by LEONARD FELDMAN
As of this writing. you can't buy a Digital Audio Tape recorder in this country no matter how much you're willing to pay. In Japan, however, DAT recorders have been on sale since the third week of March 1987. It's expected that some DAT machines will be available in the United States before the end of this year. Therefore, the Editor of Audio and I thought it would be a good idea to have a look at some of these machines. As a rule, products arriving from Japan for sale here carry different model numbers from their Japanese counterparts. Aside from that, and their slightly different power line requirements ( Japan house voltage is 100 V a.c.. 60 Hz, as opposed to our nominal 120 V). features and specifications are normally just about the same.
-------TECHNICS SV-D1000
-------SONY DTC-1000ES
-------LUXMAN KD-117
-------MITSUBISHI CAR PLAYER
Audio's representative in Japan was asked to run down to Tokyo's famous Akihabara section an electronics wonderland where everything from electric fans to the most sophisticated digital equipment can be purchased-and pick up a DAT recorder. He selected a Technics Model SV-D1000 and promptly shipped it by air to us. I borrowed another DAT recorder, a Sony DTC-1000ES, from Tom Jung, who'd purchased it for his CD company, DMP (Digital Music Products).
The Luxman division of Alpine Electronics offered to supply us with a prototype of the DAT machine they intend to sell, Model KD 117. Finally, Mitsubishi loaned us a hand made prototype of a car DAT player. Like standard car cassette units, this tiny DIN sized prototype can only play Digital Audio Tapes; it cannot record.
=====================
The R-DAT Standard
A conference on digital audio tape recording was organized way back in mid-1983; its objective was to create uniform standards for home digital audio recording. The conference eventually attracted the participation of some 84 companies, whose representatives formed a variety of working groups, study groups. and a steering committee. In June of 1985, tentative specifications for Stationary-head Digital Audio Tape recording (S-DAT) and Rotary-head Digital Audio Tape recording (R-DAT) were compiled. A year later, in June of 1986, the final R-DAT Technical Standard was agreed upon and issued publicly. This Standard is being followed by all of the manufacturers which have begun to produce DAT recorders.
Helical scanning, much like that used in home VCRs, is employed in all R-DAT machines. Once the cassette is loaded, the tape is drawn out and wrapped around the head drum, which spins at 2,000 rpm and is angled at 6°22'59.5" relative to the tape. Unlike videotape, however, the audio tape is wrapped around only one quarter of the head drum; this limited wrap angle should help minimize tape wear. The head drum and tape capstan are driven by separate motors. Each of these motors must be precisely controlled for proper tracking of the digital data as it is recorded and played back.
-------- Figure B1 R-DAT track pattern showing PCM audio, subcode, and tracking
correction data blocks.
Figure B1 shows the track pattern on a short section of the new digital tape. The drum's two heads are each angled differently with respect to the tape, which is only 3.81 mm wide, so successive tracks will have different azimuth angles. This reduces cross talk, making it possible to lay the digital tracks down right beside each other with no guardbands between them. The standard DAT system can completely correct any error caused by mistracking or dropouts over an area up to 2.6 mm in diameter (68% of the tape's width), or a stripe 0.3 mm high. Errors caused by dropouts or mistracked areas up to 8.8 mm long and 1.0 mm high can be concealed by interpolation.
The main PCM audio information contained in each angled track occupies just under 60° of the record/play head's rotation. The smaller area in Fig. B1 labeled "ATF" contains digital data used for tracking: the area marked "Subcode" contains various identification data and other sub-codes. The copy-prohibit flag is one of these subcodes. If this flag bit has a value of zero, indicating that the source is copy-protected, digital-to-digital copying onto another DAT recorder will not be possible; when the second recorder detects this bit, its recording circuitry will shut off. Since this copy-prohibit bit is present in each data block (and is fully independent of the PCM audio data), it would be extremely difficult to defeat or override it.
The Table shown here lists basic parameters of the R-DAT Standard. Note that it includes a wide-track option for prerecorded tape, to make high-speed duplication easier. Since the wider track pitch reduces the re cording density, the tape must be speeded up to restore the data-transmission rate to the standard 2.46 megabits/S. This cuts the maximum recording time from 120 to 80 minutes in wide mode.
