Miscellaneous

Introduced nearly eight years ago as the first recordable digital format for consumers, DAT both failed to appeal to its target market and was blocked from formally entering the USA for four years by the RIAA, who feared for the copyright of its members' recordings. However, the DAT medium was enthusiastically snapped up by professionals and semiprofessionals, who found its combination of reliability, CD-compatible signal format, and editing ease ideal for mastering. Sony currently offers a small range of consumer models, from the diminutive Walkman-sized TCD-D-7 DATMan to four-head cassette-deck–sized machines for the so-called "prosumer": amateur recordists who make pin money taping local concerts. The DTC-2000ES is Sony's latest entry (footnote 1).

The DTC-2000ES is the first consumer deck to offer Super Bit Mapping—Sony's proprietary noise-shaping system that claims to deliver 20-bit accuracy in the midrange and low treble by moving quantization noise from the ear's most sensitive frequency range to a higher one. Sony also modestly admits that the 2000ES is Sony's "statement" on consumer A/D and D/A conversion, about which more later.

The recorder features four heads, two of which are dedicated "assurance" heads that play back the just-recorded signal from the tape while a recording is being made. This is an invaluable feature for a professional, who's likely to be taping 100 musicians at union recording rates and needs to be certain he's actually getting usable tape from them. If he hears a dropout, he can say, "Hold it, we have a problem," and do a retake on the spot. But record assurance is of much less value to an amateur (or prosumer)—he'll probably be taping concert performances, and if he hears a glitch from the tape, there's not a damned thing he can do but bite his tongue and hope there won't be more. The extra heads add to the cost of the recorder; considering its target market, I don't think they're justified. I have never once encountered a playback glitch from DAT that wasn't in the original signal, and neither has anyone else I know.

The 2000ES is one of the few available consumer recorders that has microphone inputs as well as the usual complement of line-level Ins and Outs. It should be noted, though, that the mike inputs are for unbalanced mikes only; they will not handle a phantom-powered condenser mike, for instance, and could well be damaged by it. (Such mikes must be used with balanced/unbalanced input transformers.) You may think it ludicrous that I would suggest partnering a pair of expensive professional condenser mikes with a consumer tape deck, but the idea isn't at all absurd. People who have used top-quality condensers even with modest compact cassette decks have been astonished at how good their tapes sounded. DATs may be digital, with all that that implies, good and bad, but make no mistake: If you plan on doing live recording, microphone quality is at least as important as the quality of your deck. Digital's lack of treble softening puts formidable demands on a microphone's treble smoothness, and that smoothness doesn't come cheap.

The input lineup also includes a single pair of analog line inputs, as well as two switch-selectable TosLink digitals and a coaxial digital input. Outputs include one fixed-level analog pair and a TosLink and coax digital.

The record volume-control is a ganged, clutch-coupled pair that tracked surprisingly well all the way down to the bottom, allowing me to do perfectly balanced fadeouts down to silence. Other controls include On/Off buttons for SBM and pre-emphasis, and a switch to select 48 or 44.1kHz sampling, or a 32kHz mode that doubles the recording time.

Both the '2000's displays are backlit LCDs that are bright enough to view in a well-lit room, but not bright enough to sear your retinas when working in subdued light, such as in the wings of a concert hall. Everything you need to know is displayable—some of it gratuitously, some of it on demand by punching a button or two. You can even display the sampling frequency of an unknown tape you're playing—for whatever that's worth to you.

The record-level indicator has a useful feature called Margin, which keeps a numeric tally of how closely the record level has approached maximum since the last time you reset the Margin. It's like a Peak Hold function, except that you get a decibel value instead of a stationary segment on the bargraph. Either way, you get a frozen record of what happened when that loud part came along while you were staring lasciviously at the harpist, so you can choose to edge the level down a bit before the next one comes along. Actually, a slight overload is of little consequence; all you'll notice is a subtle loss of attack speed. A severe overload, however, will cause a flat top that adds a click (but thankfully not a bang) to the peak.

