POPMusic 版 (精华区)
发信人: crazy (雪山), 信区: Music
标 题: DAD-the absolute sound
发信站: 哈工大紫丁香 (Wed Apr 5 11:45:55 2000), 转信
From out of left field, and to virtually everyone's surprise, a small consortium of High
End companies unveiled the Digital Audio Disk at the January Consumer Electronics
Show in Las Vegas.
The handful of companies, evidently egged on by Michael Hobson of Classic Records,
are launching a preemptive strike, one reminiscent of that in 1957 when one small
audiophile label, in the face of noodling by the major labels, thumbing its nose at the
big boys, released the first stereo disc on vinyl and, de facto, set the standard for LP
stereophony.
That company was the smallish Audio Fidelity, headed by a most aggressive Sidney
Frey. The industry of the day was debating, endlessly, how best to get two
stereophonic grooves on a long playing disc. The British (headed by Decca/London)
were for the so called hill-and-dale system, with one channel recorded laterally (just
the like mono discs of the time), the other vertically, thus putting the stylus in contact
with the bottom of the groove. The Americans wanted to turn the hill-and-dale system
on its sides, 45 degrees off the hill-and-dale axis, thus the 45/45 cutting system. Jerry
Minter, who would be called a techie today, proposed an FM carrier system, with one
channel encoded (via FM) supersonically in the grooves, even though there was
considerable doubt at the time whether a cartridge could reliably trace signals above
20,000 Hz.
Each of the systems had its proponents and each its disadvantages (the
Westrex-sponsored 45/45 was, in theory anyway, the weakest of all in terms of
channel separation and therefore effective stereo). What Frey did was simple enough.
He asked Westrex to cut a test pressing, with its 45/45 system, of a Dukes of
Dixieland album. Now Westrex wasn't born yesterday and it knew of Frey's carry-like
salesmanship (a reals-bomb blest on disc? Yes!), so to frustrate any thoughts he
might have about having the disc pressed and commercially issued, it locked the end
of each cut, so that you could play only one cut at a time, before having to get up,
move the pick up arm over to the succeeding cut. Locked grooves? No problem. Frey
issued the disc anyway, even though there were no stereo cartridges, no stereo
preamps, no stereo anything to play it back with.
There was an avalanche of publicity and the mad scramble to go "stereo" started.
When the Hobson-led consortium held its Vegas press conference, there were no
commercially available High End players to play back the four discs (plus sampler)
that Classic Records had, test pressings in hand. The only players that could
reproduce these discs were a handful of Digital Video Disc players, designed largely
for home theater video playback, with audio sections that are High End NoCD (not our
class, dear). Indeed, the first player he dropped off in Sea Cliff - so we could audition
his discs - was a Pioneer 505 DVD unit.
I am assuming that you all know that the Digital Video Disc is technically capable of
reproducing program material with digital sampling at 96 kHz, as opposed to the 44.1
standard of the regular compact disc. What his means in practical terms is that 96k
sampling offers something closer to High End sound than one will ever get with the
44.1 standard. Why? Because the frequency response of a pulse-code-modulated
signal can only be half that of the sampling frequency (thanks, Dr. Nyquist). That
would mean an upper frequency limit of 22.05 kHz theoretically (to prevent the ugly
demon called Aliasing); in practical terms, the real upper limit is much lower, circa 19
kHz. On the other hand, 96 means a upper frequency (overtone) limit of somewhat
less than 48 kHz, which ought to be wide enough to put the resolution back on our
source material, and the High back into High End. And I probably should warn you
that some of the best ears in the business think a 192 kHz sampling rate to be near
the last word in transparency, and, often enough, indistinguishable from a live feed.
I also assume that most of you know that, to quote Muse's Kevin Halvorsen, "a 16-bit
sample can contain 65,536 possible levels [of sound dynamics].... In the case of
24-bit samples, these can contain up to 16,777,216 possible levels." Ergo, more
bandwidth, more resolution in the 96/24 discs, as these issues are being called. Never
mind that there is, at present, no such thing as 24-bit resolution (there is the small
matter of the inherent thermal noise barrier of resistors that may prove to be as
intractable as Einstein's speed of light), and there may never be. Still, say the industry
folks, from Halverson to Hobson, 20-bit resolution (which is about what you're getting)
is a significant - and huge - jump over 16 bits.
