Adventures in Early Digital Recording
Throughout the 1970s, digital audio
was mentioned in the hi-fi press every so often, usually in connection with its use in
Britain for ferrying FM signals around the country, but it all seemed a bit farfetched.
After all, when everything was working well, the analog audio systems of the day produced
excellent sound, so all the technical jiggery-pokery involved in digital seemed like a lot
of trouble for very little gain.
All that changed, as far as I was concerned, when I
wandered into the exhibit of a manufacturer of digital recorders at a hi-fi trade show. At
first nothing much seemed to be happening, other than people standing around chatting.
Then, all of a sudden, the sound of a low-flying jet filled the room, at
only-just-below-ear-splitting level. Then silence. Then, maybe 30 seconds later, another
startling noise. Then silence again.
The company was demonstrating not only the dynamics and
realism of digital audio, but also its superb noise performance. The tape was running all
the time, but even if you put your ears right to the speakers during the quiet bits (a
risky act, under the circumstances), there was no hiss. Even the best open-reel
decks running at that level produced some noise. We were just used to tuning it out.
Another telling moment came a couple of years later when a
colleague had managed to borrow one of the first VCR-based digital recording systems, and
we took it home to audition it on familiar equipment. The only material we had to play on
it was one of those excruciating Japanese jazz trios (no vocals, fortunately), but it did
serve to illustrate perhaps the most dramatic of digital's virtues.
The moment we rolled the tape and began listening to the
piano, we knew we were hearing something exceptional. But we didn't know what. Eventually
we concluded that we were reacting to a complete absence of flutter. It seemed as though
we were always aware of this speed fluctuation in the analog recordings of the day, even
if we might not realize it; when it was gone, that was immediately noticeable.
In those days, it seemed unlikely that any medium using
digital sound would be available to ordinary audiophiles anytime soon. But that proved to
be wrong: Philips showed a prototype CD before the end of the decade, and both discs and
players were on the shelves by 1982.
Audio enthusiasts are a conservative lot, some of them
anyway, and quite a few were appalled by the new technology. The "digital sound"
was edgy, they said. It lacked detail; it couldn't reproduce reverberations properly; it
sounded lousy. And in truth, a lot of the early CDs did sound pretty bad.
But now most would agree that the early problems had little
to do with the medium itself. Rather, reissuing material originally designed for vinyl
release was risky unless care was taken to undo the various tricks the engineers had used
to overcome what they perceived as the shortcomings of the LP. Even the most virulent
digiphobes conceded that there were some excellent compact discs, and the pro-digital
forces argued that if even one good CD could be made, there was nothing inherently wrong
with the medium.
Once the CD had become established, it was assumed that
some form of digital recording medium would be developed, to function in relation to the
compact disc as the cassette had to the LP. The existing VCR-based systems, while they
certainly worked well enough, were too cumbersome to be the answer. And so, the hardware
and record companies began a decade-long dance to come up with an acceptable digital
medium.
It seemed like a fairly simple challenge, and it was. The
main drawback in using VCRs for digital audio was that they had more capacity than was
necessary. By scaling things down to the point where they could handle only what was
necessary for audio, much smaller machines and tapes were possible.
And so was born the digital audio tape (DAT), resulting
from the joint efforts of the major Japanese audio manufacturers to come up with a single
standard right from the start. In fact, they came up with two: the R-DAT, which used a
rotary spinning head like that in a VCR; and the S-DAT, which had a stationary head and
much simpler mechanism, but which required a prohibitively expensive tape.
The industry adopted R-DAT and began making machines, but
the format was immediately embroiled in legal hassles at the hands of the record
companies, who were already smarting from what they saw as huge losses caused by consumers
copying their musical property onto ordinary cassettes. They were determined not to
countenance something that would allow Joe Public to make perfect copies. At least
a cassette recording deteriorated considerably in sound quality from generation to
generation. But in theory, you could make an endless chain of digital recordings without
hearing any difference.
The DAT format did become a professional standard, and
ultimately a few machines appeared for consumer use. But as the sun began to set on DAT a
few years ago, there was a ray of hope for those yearning for a digital recording medium
in those pre-MP3, pre-CD-R days. Ever-ingenious Philips came up with the Digital Compact
Cassette (DCC), which had the advantage that you could play your old analog tapes in the
new machines, along with digital material.