The Table also includes the additional home-recording options available at a sampling rate of 32 kHz. However, I do not feel most audiophiles will want to use the machines for such limited-fidelity applications. -L.F.
-----------Basic parameters of the R-DAT Standard.
=====================
I managed to acquire an English translation of the operating manual for the Sony unit, but I was pretty much on my own when it came to the other three machines. Fortunately, I was able to figure out and use most of their features thanks to the clear and logical English nomenclature on their front and rear panels. With rare exception, Japanese manufacturers label electronic products' controls and switches in English even when they are intended solely for local consumption. (I guess that saves the cost of tooling up for two kinds of panels and having smaller production runs of each.) For bench tests, measurements, and listening tests, I obtained several samples of blank Digital Audio Tape (again, much of it purchased in Japan, the rest provided as prototype samples by TDK) as well as a few prerecorded DAT cassettes of classical and popular music. DMP also supplied a demo tape for use in this project.
It was obvious from the first that the manufacturers have learned a great deal from the successive generations of CD players they have developed and produced. Each DAT data block has a subcode area that can be used to identify the start of a selection by number. Other control and identification data is also included in each block. (For a de tailed view of the standard DAT data block format, see the accompanying sidebar, "The R-DAT Standard.") It appears that all DAT decks will be able to record this identifying mark automatically at each point on the tape where the deck is put into record mode; some, including the Sony, also allow the user to record it manually at any point on the tape.
Codes assigning program numbers can also be recorded automatically by any deck, and manually with the Sony and some others.
These codes are, of course, completely independent of the audio data and do not in any way interfere with it or degrade audio quality.
Once identification codes have been entered in the special subcode areas, the DAT unit can search at high speed for any de sired point that has been so marked. The "start ID" codes can be detected in sequence, for example, or one can search for a specific program number. The Sony DTC 1000ES also lets you add a "skip ID" signal to any selection on the tape, during either recording or playback. During playback, if you press the "skip button, the tape will skip over any start-ID signal that is accompanied by a skip-ID signal. The skip-ID signals can also be erased. in which case the start-ID signals will once again become detectable.
SONY DTC-1000ES
----------The tape drawer of the Sony DTC-1000ES DAT recorder.
The front panels of DAT recorders resemble those of CD players. The Sony's panel starts at the extreme left with a "Power" but ton, a "Timer" switch (for use with an external clock timer), and a stereo headphone output jack. The cassette compartment. next to the power switch. is similar to a CD drawer except that it is much less wide: The standard DAT cassette measures only 7.3 cm wide by 5.4 cm deep by 1.05 cm thick (2 7/8 x 2 7/8 x 13/32 in.). An "Open/Close" button operates the cassette drawer, which can also be closed with a light press of its front.
An elaborate display area is to the right of the cassette drawer. The main display is a real-time tape counter which shows elapsed time of the selection being played, tape running time. or total remaining time, depending upon the setting of a counter mode button.
Program-number indicators show the number of the selection being played, providing the tape has been recorded with program-number subcodes. An "AMS" (Automatic Music Search) indicator displays the number of selections to be skipped backwards or forwards when this function is in operation. If a numeric button is pressed (to call up a specific selection by its program number), the display shows the target selection number while the tape transport searches for it. The Sony display includes a "Copy Prohibit" light. This illuminates whenever an attempt s made to produce a direct digital-to-digital copy of a CD or digital tape which has been copy-protected by a digital flag code. This digital-to-digital copy-inhibit flag is present n every CD in my collection and, according o the best information I have, is on just about every CD produced to date. It will also be present on prerecorded DATs. For now, to record digitally protected sources onto DAT, you must feed the source player's analog outputs to the DAT recorder's analog inputs.
Sampling-rate indicators on the display how whether the recorder is sampling at 48, 4.1, or 32 kHz, all three of which are covered by the DAT Standard. In playback, DAT machines automatically recognize the sampling rate of the tape and set themselves accordingly. The 48-kHz sampling rate is the only one used for recording on the DTC-1000ES and most home DAT recorders. Pre-recorded DATs are expected to use the 4.1-kHz sampling rate, which will allow record companies to produce them easily from D master tapes; the inability of DAT home decks to record at this sampling rate is a further inhibition against copying CDs in a digital-to-digital manner. The 32-kHz sampling rate is used in direct-broadcast satellite transmissions of digital audio; most home CM' Sate, DAT decks can play but not record tapes made at this rate.