There is also some confusion about what we're going to call this new disc. Halverson
likes the Advanced Audio Disc. It, like the CD, can be played back on a standard
DVD transport. But it can't be played back on any standard CD deck. You see,
Toshiba holds the patents on the DVD system. And Sony and Philips all the patents
for the compact disc (they get seven cents per disc!). If a DVD-based system
triumphs in the marketplace, the big boys have lost all those millions upon millions
upon millions from their CD patent royalties. So Sony and Philips have come up with
their own high-resolution answer to the threat posed by a high-resolution DVD-based
audio carrier (about which you may read in Daniel Sweeney's accompanying article).
No more of this.
The notion behind the AAD, or, as we prefer, the DAD at 96/24 (okay, 20 something,
maybe), is that more than 98 percent of the recordings made in the past 40 years or
so were intended to be heard through a two-channel stereo set-up and nearly as many
of today's home stereo systems are plain ole two-channel. Hence we have an archival
medium (though some would argue 192 is better) for the Golden Age classics and a
recording technique said to better 30 ips analogue technology- the faint chortling in
the background is the sound of those who know that multichannel sound will have to
be data-compressed and thus sonically compromised.
Thus, we see the High End's young turks lining up behind the 96/24 DAD, including
Ayre Acoustics, Bel Canto Design, Conrad-Johnson, Muse Electronics, Resolution
Audio and Theta Digital. Ditto for Chesky Records.
Dangerously close to press time, we finally laid hands on a batch of players, most of
them fresh out of prototyping and few really in production (we're talking early April).
We got in house a prototype of the Ayre Acoustics effort, the Muse system, the
Resolution audio system and we received the Theta 96/24 D-to-A unit, as well as the
Elgar. The four Classic discs:
1.Rachmaninoff's Symphonic Dances, taken from a quarter-inch two-track
30 ips per second master made originally made by David Hancock in 1967.
This is the famous, in audio circles, Johanos performance with the Dallas
symphony that first saw light on the Turnabout label.
2.Pulse: classical percussion-recorded by Paul Goodman in 1983 on
quarter-inch two-track tape at 15 ips.
3.1957: Red Rodney's rare jazz album, recorded by Rudy Van Gelder in
1957, again two-track, quarter-inch 15 ips.
4.A Time Remembered, jazz from Art Davis (and friends), recorded in 1995
by Rik Pekkonen on quarter-inch, two-track analogue running at 30 ips.
And there is a sampler disc or so (one hush-hush, not for public
consumption, with you-know-who conducting the 1812 Overture). Also, I
had the briefest encounter with a Pioneer disc, piano and singers in classical
repertoire, much beloved by our very own Dan Schwartz, one which I'm
not going to discuss here.
Conclusions? Well, none of the material exactly gives the new technology the
showcase it may well deserve. Two are recorded at 15 ips and on quarter-track tape.
One of the 30 ips tapes, the interesting Rachaminoff, is 31 years old; the other 30 ips
tape is of a jazz combo, and that simply isn't, harmonically, the last word in
symphonic or even stadium rock complexity.
It certainly would not be fair to attempt an evaluation of the players (all of which will
show video in a home theater setup and, just maybe, all of which could turn the
hometheater crowd on its sonic ear - I don't know, I didn't have time to test the audio
electronics on the home theater system). The Ayre was fresh out of Charles Hansen's
hands; the Muse and Resolution Audio units were two of the four in existence.
Given these conditions, I proceed to wade in the water:
At its best, with the Theta 96/24 decoder and the Ayre-modified Pioneer 505 deck, the
Rachmaninoff disc had astonishing dynamic range, and I mean astonishing. As well
as an extraordinary degree of slam. (We were using the Genesis APM speaker
system and the Melos Model 1000 monoblocks with 400 watts, per side, of push-put/
triode power.) What this disc did, under this singular circumstance, was reproduce
both bass and high-frequency transients in a way you're never going to hear out of a
conventional LP playback system.
None of the other players/DAC combinations exhibited the startling dynamics and
sheer impact of the Resolution/Theta, including Resolution's own DAC (but what the
hey, it's a prototype, guys - the Theta looks, feels and acts like a finished product,
which it probably is).
