Sony DTC-60ES DAT Deck Review

Home decks for the digital audio tape (DAT) format have always been a rare breed. They have never been an endangered species, though, because professionals have enthusiastically embraced the format as the most cost-effective way of making CD-quality 16-bit digital master tapes. In fact, the audio data on many music CD’s passed through a DAT stage.

Sony, long a leader in the DAT format, makes three home models, the middle of which is tested here. All are suitable for making dubs of LP’s and digital “clone” copies of CD’s as well as professional-quality live or garage-studio recordings. With the $1,200 DTC-60ES, however, Sony has gone beyond the capabilities of most professional DAT machines: It is one of two Sony home decks that incorporate Super Bit Mapping, or SBM (the other is the $2,500 DTC-2000ES). Simply put, SBM enables the recording of a wider dynamic range than is theoretically possible with a standard 16-bit recording system. It does this by taking a better-than-16-bit signal from a high-resolution analog-to-digital converter and turning it into a 16-bit signal while shifting the irreducible, residual quantization noise away from where it would be most audible to where it would be least audible, above 15 kHz.

SBM is a recording-only process; SBM tapes are playable on any DAT machine. The noise and distortion reduction you obtain in playback of an SBM tape are proportional to the general quality of the digital playback system (more on this later). In the DTC-60ES, SBM is an analog-input-only operation and is turned on and off by a rather inconspicuous button to the upper right of the display window.

Besides SBM, the DTC-60ES has a host of useful features. Those directly affecting sound quality are 1-bit analog-to-digital (A/D) and digital-to-analog (D/A) converters and three selectable sampling frequencies for the analog input: 48 kHz (the DAT and professional standard), 44.1 kHz for CD and MD compatibility, and 32 kHz to provide a 4-hour long-play mode. Analog-input sampling rate is determined by a front-panel slide switch. Sampling rate is determined automatically for digital input signals.

The availability of 44.1-kHz sampling with analog signals is important for would-be record producers, since the resulting tapes will be directly transferable to a CD mastering system without going through sampling-rate conversion, a process that under even the best of mathematical circumstances adds some noise to the signal. There is no audible advantage to the 48-kHz rate, and I use 44.1 kHz for all my live recordings. (Lack of 44.1-kHz analog-input recording is the principal fault of Sony’s otherwise admirable TDC-D7 portable DAT recorder.)

Other features include an automatic in/out fader that operates on either analog or digital inputs, a clock whose time is automatically recorded during all taping, and a complete DAT subcode system that automatically inserts the DAT equivalent of CD track locators while recording. The subcodes give the DTC-60ES such CD-player features as track scanning, automatic track skipping, fast cueing, and programmed playback. They can be added, erased, or moved after the initial recording. There is also a CD-sync mode that starts recording in synchrony with compatible Sony CD players, as well as a crude timer-recording system that requires an external power on/off timer (this would have been the perfect use for that internal clock).

The DTC-60ES shows the status of most of these features, with the result that the indicators are rather crowded on the front-panel fluorescent display and illegible from a distance. The display can show DAT track number, four time functions (date, elapsed time, remaining time, elapsed track time), tape-transport status (record, play, and pause), sampling rate, selected input, overload margin, the status of the various subcodes, and, naturally, the signal level. I would have liked more, and more prominent, transport indicators; from across the room it’s hard to tell in which direction rapidly changing timer digits are going. The useful recording-overload margin indicator shows in decibels how close the largest peak encountered approaches digital overload.

The deck has a full complement of back-panel connectors: phono-jack analog inputs and outputs and both optical and coaxial digital inputs and outputs. The recorder’s remote duplicates most of the front-panel functions (including the subcode-editing functions) and adds controls for the various programmed-playback functions. Its buttons are too uniform in size and arrangement, however, and the transport controls are arranged differently from those on the front panel, which is an annoyance.

All our lab data involving the analog input stage and its A/D converters are referred to maximum input level, which with the DTC-60ES was just below the point where the red overload segments in the level meters lit up. That’s about 0.3 dB below maximum digital data output, a very slight and absolutely inconsequential loss of dynamic range. Take those overload indicators seriously, by the way, because they really do mean overload. The instant those red lights go on, you are clipping the signal and introducing very large amounts of distortion; you cannot allow the signal to go “into the red” as you can with analog cassettes, although the audibility of the clipping will depend on the degree of overload and how long it lasts. On the other hand, you can run signals right up to that overload point with no audible change in distortion.

In contrast, the digital input cannot be overloaded, since it simply passes the digital data stream through, though there is nothing to prevent recording an already-clipped digital signal originating from, for example, a poorly recorded CD with its pre-existing distortion perfectly preserved. On the other hand, the digital input will correctly record only 16-bit signals. Greater-than-16-bit data seem to be truncated to 16 bits, which can produce more distortion in the signal than if it had been 16-bit to begin with. This limitation means that the benefits of SBM are not available to incoming digital signals, but then again, very few nonprofessionals have access to the 18- or 20-bit digital signals that could make it worthwhile.