The only real compromise in the new medium was that it used
a data-reduction scheme to fit a digital signal onto a tape with the same physical
characteristics as a normal cassette. Audio observers worried that, with only a quarter of
the original data recorded on the tape, there might be audible artifacts. But extensive
demonstrations reassured virtually everyone that this was an audibly transparent system.
The unexpected announcement of DCC reputedly forced Sony to
announce its new format, MiniDisc, prematurely. And most early reviews of the system,
including my own, found the sound inferior to a CD or DAT recording (or, for that matter,
DCC). Sony began claiming that this wasn't meant to be the equal of those systems,
but was built for convenience and portability. Later reports did suggest that they had
their algorithms together, however. (MiniDisc, incidentally, uses a 5:1 reduction system
called ATRAC, which early wags noticed was pronounced exactly like "eight
track.")
A number of years ago, in the dying days of DAT, the one
digital recording system that played it straight -- recorded all the data -- some fellow
audio journalists and I decided to do some listening tests to see what sort of sonic
compromises we would have to put up with in using one of the data-reduced systems. We
collected one of the last remaining DAT models (from Sony), a DCC deck from Technics, and
a MiniDisc player from Sanyo (actually a glorified boombox, but with line inputs and
outputs so we could access just the MD section). The purpose was to make A-B comparisons
among these three recorders, and also to compare the sound to the best analog had to
offer.
To make our lives easier, we decided that our source
material would be a DAT tape rather than straight CDs, because that would allow us to
create a standard program consisting of a number of different cuts of music, rather than
having to flip individual CDs about for each test. To make this valid, however, we had to
assure ourselves that there was no sonic difference between DAT and CD. We were fortunate
to have a sampler issued in both DAT and compact disc, and were able to switch back and
forth between the two. We also dubbed a couple of cuts from other CDs onto the DAT and
compared their sounds. In no case could we hear the slightest difference -- just as we had
expected.
Once we had created the listening program on DAT, we simply
copied it to DCC and MiniDisc by the best method available (digital in the case of DCC,
analog for MiniDisc). Then, using a countdown at the beginning of the recording to get
things synchronized, and adjusting an external network to make sure levels were exactly
even, we played the recordings, switching back and forth among them. DAT was compared
individually to DCC and MiniDisc, and then to other formats.
It was no real surprise that our listening panel couldn't
hear any difference whatsoever between DAT and DCC -- that had been our experience in
earlier sessions, although none of those had been this intensive. We were more interested
-- and pleased -- to find out that the MiniDisc in its later incarnation was just as good
a performer. All three digital formats seemed to be identical in sound; if there were
subtle things we missed, even in our direct A-B tests, they would be so small as to be
immaterial when it came to choosing one format over another.
Part of our test was to compare the digital media to analog
recording, in an attempt to interpret the specific nature of digital's improvements (or
not). Certainly, a digital tape and an analog cassette made with an off-the-shelf deck and
off-the-shelf tape would bear no comparison.
But one of our most revealing tests put DAT and the others
up against a truly optimized analog cassette. We chose the most advanced tape (metal), the
most advanced noise reduction (Dolby S, with HX Pro in the record mode), and a very
effective deck-to-tape matching system (Pioneer's Super Auto BLE). With everything going
for it, the analog recording was indistinguishable from any of the digital formats. Tape
companies have been saying that for years, of course, but we've tended to take that as a
bit of typical hyperbole.
It's not, or at least not always. It's not enough to use a
good machine and a high-quality tape; the matching between the two is the critical factor,
and a machine that doesn't offer this facility is only likely to make recordings this good
by blind chance.
And no amount of electronic wizardry can help when it comes
to speed problems. We used two cassette decks in our extended listening, one with
excellent flutter, the other with only average performance. The latter always tipped its
hand in listening tests.
Our main conclusion was that digital systems can do
effortlessly what the older analog technology can only do with difficulty (and expense).
But when it's done right, analog recording can be staggeringly good.
But as disc-based digital media gain acceptance, analog's
fate is probably sealed.
...Ian G. Masters
ian@mastersonaudio.com
|