And we tried all sorts of combinations with the Hobson discs. We used the Von
Schweikert VR-8s with the Audio Research Reference 600s, we used the Viva
amplifiers with the InnerSound Eros hybrid electrostatic. And we came back, again
and again over a three-week period, to the units we had. Interestingly enough, the
discs, particularly the Rachmaninoff, didn't sound horrid as I might have expected from
the solo Pioneer 505. In fact, it sounded pretty good, which led me to believe that
potentially the strengths of this software might, like the Mercurys and RCAs of old,
shine through even mediocre IC-oriented electronics.
I had high hopes that the Chesky 96/24 disc would be ready in time for this issue and
that our order of the Pioneer 96/24 discs from Japan would reach us before the
last-minute inflexible deadline. I would and will be particularly interested to see how
other discs from other manufacturers sound when played back through this, the first
generation of 96/24 players and decoders. I even had hopes that Hobson might push
along his second-generation issues of DADs, so that I might have more material at my
disposal in order to get a better picture of just how good this new medium can be.
I can say this: none of the discs we played (and this was true no matter what the
playback gear) sounded particularly digital. Indeed, it was difficult, if not impossible, to
detect any digital artifacts at all, although I thought there was a distressing lack of hall
ambience in the 1812 test cut and throughout the Rachmaninoff (keeping in mind that
the auditorium in Dallas where the recording was made is both small and dryish).
Transparent it was, at least in the forefront of the stage. Layered depth, there wasn't.
And how massed strings will turn out to sound is strictly a matter of conjecture, when
and if, say, Hobson lets out the classic Vox/Ravel set from the old Minneapolis
orchestra.
What I am saying, in effect there wasn't really enough harmonic information on these
discs for me to divine the potential in any realistic way. I've no doubt, putting some
disparate pieces of the puzzle together, that 96/24 will wipe conventional 44/16, all the
more so when engineers and designers have enough experience with the system to
bring out its full potential (remember, it was more than a decade before the
technocrats got 44/16 to begin to reveal what it could, at best, do). No need to say
that I vigorously disagree with Hobson's choice (and it is that) of the software released
- he seemed to be trying to please irreconcilable sides of the audiophile community. If
it had been up to me, I would have picked some s.o.b. Blockbusters and let 'er rip.
As it is, we but scratched the surface of what could be a potent and even revolutionary
force for better sound from recordings of musicks old and new, and could start the
scramble for better High End components to reveal all the nuances that might therein
be contained. HP
Addendum: At [past -PB] the very last moment, what should arrive but three "super
audio" discs? 1. a 96/24 sampler from Chesky; 2. the Pioneer disc, Songs My Mother
Taught Me; and 3. from John Marks Records, The Harry Allen Quartet The Pioneer
disc sent me screaming from the room. The Chesky sampler doesn't push the | limits
of reproduction much l farther than the best CDs l might. The John Marks jazz
recording has the sweet airiness of a great analogue LP and is the most conically
satisfactory of the lot. Moral: if 96/24 is to be sold as a super audio disc, it had better
sound perceptibly better than the best of the current 44/16 CDs or it will find a tough
row to hoe beyond the audiophile community.
Hobson's Choice: The Dawn of High Resolution Digital Audio?
"There are no trade offs," says Mike Hobson, president of Classic Records and for
now chief purveyor of music software hewing to the optional 96 kHz, 24-bit audio
standard for DVD-video. "With vinyl there were major trade offs in convenience, and
some in fidelity, while in CD there were serious compromises in fidelity. This is the
closest thing in a commercial format to the master tape, and, for once, comparisons
with vinyl reveal far more sonic drawbacks than strengths in the older format."
Bold words, echoed by others of the DVD video-audio contingent. "Ninety-six
kilohertz, 24 bits is a lot better," David Chesky, artistic director of Chesky Records,
agrees. "We've been mastering to this standard for two years, and I could never go
back to the lower rates. Now we'll be making it available to the consumer."
George Cardas, High End cable manufacturer and recordist, is similarly ebullient.
"Absolutely the best consumer format. It approaches the mastertape for the first
time."
Finally, Kevin Halverson, CEO of Muse Electronics and chief missionary in promoting
the new standard within the High End community, says: "You'll get decisively better
sound from these recordings on the cheapest DVD player capable of playing them
than you'll obtain from the CD version on any available playback system. Every High
End recording label has expressed interest."