A "Caution" light illuminates in the display if moisture condensation is present on a tape. A "Search" indicator lights up while tape search is in progress, and various other indicators illuminate when such functions as "Start ID" and "Skip ID" are being entered or deleted.
-----------Most of the Sony's controls are duplicated on this remote transmitter.
------------Like the other two DAT decks. the DTC-1000ES combines counter
program, recording-level meter. and many other indicators in one centralized
display.
-----------The Sony DTC-1000ES gives the user the options of manually adding or erasing, start and skip ID codes. Its recording-level controls resemble those of many analog cassette decks.
The lower section of the Sony's display area is dedicated to separate fluorescent record-level meter indications for left and right inputs. Because of DAT's wide dynamic range, this metering system is calibrated from 0 dB (maximum recording level) down to-50 dB. with an infinity symbol still further down the scale. Above zero, instead of the numerical calibrations familiar from analog recorders. there is only a red "Over" warning. As I was soon to learn, the recording level denoted by the "0-dB" mark is truly a maximum. Analog recorders can tolerate some degree of over-recording, but if you go over the 0-dB mark on a DAT recorder, you hit a brick wall of horrendous distortion. I suspect this will take some getting used to on the part of experienced home recordists who are accustomed to pushing past recommended maximum levels in order to get the best dynamic range and signal-to-noise ratios possible.
Major tape-transport operating controls are located beneath the Sony's display area. Included are "Stop," "Play," "Cue" (fast for ward), "Review" (rewind), forward and re verse "AMS" buttons, "Rec," "Pause." and "Rec Mute." Counter "Mode," "Reset," and rewind "Memory" buttons are to the right of the display. as are buttons ("Write" and "Erase") for inserting and deleting the start-ID and skip-ID subcodes described earlier.
There's even a button for renumbering your selections in consecutive order on a tape. If, for example, you had only recorded starting IDs for a given tape (but had assigned no program numbers), pressing the "Re-number" button would rewind the tape to the beginning; program numbers would be entered automatically in numerical order for each of the selections you had recorded on the tape.
The number keys. a "Clear" button (for cancelling incorrect entries), and a "Start" button (for initiating play of a randomly selected program) are farther to the right. Con centric rotary record-level controls are at the far right of the panel: below them are a "Phone Level" control. a "Skip" switch. and a two-position switch that selects either the analog (line-level) inputs or the digital input.
The remote control supplied with the DTC 1000ES duplicates virtually all of the control functions of the front panel with the exception of such operations as setting record levels, selecting inputs, and opening and closing the cassette drawer.
The rear panel carries the usual pairs of phono jacks for analog line input and output.
It also has separate digital input and output jacks (each digital jack carries both channels). These jacks allow digital-to-digital copying of home recordings, using this deck and a second DAT recorder. Another purpose of the digital output jack is to permit D/A conversion of recorded digital data using a separate D/A converter such as Sony's DAS-703ES. It's worth repeating that the digital input jack will not allow you to record any digital program source that is protected by a digital flag code, as just about all CDs are and as all future DAT prerecorded cassettes are expected to be.
Technics SV-D1000
Technics too seems to have kept the most desirable convenience features of CD players in mind when they designed their first DAT recorder. As with the Sony, program numbers can be assigned to selections as they are recorded on a tape by the user;
unlike the Sony, however, start-ID coding is automatic only and cannot be added by the user. Search facilities and tape-counter modes are similar to those on the Sony unit, with the addition of a purely numeric counter mode. The Technics player also offers a few other controls and features not found on the Sony. For example, the A-to-B phrase repeat popularized by CD players is available on the SV-D1000, as is an output level control that affects both the line and phone outputs.
Another useful addition is a recording channel-balance control that allows you to adjust for any channel balance disparities present in the program source you are recording. Of course, such adjustments can usually be made at the program source, but it's nice to have this facility on the front panel of the DAT recorder itself.
---------The tape drawer of the Technics tilts the cassette up towards
the user.
--------The yellow and red segments on the Technics SV-D1000's recording meter help warn the user when levels approach the point of overload distortion.
--------Only the Technics has a recording balance control and it proved useful.
