excellent. Overall the demonstration was effective and well done, with Philips getting extra points for turning off the air handling equipment before starting the music.

DVD Audio

There were numerous DVD Audio demos at the show. The one at the Panasonic booth had problems compared to the quiet Philips venue -- as usual for booths in the main hall, there was a fair amount of the din outside leaking into the room. Nevertheless, the sound quality of the various selections was vivid and convincing on the large Technics monitor speakers (there are expensive Japanese monitors, most of which are not sold here, and the best are very good). I wasn't expecting to be impressed, but I was, not just by the multichannel demos but by the overall quality of the sound on everything. The DVD-Audio people have emphasized the extra channels from the very beginning, so it's logical for them to assume that their productions will be played primarily, for now, on existing video systems.

At the DVD consortium's event high atop the Mandalay Bay hotel were these extremely slim Panasonic portables -- DVD Video on the left, DVD Audio on the right.

Since those will contain a DVD player, the DVD-Audio camp hasn't paid any attention to the notion of dual-layer discs. This is going to cost them in the near term and possibly for longer.

High-bit Audio -- Myth and Reality

Despite widespread agreement to the contrary among audio writers who refuse to do controlled testing, there is no reason that the extra bits, helpful as they may be to the marketing of DVD-based audio, should actually make any audible difference 99.99% of two-channel material, especially older analog recordings. More frequent sampling, which intuitively seems like it would make the waveform more accurate, in fact does nothing but extend the high-frequency response farther into inaudible regions. More bits per word, similarly, do not decrease distortion, but only lower the noise floor below the already inaudible -92 dBFS of 16-bit CD audio. (For an explanation of how it all really works, see The BAS Speaker, Volume 22, Issue 5/6.) So why do so many high-bit demonstrations sound as good as they do? Even Neil Young, one of the fiercer and more insistent anti-CD voices over the years, is now on record as saying that DVD audio finally delivers to consumers the sound he wants them to hear.

The primary difference I've heard in the new releases is that the all-important frequency balance is better on the new productions -- and also on remasters of old ones, where peaks, particularly in the upper midrange and treble, are more assiduously corrected than formerly. (The old Vanguard release we heard at the SACD demo showed that some of these effects remain, even though the label's founder, tk Solomon, declared it superior to what he has heard on CD.) I'm not saying that ordinary CDs have been getting steadily better since 1983; the very first releases were generally very good, and it took a year or so before really nasty-sounding productions began to appear. Another difference is that the goal of the high-bit mastering engineer (who in in a surprising number of cases at this point is probably one man, Robert Ludwig; Ludwig was at the show and you'll read more about him in the next audio report). No one is yet mastering high-bit audio the way engineers do some popular CDs, namely so that they sound the loudest and most aggressive on the radio. The producers and engineers on these projects have the unprecedented luxury of listening to them on Ludwig's or other top-drawer monitoring facilities, and of making them sound as good as possible on those systems, without worrying about the rest of the world. The nature of this assignment, and their appreciation for it, shows in their results.

Now that high-bit hardware is becoming more common, we cheapskates at the BAS will get our hands on it and start to do the tests that are so carefully avoided by everyone else -- level-matched, blind comparisons of recent high-bit releases and their 16-bit equivalents. Unfortunately you can't switch quickly between the CD and high-bit versions on a single disc, as the two programs have to be loaded separately -- prepare yourself to be just as annoyed at the wait when you load these discs as you now are by DVD movies -- but we will find ways around all the obstacles. We have already seen from earlier attempts to improve on 16-bit audio like Sony's super bit-mapping demos that remasters of old productions are always equalized so as to be audibly different from prior versions, so comparisons with previously released CDs aren't going to tell us anything.

Stay tuned to this site for more on high-bit audio as it develops, and to the main CES for more audio reports. Coming soon will be an account of the mammoth 10.1-channel demonstration put on by Tomlinson Holman and his partners in crime, including the reason it really had 12.1 channels, plus more pictures of actual audio hardware.

© 2001 by E. Brad Meyer

Send any questions or comments to EBradMeyer@att.net.

 

The Boston Audio Society
PO BOX 260211
Boston MA 02126

problems? email Barry: webmaster@bostonaudiosociety.org

updated 11/11/04