The positive spin on high-resolution digital is, if not universal, extremely widespread,
and is even heard in such unlikely pieces as the corporate board rooms of Samsung,
which favors an even higher standard of 192 kHz, 24 bits. Indeed, many see the new
standards revitalizing the listening experience, and coincidentally, the languishing
music-oriented audio business. But having heard similarly emphatic testimonials for
past enhancements in digital audio, such as four-times oversampling/true 16-bit,
single-bit conversion, Super-Bit Mapping, and, most recently, HDCD, I was not
prepared to take such statements at face value. Still, I was impressed by the passion
and sincerity of the people I talked to and by the scope of expressed support for the
new format within the High End. While not universal, it is still considerably more
extensive than that associated with older innovations, with the possible exception of
HDCD.
I must point out, however, that the strong support for an enhanced digital audio
standard is not quite the same thing as specific support for 96 kHz 24 bit. The current
situation in the larger music industry concerning a high-resolution consumer digital
audio disc is extremely volatile, and a format war appears to be underway. The course
of that war is apt to diverge from that of past conflicts because of some fundamental
changes in audio technology that change the very rules of warfare. Analysis rather
than partisanship should engage the energies of both the consumer and the
manufacturers today.
Of Facts and Factions
As anyone interested in High End audio is now aware, a small cabal comprised of
Conrad-Johnson, Classic Records, Muse, Chesky Records, Bel Canto, Resolution
Audio, Cardas and Ayre Acoustics recently announced support for the optional
high-resolution audio standard for DVD-video, which is set at a 96 kHz sampling rate
and 24 bits of resolution (see Robert Greene's article, this issue, explaining the
technology), and has been since the end of 1996 when an overall video standard was
adopted by the DVD consortium. The announcement was made at the January 1998
Consumer Electronics Show in Las Vegas. A well-attended press conference was
held there by the High End Group, who distributed a number of selections from the
Classic catalogue on DVD, showcasing the new standard and providing a basis for
comparing the sound quality with that of prior vinyl and CD releases of identical
material.
"Let's make It clear that upon the advent of digital, the sound
was as I described It horrid and godawful. That is not my
current position. ... I may have led the charge against the
horrors of early digital, but I have been forthright on its speedy
(thanks to the High End's evolving) evolution." -HP
These were not the first 96 kHz/24 bit recordings available on DVD. In February 1997,
Pioneer Video released a couple of classical music videos so recorded. But in a real
sense, the CES announcement marked a beginning, because what was initially an
afterthought was here being seriously positioned as the de facto standard for all
perfectionist audio manufacturers to follow.
According to Mike Hobson, Classic itself will release at least 15 titles in support of the
new standard and possibly more by year end. A couple of Classic CDs are even now
coming on the market, and Chesky's Sara K is available in Europe. Chesky has
committed to several more, Cardas Recordings to a few, and Golden String to one.
Beyond these few labels, no others have definitely committed, rumors to the contrary
notwithstanding.
In playback hardware, Theta Digital offers a 96 kHz, quasi-24 bit option on its
Generation V - a DAC - while Thor Audio has been selling a 96 kHz 20-bit DAC for
several months. Muse is reportedly dose to shipping both a DVD transport and a
DAC. Resolution Audio, Ayre Acoustics, Bel Canto, Conrad-Johnson, Audio Logic and
Arcam (the last two are not part of the original group) have promised D-to-A converters
or
transports (or both) in the near future, and several other outboard D-to-A converter
manufacturers are contemplating such an option. But for now the hardware choices
are as few as those for programming. One can, it is true, resort to professional
equipment from Nagra, dSC, and others, but according to Alistair Roxbury of
Enlightened Audio Design, the standard Sony/Philips digital interface will not permit
docking with such converters without modifications. One can also utilize the built-in
DACs in a number of current DVD-video machines from Toshiba and Pioneer (by no
means all DVD players have suitable DACs, and those that do not convert the 96 kHz
signal to a 48 kHz output), but obviously no audiophile pretensions are made for these
by the manufacturer, and listeners accustomed to pedigreed hardware might be
inclined to hesitate. So at this point, probably only the most dauntless of early
adopters will take the plunge, while more cautious souls will hearken to the intentions
of their favorite marques.