----------Rear-panel jacks of the SV-D1000 include both coaxial-cable and fiber-optic digital connections. as well as the analog input and output jacks.
The Technics' record-level display is far more extensive than the Sony's. Its calibration extends across more than one-third of the front panel and is numbered from 0 dB (maximum record level) all the way down to-60 dB. As with the Sony, there are no calibrations above 0 dB, merely an "Over" indicator which lights when low-distortion levels are exceeded. The meter also has peak hold, a feature found on better analog cassette recorders, which shows record-level peaks for several seconds after they have occurred.
The Technics unit, in addition to its other programming options, has a "Recall" button which lets you review your programming to make sure you haven't entered an incorrect number. Just about all the remaining controls on the Technics perform the same functions as those on the Sony, with most of those functions, again, available on a supplied remote control.
In addition to its analog inputs and outputs, the Technics has digital input and out put jacks. There are two sets of them, in fact, one for coaxial cable and the other for fiber optics. Its circuitry also senses when an at tempt is made to duplicate a copy-protected digital recording. If you attempt to copy such program material by direct digital-to-digital data transfer, a light associated with the digital input's, selector button flashes on and off (instead of lighting continuously) and you cannot get into record mode.
Luxman KD-117 Prototype
The Luxman KD-117 sample was supplied with neither an English nor a Japanese owner's manual, so it was a bit trickier to figure out and use than the other machines. Nevertheless, I found that most of its features lined up pretty well with those of the other home DAT recorders.
-----------The tape drawer of the Luxman.
-----------Indicators for the three standard sampling rates are on the front of the Luxman's tape drawer.
------------The Luxman KD-117's comprehensive display includes a recording-level meter calibrated from 0 down to-80 dB.
---------The Luxman has two sets of analog inputs, one intended specifically for recording from CD players.
Again, the level-meter scales are integrated within the main display area. These meters, like those on the Technics unit, are linearly calibrated but go down to-80 dB.
Like the Sony unit, the Luxman lets the user add start-ID signals manually as well as automatically.
The Luxman has two sets of line inputs (the first labeled for CD use, the second for other line-level program sources) as well as a digital input. Next to the input selector switch on the front panel are left and right level-calibration controls which can be used when re cording sources with right/left level imbalances. Though the front-panel markings make these controls look as if they apply only to the CD input, they also work with the other line-level analog inputs.
Since the Luxman does have separate in puts for a CD player, I presume the "Low Pass Filter" switch on the front panel is to remove any ultrasonic products from the in put signal's path when transcribing a CD via the digital-to-analog-to-digital route.
The remaining controls are substantially similar to those on the first two machines I evaluated and analyzed. The designers of these machines all favor putting the cassette drawer at the left, the main display area near the center, and operating controls to the right of the display and/or beneath it. The designers of the Luxman unit chose to place the main tape-transport and record buttons at the extreme right of the panel and the number keys (for programming and direct accessing of selections on a tape) below the display. Sony and Technics designers chose the reverse, placing the number buttons at the extreme right and the main controls beneath the central display. Either arrangement is perfectly fine with me.
------------ A prerecorded DAT sample from a trial run by GRP.
Preliminary Measurements
I'm calling the lab tests I performed on these three recorders "preliminary" because I did not measure them as comprehensively as I would if they were currently available for purchase. (The same holds true for my measurements of the Mitsubishi DAT player for the car, discussed at the end of this article.) The object of these tests (during which I fed generator-produced test signals into the machines' analog line inputs) was to get a good first-hand look at how well these models re cord and play digital tapes.
Frequency response curves for the three DAT recorders were all pretty similar, so I am showing only the Luxman's (Fig. 1). Note that overall record/play response extended to 22 kHz, thanks to the 48-kHz sampling rate I used in recording. With this sampling rate, output filters, whether analog or digital, need be concerned only with cutting frequencies above 24 kHz (half of 48 kHz), rather than with everything above 22.05 kHz, as would be needed with CD's 44.1-kHz sampling frequency. At 22 kHz, response was down only 0.1 dB for the Sony and Luxman units; the Technics was just about as flat, down only 0.2 dB. Both the Technics and Sony deviated less than 0.1 dB from perfect uniformity at any frequency within the audio band.