Our measurements are broken down into three categories: analog-to-digital, digital-to-analog, and analog-to-digital-to-analog. Data in the first category indicate the quality of the DTC-60ES’s input stages, including the critical A/D converters. Good performance here is important for those who do live recording or hope that their recordings may one day be released on CD. The sound of such recordings will be only as good as the DTC-60ES’s analog-to-digital conversion. The digital-to-analog measurements show how well the DTC-60ES reproduces DAT’s, including those made directly from a digital source. These data also show how good the deck is when acting as an outboard D/A converter for a CD player or other digital source (it’s easy to switch the deck into this mode). The analog-to-digital-to-analog measurements show the deck’s overall performance—from analog input to tape to analog output— and are analogous to a cassette deck’s record/play performance. A fourth set of data, for 16-bit digital-to-digital recording, is not printed because the results were perfect: Precisely the same audio data came out of the digital output on playback as were received at the digital input during recording. This digital cloning capability is one of the most important characteristics of the DAT format, one that is not obtainable with DCC or MD or with almost any consumer-level computer hard-disk recording system. Like all other consumer digital recorders, however, the DTC-60ES incorporates SCMS, the Serial Copy Management System, which prevents second-generation direct-digital copying—that is, digital copying of a recording that is itself a digital copy.

As for those lab measurements, practically all the numerical data show what would be excellent performance for an amplifier, let alone a tape recorder. The DTC-60ES is a stellar performer, better in all audibly significant respects than most of the first few generations of professional digital recorders (many of which are still in studio use). Of particular note are the linearity figures, which are superb in the dig-ital-to-analog mode but, surprisingly, just as good in the much more difficult analog-to-digital direction.

Also of note, and deserving of some explanation, are the noise and distortion measurements, especially those for SBM operation. First, the numbers. The reason the signal-to-noise ratio (S/N) decreases slightly when Super Bit Mapping is turned on is that the conventional A-weighting scale used does not “ignore” frequencies above 15 kHz as much as your ear would at these low levels. Remember, it is into this very-high-frequency region that SBM moves lower-frequency quantization noise, hence the decrease in measured S/N with A weighting. Use of CCIR weighting, which more closely models the ear’s response to low level signals, does show a slight improvement in S/N. Our distortion figures are for total harmonic distortion plus noise (THD+N) and thus reflect the increase in noise above 15 kHz when SBM is on.

The SBM improvement is quite a bit smaller than what one would expect from the graph on page 18 of the manual and the nearby statement that “the improvement in noise level for ... frequencies lower than 3 kHz exceeds 10 dB when SBM is activated.” As our own graph shows, turning SBM on gives an S/N improvement of only about 4 dB at the ear’s most sensitive frequencies (3 to 4 kHz), along with a telltale but inaudible rise in noise above 15 kHz. The benefit of SBM diminishes to about 2 dB below around 500 Hz, but the noise in that frequency region is dominated by comparatively large, but also inaudible, power-line harmonics. Although the improvement afforded by the Super Bit Mapping circuit in the DTC-60ES is not spectacular, every decibel around 3 kHz helps. The most likely reason for SBM’s failure to come even halfway to the claims made for it is that the A/D converters in the deck are not enough better than 16-bit devices to permit a greater improvement from the application of SBM. The manual, by the way, makes some statements that suggest the deck’s A/D converters have 24-bit resolution, which clearly they do not (our measurements imply, at best, 17-bit performance); it’s simply that data fed from the converters to the SBM system are in 24-bit format.

We were also interested in seeing how well the DTC-60ES’s D/A converters preserved the benefits of noise-shaping 20-to-16-bit systems similar in effect to Super Bit Mapping. Our second graph shows both theoretically perfect performance and the DTC-60ES’s actual performance when reproducing, through its analog output, signals fed directly to its D/A converters: the residual digital background noise of noise-shaped “20-bit” recordings from Telarc and Chesky (the 1-kHz hump in the Chesky traces is a test tone) and the dither noise of a theoretically perfect 16-bit recording. Although the DTC-60ES’s D/A performance in no case matched theoretical limits, it did very well with these extremely difficult signals.

These observations are really nitpicking, however. We heard absolutely no noise or distortion from the DTC-60ES in actual use, playing all types of music at elevated, but still domestic, levels. The deck was sonically transparent not only when recording through the digital input but also when recording even the most difficult material through the analog input. This superb sound quality, in combination with the deck’s outstanding lab measurements, ease of use, and full complement of features, means that even with its Super Bit Mapping switched off the Sony DTC-60ES is easily the best audio recorder has ever tested.

Read all

Sony reviews

Read all