There, unhappily, they will find caution more them equal to their own While I could not
interview every manufacturer of specialty digital products, I did attempt to construct a
sample. Among the hardware companies, I spoke with representatives of Accuphase,
A/D/S-Museatex, Boulder, Enlightened Audio Design, Sony, Spectral, Timbre and
Wadia, as well as the acknowledged supporters of 96/24. On the software side, I
contacted Acoustic Sounds, AudioQuest, DMP, MoA, Mobile Fidelity, Reference
Recordings and Telarc, in addition to Classic, Cardas and Chesky. I also contacted
recording engineers of my acquaintance to attempt to ascertain the level of major label
interest. Interestingly, a number companies I attempted to contact did not respond to
repeated phone calls, perhaps because the subject has become so sensitive. The
reactions of the fence-sitters in the survey are interesting. Although generally
endorsing the notion of a higher digital standard, the spokespeople for the
nonparticipating specialty audio companies tended to express apprehension at the
immediate implications of adopting such a standard. All sorts of reservations were
cited, both technical and market-oriented, and several people frankly stated that
allocation of resources to the new standard would put their companies at extreme
risk.
Such fears may well be justified. Kevin Halverson told me that he was betting his
company on the acceptance of the format, having devoted more than a million dollars
for R & D m support of the new standards. I should also point out that some
companies contacted, namely, Accuphase, Museautex, DMP and, of course, Sony
and Marantz, are frankly opposed to 96/24. These companies have already embraced
the Sony/Philips Super Audio Disc, described below, and view the activities of
Hobson, Halverson, et al. with dismay. Several other companies have indicated that
they will go the way the wind blows. "We will support the dominant standard, and only
the dominant standard," says Steve Huntley, National Sales Manager of Wadia,
speaking for many in the industry.
Mixing It Up
Any new format has to start somewhere, and the limited availability of program
material and playback equipment might be expected to take care of itself if the format
gains acceptance. By the far most formidable competitor is garden-variety Philips Red
Book CD. Though music sales of CD slumped in 1997, sales slumped in cassettes as
well, and few music industry analysts detect any widespread consumer
dissatisfaction with the CD format itself, which in truth has increased its market share
and total sales volume every year since its introduction until 1997.
In contrast, the LP had been a declining format for years before the record industry as
a whole decided to replace it with the compact disc. Then too, there is nothing like the
consumer electronics industry consensus on how to market the new audio disc that
marked the introduction of the CD - again in sharp contrast to the highly focused
marketing blitz of the old Compact Disc Group. Some suggest that multi-channel is
the chief selling point, some argue for enhanced resolution within a two-channel
platform, and some for a combination of the two. Everyone has a theory, but no one
has much market research to back it up. Certainly, multichannel music programming
m the DTS and Dolby Digital formats, discussed briefly below, has not enjoyed much
success, but then how much success will enhanced resolution enjoy? No one really
knows.
Now, one might assume that, in our own small marketplace, an advanced resolution
disc would quickly gain allegiance. And perhaps it will, but unless it can spread
beyond that market and into a larger population of nonaudiophile music listeners, the
future of the new format must remain in doubt. Some have speculated that the majors
might support such a standard as a limited release, premium price format - "sort of
like gold CDs," suggests Charles Hansen of Ayre Acoustics - and maybe they will.
But maybe they won't.
According to legendary mastering engineer Bob Ludwig, "The majors [apart from Sony
Music] have no interest in the new disc formats."
According to Jordan Rost, Senior Vice President of New Technology for the Warner
Music Group, Warner is strongly committed to providing DVD-audio software. But m a
mass market setting where hometheater systems, mini-racks, computer speakers,
boom boxes, personal stereos, and car stereos will be the principal delivery systems,
Rost, and presumably his colleagues, see DVD-audio largely as an epi-phenomenon
accompanying the general transition to higher density discs in all information storage
applications, and not as any particular stimulus to that small sect, the High End. Rost
told me that higher resolution "is just numbers as far as most people are concerned,
but they lend credibility to the format. What creates excitement is the multiple
channels. Frankly we're not after the audiophile early adopter."
Of course, in the mainstream of audio journalists who write for Stereo Review and the
trade publications, the utter transparency of 44.1 kHz 16-bit has never been in doubt.
And since those individuals are the ones consulted by writers for general news
publications, it is their views that are communicated to the public. Thus any new
format must contend against the received wisdom that the Red Book is perfect.
Obviously, such mainstream audio writers cannot readily endorse an advanced
resolution disc without endangering their credibility, and many are sorely aggrieved
that Sony and Philips would presume to undercut the earlier stated position that CD
represents "perfect sound forever." The prospect of a major manufacturer push for a
new higher standard would be for such writers a sort of Consumer Electronics Bay of
Pigs, where they, like those intrepid reactionaries of yore, suddenly find themselves
abandoned on the beach, bereft of superpower support. Indeed, I wouldn't be surprised
to see very soon a double-blind listening test report in the pages of Stereo Review
"proving" that high resolution is indistinguishable from Red Book.