Since THD was too low to measure with my Sound Technology test instrumentation, I used a conventional, manually operated distortion analyzer. At 0-dB recording level, THD was approximately 0.005% for all three recorders. As is usual with any digital audio medium, lower recording levels yielded linearly higher levels of harmonic distortion, as the system had fewer and fewer digital "bits" with which to describe sampled amplitudes.
Figure 2 shows what happened when I tried to push recording levels above the 0-dB mark on the Sony unit; I got virtually the same results on the Technics and the Luxman. At +3 dB recording level on the Sony machine, third-order distortion reached 3.9%; had I pushed harder, to +4 dB above nominal maximum recording level, it would have approached 10%. Backing off to +2 dB dropped the distortion level down to little more than 0.1%, and at +1 dB it fell off the bottom of the graph (my Sound Technology tape tester can only read distortion levels down to around 0.01%). When measured by other means, the residual THD of these machines at 0-dB record level proved to be no more than around 0.005%.
Fig. 1--Record play frequency response of the Luxman prototype. Like the response
of the other decks. I was extremely flat from 20 Hz to 22 kHz.
Fig. 2--Overload distortion on the Sony DAT recorder. Below 0-dB recording level. distortion on all three decks was less than 0.01% but rose rapidly. as shown when level was raised beyond 0 dB.
Fig. 3A--A-weighted S/N analysis, DTC-1000ES.
Fig. 3B--A-weighted S'N analysis, SV-D1000.
Fig. 3C--A-weighted S N analysis, KD-117 prototype.
Similarly, the Technics sample offered vanishingly low distortion at or around 0-dB recording level, but when pushed to +2 dB, its third-order distortion zoomed up to 4.9%.
Like the Sony unit, its distortion would also have climbed to around 10% had I raised the recording level another dB or so. The Lux man delivered very similar results, with distortion around 0.005% at 0-dB record level, rising to around 4.0% at the +3 dB record level shown on its metering system.
Figures 3A, 3B, and 3C show the results of A-weighted signal-to-noise analyses done for the three recorders. The Technics and Sony units registered readings (on a one third-octave by one-third-octave basis) of over 100 dB for most of the audio spectrum and overall composite S/N readings in ex cess of 90 dB. The Luxman unit registered an overall composite S/N reading of 87.5 dB.
These figures are more comparable to those I obtained from first-generation CD players, nearly four years ago, than to those I get from CD players today. Perhaps this is because these DAT recorders are first-generation machines, too; also, I was forced to use analog test signals, for reasons I'll get to shortly.
If these three units had any wow and flutter, it was well below the level detectable by my test system. In other words, any wow and flutter produced in recording and playback together was below about 0.002%! Channel separation was of the same order as that measured for most CD players I have tested recently, as illustrated in Figs. 4A, 4B, and 4C. I attribute the slight decrease in separation shown for the Luxman and Technics units to some minimal amount of capacitive crosstalk present in their analog output stages rather than to any part of the digital circuitry. In any case, I think it's safe to say that- 70 dB of crosstalk at 10 kHz is hardly going to affect stereo imaging!
Use and Listening Tests
One of the things that particularly impressed me about all three machines was their speed of access. Most of us have got ten used to the rather slow fast-forward and rewind speeds of analog cassette decks.
Such decks must run even slower in their "fast" music-search modes so they can detect the. pause between selections. In contrast, the DAT machines were able to find any desired selection on a two-hour tape in times that ranged from a few seconds to slightly more than 30 S, depending on how far away the desired selection was from the tape's current position.
This rapid access is possible for two reasons. First, longitudinal tape travel in the play and record modes is just a bit less than one third inch per second (8.15 mm/S), so it takes less than 100 feet of tape to record a two-hour program. In contrast, a standard C 90 analog cassette contains about 420 feet of tape for 45 minutes of recording in each direction.
The second reason for R-DAT's quick ac cess has to do with the way its high recording density is achieved. The system uses an angled, spinning head assembly (like that on a VCR) which lays very narrow tracks diagonally across the tape. This raises the writing (head-to-track) speed to 3.13 meters/S, or 10.3 feet/S. In fast-search modes, the tape remains in its normal 90° wrap around the head drum, which continues spinning at the same 2,000 rpm as in playback or recording.
Those spinning heads can therefore detect start IDs and other subcode information even during fast-wind modes, which shortens ac cess times.