Because DAT, like all digital recording systems, records in a series of discrete "frames," the 2000ES's minutes/seconds counter is precisely and repeatably accurate to within a second. The counter Mode selector allows display of absolute time (from the start of the tape), elapsed time of the current selection, remaining time to the end of the tape, and elapsed time from when you last punched Play or Record, regardless of where you started from.

Construction is excellent—well up to Sony's ES standard—and the unit looks as easy to service as anything this complicated could possibly be. Its 25-lb weight tells the power-supply story. (Unlike some high-priced decks, which shall remain nameless, the 2000ES doesn't embroider its own heft with a 25-lb cast-iron bottom plate.) Appearance and ergonomics are—well, Sony ES. Enough said.

Although there's a lot to learn about using this deck, making a tape with it is as easy as recording on a compact cassette deck. But it does take a little while to learn all the available functions well enough to be able to invoke them without peering at the button labels or digging out the instructions. And the functions all work flawlessly. The only problems I had were my own fault—like putting two Start IDs (see "Dis 'n' DAT" sidebar) too close together when I'd been warned by the instructions not to. (On one occasion, I had to re-record over that section of tape to get rid of them.)

Listening
Other equipment used included a Fosgate THX Home Theater amplification/loudspeaker system, an EAD 7000 A/D converter, an Arcam Delta 100 Dolby S cassette deck (reviewed last month, Vol.17 No.10), and a Sony CDP-X779ES CD player. Audio interconnects were 75 ohm homebrews; the main loudspeaker cables were AudioQuest Greens.

Listening straight through the 2000ES—line input to output—revealed no more difference than an extremely slight HF softening, and it was so slight that I'm still not certain it was there at all. Neither bypass testing nor prolonged listening to the in-circuit 2000ES confirmed my observation, so I would say that the deck's analog circuitry is, for all intents and purposes, beyond reproach. Going into the 2000ES through its mike inputs (with the source signals attenuated by 20dB) made no qualitative difference whatsoever, which suggests that the preamp stages are probably in-circuit all the time, their inputs simply padded-down for line-level sources.

I made a digital copy from a CD through a TosLink optical interconnect, took a long listen to it, and quit right there. If there was any difference at all, I couldn't hear it. (This makes it all the more difficult for me to understand how CDs can so often sound inferior to their digital originals. In the case of DAT, it really does seem as if bits is bits.)

For the analog/digital (and vice versa) tests involving the 2000ES's own digital converters, I used CDs and recent analog-mastered LPs (from Reference Recordings and Wilson Audio as signal sources)—the CDs because of their more extended low end and slightly better transient response, the LPs because they can't have contained any digital artifacts that might make them sound less different than their copies.

Comparing analog-copied DATs with their CD originals revealed a few small differences. The copies had identical frequency range and balance, a bit less depth, and a slight edge of dryness. Noise was not a consideration.

Overall, the similarity between good CDs and their analog copies to DAT was greater than between any original and its analog copy that I've ever heard. Dolby S cassettes copied (from CD) on the Arcam Delta 100 deck I reviewed last month sounded a little more "musical" than the 2000ES DATs, in that they were slightly sweeter than the originals; but the 2000ES's tapes were unmistakably more like the CD originals than were the S cassettes. Again, the major difference was that the DAT copies were slightly drier, and seemed to have a very subtle "zzz" riding on their treble content. DAT copies of LPs had even less of this, suggesting that Sony's A/D and D/A conversions were adding to something uniquely digital that was already present in the original CD sounds. In short, Sony's "statement" A/D and D/A converters do a very good job.

It's not, however, a statement of the art. While I didn't have an expensive outboard A/D converter to compare with, the Enlightened Audio Design DSP-7000 II D/A I had on hand was clearly better than the 2000ES's internal D/A. (It should have been—it costs more than the whole 2000ES.) The EAD sounded sweeter, more liquid, and—dare I say it—more like good analog.