But even assuming that most High End as well as mass market consumers and the
manufacturers who serve them are perfectly satisfied with CD, mightn't a segment of
the High End be enough to launch an improved digital format? After all, vinyl survives
with only a tiny subset of the High End supporting it. The answer, I suppose, depends
upon the size of the segment, and right now the segment of committed software
manufacturers is critically limited. Among the largest audiophile labels, Mobile
Fidelity, Reference Recordings and Acoustic Sounds are holding back, while Telarc,
the giant of our industry, has expressed interest, but hasn't settled on whether to
support DVD or the Sony/ Philips Super Audio Disc discussed below.
So, dearly, CD is still entrenched on all fronts, and record labels are not about to retire
it any time soon. That means chat the new disc, whatever form or forms it will take,
must make its way along the fringes, at least through its infancy - rather like a
hatchling crocodile keeping to the reeds and waiting for the time when it can be
formidable.
Unfortunately those fringes will soon be crowded with competing life forms. We
already have DTS and Dolby Digital audio discs. The former began as a specialized
form, a&Bring in part to the general Red Book standard, but carrying multiple streams
of compressed PCM audio, while more recent releases have been contrived as
DVD-video discs without the video track. The few Dolby Digital music-only discs follow
that form as well. Market performance of either format has not been spectacular,
particularly in the case of the audio-only Dolby Digital, but you never know. But
lossy-compressed formats such as Dolby Digital and DTS are not properly our focus
here. Rather, we are concerned with formats that hew to a higher standard of fidelity
for individual channels than does Red Book CD. And here we find at least four
competing standards, whose adherents will comprise the principal combatants in the
format wars to follow.
The Rivals
Amidst the proposals for new standards, two are of paramount importance, those of
Toshiba, which have gained broad acceptance within the DVD-audio working group,
and those of Sony and Philips, which have set those two companies at odds with
other members of the group. There are two other standards vying for acceptance m
the marketplace, and m the interest thoroughness, I shall cover all four.
The first high-resolution audio standard, the only one that has really been set, is that
for DVD-video, and though it is only an option, it is an option that any hardware or
software manufacturer can choose. It's here for anyone who wants to use it, and
Messrs. Hobson and company dearly want to use it.
The second standard is the as-yet-unannounced DVD-audio disc standard, which will
probably encompass the first standard but also allow an extraordinary range of
options. Recent announcements made by Working Group Four, within the DVD
Group, confirm that software producers will be free to select, independently, resolution
and bandwidth for individual audio channels. One can have any multiple of 44.1 kHz up
to 176.4 and any multiple of 48 kHz up to 192. Choice of 16, 20 and 24 bits will also
be available at any given bandwidth. The producer may also choose a number of
channels from one to six, and may assign different bit levels and bandwidths to each
of the channels, though, according to reports, 24 bits and 96 kHz or higher cannot be
allocated to all six channels because of the resultant limitations in playing time.
Provisions for folding down multiple channels into two will also be included.
No one has ever before introduced a consumer audio format adhering to such an open
standard. How will the music industry respond? Recording engineers and producers
are known to dislike multiple-release mixes, and in terms of minimalist recordings, a
fixed-channel allocation is virtually a given (Ambisonics is a special case I won't
explore here), but conceivably the benefits two-, three-, four-, five- and six-channel
recordings at various levels of resolution could be espoused by various individual
producers.
So what are people going to make of all these choices? Most of the folks promoting
96/24 two-channel are betting that multi-channel won't catch on, at least not in the
High End. Steven Lee of Canoris, distributor of the Nagra and dSC professional lines,
expressed the view, based upon some well publicized blind listening tests held at the
last AES convention, that a 192 kHz sampling rate is really needed to achieve
transparency. Such a rate is permitted in the standard, but is not generally supported
in the High End, except by the proponents of the Sony/Philips DSD system, who
actually go one better with 200 kHz. Most people advocating anything are going with
96 kHz or 88 kHz.
Similar controversies swirl about the proposed bit rates. From what I've been told by a
number of engineers, DACs of true 24-bit resolution are not yet available, and certainly
not on any consumer machine. Thus, playback obtainable today is at the 20-bit level
maximally. This, to be sure, is a considerable improvement over 16-bit CD, 20-bit
representing 120 dB of dynamic range altogether. Full 24-bit improves dynamic range
to 144 dB, though that is well beyond the range of the quietest analogue electronics
made today and thus arguably is not perceptible.
Interestingly, some in the High End, including Bob Stuart (see his article in Audio,
March 1998), argue that lower-bit recordings, as low as 14 bits in fact, will, with
suitable noise-shaping techniques, achieve resolution and transparent equivalent to
recordings made at the higher bit rates. Remember, the standard permits the options
of 16, 20 and 24 bits, and so the manufacturers are free to go with any of them. And
they probably will.
Here I should point out that within these myriad configuration choices, a couple of
options of really fundamental importance exist, options having to do not just with bit
rate, channel number and sampling rate, but with core technology. And while likely to
be accommodated within the DVD-audio standard, these basic technology choices
are so axial in their significance, they constitute rival formats, veritable jokers within
the DVD pack. One of these is the Sony/Philips DSD (Direct Stream Digital)
modulation system; the other is HDCD, which is essentially a group of digital signal
processing strategies. Since DSD may be the joker that is also the trump card, we'll
look at that first.
DSD is an advanced modulation scheme cooked up by Sony and Philips that is not
PCM digital at all - in other words, it doesn't have much in common with what's on CD
today. Instead of representing audio information by digital words of varying values
created at minimally twice the rate of the highest frequency of interest, DSD is a
single-bit system related to the single-bit conversion schemes used in many D-to-A
converters today. Now, the single-bit converters used in existing PC players are
actually double converters in that they first convert a pulse code modulation (PCM)
signal into either a pulse-width modulation or a pulse-density modulation, which is
then smoothed into an analogue waveform by means of a relatively simple analogue
output filter. But in a DSD recording/playback system, there's no PCM at all.. Instead,
analogue waveforms are represented by the varying density with which pulses of a
single amplitude are distributed in a given time interval. It's bitstream all the way, with
no conversion and no digital words.
Interestingly, on a scope, a bitstream at a high enough sampling frequency looks
remarkably like an analogue waveform. There's no ragged staircase to deconstruct,
and in fact no D-to-A converter is needed at all, just a suitable analogue output filter.
And no brick wall filter is needed at input, either, because a single-pole input filter at
100 kHz will eliminate aliasing problems. Conceptually, the system is extremely
appealing because it is extremely simple. It's digital reduced to human terms rather
than machine logic.
Actually, there's a lot more going on in DSD than just the recording and transmission
of a bitstream, because a lot of signal processing is needed to allow the system to
operate optimally on a real-world optical disc. But the basic conversion operations
ready are extraordinarily simple and elegant. It's almost a kind of pointillistic analogue
recording. But rather than being universally applauded for their ingenuity in creating
such a system, the Sony/Philips DSD project team has aroused the most intense
controversy I have ever witnessed in the specialty audio business. Indeed, the
audiophile labels are divided into armed camps on the subject of DSD versus PCM,
with most favoring PCM.
"DSD is a joke," declares George Cardas, who admits he has never heard the
system. "Sony and Philips are farting in the wind."
"Nobody thinks it has merit," says Michael Ritter, president of Pacific Microsonics.
"Even the people at Sony know it's a loser."
Bob Stuart of Meridian and Malcolm Hawksford, both charter members of the ARA
(Acoustic Renaissance for Audio) group, have been attacking bitstream for years in
technical publications and Internet squibs, and recently dSC joined them with a
technical expose published in the March edition of Stereophile. "The Brits hate
bitstream," confided one source.
Charles Hansen, President of Ayre Acoustics, suggested that some of the Sony
demonstrations may have gilded the lily somewhat by employing a better recording or
playback system for the DSD than for the 96k. "They're desperate to maintain their
CD licensing fees," he said, repeating a charge that I heard incessantly in the course
of my interviews.
Such sentiments are not new. Two years ago when DSD was first announced, I was
told by Kathy Gornick of Thiel, Mike Moffat of Angstrom and Mike Elliott of the
now-defunct Counterpoint, that DSD sounded dreadful, and that everyone in the High
End was united against it. Interestingly, none of these people claimed to have the
heard system at the time; the source of their information appeared to be Bob Stuart,
who then, more generally than now, was accepted as the High End leader in industry
discussions about a new standard.