What about the Super Bit Mapping? Well, kiddies, there is a difference, and the difference is in favor of SBM. If my hearing still went out to batsville and my speakers had a nice hi-fi treble rise in them, I might find non-SBM digital so obnoxious that I'd write a paragraph of paean about SBM. Neither is the case. Yes, SBM sounded perceptibly smoother and cleaner—more listenable, if you will—than the plain vanilla digital, but not by much. Is it worth the added expense, then? You bet your bippie it is. The ultimate quality of a high-end system depends on the sum of many small improvements, so none of them can ever be dismissed as "insignificant." And there's no question that SBM brings about an audible improvement.

JGH Concludes & Rails
This is probably as good an all-in-one DAT recorder/player as money can buy. Its D/A can be improved with an expensive outboard processor, as probably could its A/D, but I would be prepared to pit the 2000ES against any other recording machine that could be bought (or assembled) for under $10,000, and make book on the outcome. That this level of quality can be acquired for substantially less than $3000 is, I feel, a triumph of design and manufacturing.

But would I buy one for my own use? On the basis of sound alone, I wouldn't hesitate; but there's a yellowjacket in the ointment, and its name is SCMS.

SCMS (Serial Copy Management System) is a copy-protection scheme that automatically identifies any recording made from a digital source as a copy, and prevents any such copy from being digitally copied again (footnote 2). This means that you can make a digital copy of a CD, but you can't make a digital copy from that copy. You can make a digital copy from one of your own live recordings on DAT, but, again, you can't copy that copy (see "SCMS Scams" sidebar). For a "serious" amateur recordist like myself, that spells trouble.

For example: Suppose I make a recording of an entire concert season, and the performing group wants me to compile a "sampler" tape to give to its patrons. The only way to do this without some quality loss is by digital duping, which SCMS will allow—once. But that compilation tape is now branded $Bcopy. Suppose, then, that I need two DAT copies of the compilation—one for broadcast and the other to make analog cassettes from. Sorry—SCMS won't allow that. The choice is to either make analog copies, or spend another three hours reassembling all the excerpts from a half dozen or so tapes for each new DAT copy (footnote 3).

There are professional "black boxes" available that will quite legally strip the SCMS flag out of the bitstream—Digital Domain's FCN-1 Format Converter, for example, which costs around $600, or the Core Sound Digital Patch Panel—while it is rumored that some of the popular "jitterbugs" will do the same. However, "professional" DAT machines don't have SCMS protection; couldn't I just rent or borrow one of those to make additional copies? Well, I could, but the only digital connections to and from pro decks are via balanced AES/EBU ports. A simple adapter lead will get the data signal into the machine, but it's not a sure thing whether the pro deck's input receiver will recognize the 20dB lower signal level from the consumer coax interface. [I've had no problems making copies on a professional Panasonic 3700 recorder, driving one half of its balanced AES/EBU data input with the coaxial S/PDIF outputs of both my Aiwa portable DAT recorder and a Radio Shack CD-3400 CD player.—Ed.]

Why not just buy a couple of pro DATs and be done with it? Unfortunately, affordable pro is not where DAT's sonic state of the art is; consumer is where it's at, as exemplified by the 2000ES. The choice, then, would be to pay more for a pro machine that doesn't sound quite as good as the 2000ES but allows unlimited copying, or opting for the best sound and learning to live within the SCMS wall. Some choice!

I'll probably—grudgingly—buy the DTC-2000ES anyway when Sony's accounting department starts getting antsy, simply because it's such a good recorder and recording is my favorite hobby. But that doesn't mean I'll quit railing at SCMS.

---

Footnote 1: Since this was written, Sony has announced a two-headed machine for $1800.

Footnote 2: America's music industry, which viewed with horror the prospect of every Tom, Dick, and Eduardo being able to make "perfect" copies of CDs, rammed the SCMS mandate through Congress before it would allow any DAT machines into the country. (As far as I'm concerned, when an industry gets powerful enough to prohibit public access to an important new technology, it's too damned powerful for the public good.)

Footnote 3: Sony's $700 TCD-D7 DATMan is perfect for digital copying—once.