Whatever the source of the DSD canards, the validity of such damning assessments
are open to question, and indeed are not universal in the High End despite the
statements of various self-appointed industry spokesmen. Tom Jung of DMP Records,
a strong proponent of the DSD system, told me he believed it offers the best digital
reproduction currently available. Mike Bishop of Telarc and John Wood of Mobile
Fidelity also spoke highly of the transparency of DSD. Numerous other recording
industry heavyweights, such as Bruce Swedien and Bob Clearmountain, have also
commented favorably on the sound quality of DSD. Is it better than 96/24?
"Unquestionably," says Tom Jung. "I've heard them side by side."
I heard the system myself and was impressed. I heard - nothing! Transparency
personified.
Is it as good or better than 192 kHz 24-bit PCM? "Absolutely, says Ed Meitner of
Museatex Audio and A/D/S, who has contracted with Sony to work on the system.
"Higher sampling rates and word lengths fail to eliminate several fundamental
problems with PCM systems, such as phase reversals in the presence of most
significant bit errors, and the need for elaborate error correction schemes. PCM never
should have been adopted in the first place."
Bob Ludwig, on the other hand, is ambivalent. "PCM and bitstream both have their
strengths and weaknesses." Still, knowledgeable digital engineers have advanced
plausible technical arguments against the system with no obvious axe to grind.
David Rich, an engineer with Lucent Technologies and sometime writer for the Audio
Critic says, "I don't like the system at all. From the perspective of information theory,
it is massively inefficient, and it's extremely noisy. True, they can use noise shaping
to move the noise out of band, but it doesn't go away. They can also use lossless
compression to improve the efficiency, but then they're adding tremendous complexity
to a system whose virtue is supposed to be its simplicity." Rich, whose attacks on
High End manufacturers and other writers have earned him many enemies in specialty
audio, is yet generally respected for his technical expertise. His statements on DSD
summarize many critiques made by other commentators.
Leaving aside for the moment questions of either technical soundness or audio
quality, what bearing has DSD on the overall DVD-audio standard? According to
leaked reports, Sony and Philips, who are among the ten members of the DVD-audio
working group with voting rights, wanted DSD to be the sole standard for
high-resolution audio on the proposed new format. The other members, who have been
reluctantly paying compact disc royalties to Sony and Philips for years, were not
persuaded. But according to a recent DVD WG-4 press release, the other members
have allowed DSD as an option. Meanwhile, Sony and Philips have jointly announced
that they will support a proprietary format called the Super Audio Compact Disc that
may not conform to the overall DVD spec and may
not even be playable on DVD transports. Shades of Beta? Yes, according to Sony
detractors. But in several respects, Sony's position appears to be far stronger here
than in the video cassette wars.
First of all, the Sony/Philips proposal calls for a dual-layer disc with a surface layer
consisting of standard CD audio, playable on any machine ever manufactured, for full
backward compatibility. That probably won't be true of DVD-audio, at least not at first,
which raises the specter of double inventory. Second, Sony and Philips together
command about a third of the recorded material currently in release. Of the other DVD
group members, only Warner has a significant catalog of its own. Third, the Super
Audio Compact Disc offers six channels at full resolution along with a full-resolution
two-channel mix and a two-channel CD track, though the six-channel mix can be
optimal. Finally, and this is probably the strongest factor favoring Sony and Philips,
they are the only major music labels aside from Warner with a dear commitment to a
new audio disc format. DVD-audio is essentially a hardware manufacturers' proposal,
not a music industry proposal.
Still, strong arguments have been voiced against the feasibility of the proposal.
"Double-layer discs have proven very expensive to manufacture so far," notes Kevin
Halverson of Muse. "Can Sony do it cheaper? Maybe they'll try to buy the market by
selling discs at a loss, but I'm doubtful"
Another question is whether the Super Audio Disc will be playable on a DVD
--
天涯何处无芳草,活人能让!憋死?
天涯处处皆芳草,活人让!憋死了。
命苦不怨中国政府,点儿背就怪美国社会.
※ 来源:·BBS 水木清华站 bbs.net.tsinghua.edu.cn·[FROM: 202.96.151.222]
--
☆ 来源:．哈工大紫丁香 bbs.hit.edu.cn．[FROM: crazybaby.bbs@smth.o]
Powered by KBS BBS 2.0 (http://dev.kcn.cn)
页面执行时间：604.683毫秒