When you call for a given selection on a DAT recording, the transport goes into the fast-wind or fast-rewind mode. The deck reads the program-number codes as they pass: as the desired selection approaches, the tape slows down. The process is interesting to follow on the tape counter. It's as if some unseen hand were guiding the fast-wind process, controlling the tape's speed to find the precise start of a selection quickly and accurately.
I made several recordings on each of the three machines. For test purposes, I had hoped to be able to copy my latest test CD onto DAT, using digital-to-digital connections. As you might have guessed, the copy-inhibit system foiled this attempt, and I had to settle for a digital-analog-digital transcription. Interestingly, one of the prerecorded tapes given to me did not have the copy-prohibit flag encoded in its subcode control block, so I was able to make digital-to-digital copies of it. (It was also recorded at the 48-kHz sampling rate, of course.) The other two music tapes, and all the CDs I tried to transcribe, did have the copy-inhibit flag. As a result, when I tried to record them by feeding my CD player's digital output to the recorders' digital inputs, the Sony's "Copy Prohibit" light and the LED on the Technics' "Digital" input-selector button began to flash within seconds. I had to settle for recording via my CD player's analog output and the recorders' analog inputs, adding extra D/A and A/D stages and analog circuitry to the chain.
I mentioned earlier that maximum recording levels are extremely critical when using DAT machines. As careful as I tried to be, I invariably pushed the meters above the 0-dB mark whenever I attempted to obtain maxi mum dynamic range. I soon decided that it's better to stay well below 0 dB to ensure that peaks never exceed that mark. I learned, too, that manufacturers are going to have to be extremely careful in calibrating the re cording-level indicators. An error of 1 or 2 dB can make a big difference if the 0-dB reading actually corresponds to + 1 or +2 dB above true maximum recording level (the level at which 16-bit samples consist of 16 ones and no zeros). In fact, I discovered that the calibration of the Sony unit made setting levels a bit more critical than with the Technics or Luxman. At 0-dB record level (as indicated on the Sony's meters), THD for the record/ play cycle was already beginning to creep up just a bit, though it was still below 0.01%.
Backing off by as little as 0.5 dB brought THD down to the 0.005% level exhibited by the other two units when they were set for their "0" record-level indications. It's possible, of course, that the slight rise in distortion may have been caused by overload limitations of the record amplifier's analog stage or even by slight overload of the playback amp's analog stage, but my guess is that I was coming up against the limitations of the 16-bit sampling system.
Fig. 4A-Channel separation vs. frequency, DTC-1000ES.
Fig. 4B-Channel separation vs. frequency. SV-D1000.
Fig. 4C-Channel separation vs. frequency, KD-117 prototype.
Making a DAT recording turns out to be as simple as operating the most basic of cassette decks. You don't have to add starting IDs to your individual selections if you don't want to, but I found this feature to be very desirable, especially when recording a series of signals from my test CD onto a tape.
As each new test began, I simply pressed the "ID" button. Then, when I needed a tape with which to test the play-only Mitsubishi prototype car unit, I was able to skip easily
back and forth to the test signal I required instead of having to hunt for it with fast for ward and rewind. On all three decks, counter readings change so quickly during fast winds that even if you know the tape count (or real-time indication) where a given selection starts, you will have a hard time stopping precisely at that spot. Much better, I soon discovered, to let the "smart" microprocessor read subcode data and do the cueing job for you.
There was no mistaking the sound quality of these units during playback. The prerecorded material had that familiar "CD sheen" about it, with a total absence of any audible background noise and with a dynamic range that was unmistakably "digital" in character and intensity.
I hope to be in a much better position to evaluate sound quality a few months from now, if and when a greater variety of prerecorded DAT software becomes available. I well remember when my CD collection consisted of five CDs, none of which contained particularly outstanding examples of the re cording or performing arts. Testing these DAT recorders with the even more limited selection of software available to me recalled those early days of CD excitement. Of course, there is one important difference. I was able to record good-quality CD material onto DATs, albeit by going from digital to analog and back to digital. Even those recordings sounded very clean and offered wide dynamic range as well as low noise levels (once I learned how to adjust recording-level controls properly).
I am not about to rank these three DAT recorders in any order of quality or preference; it's much too early to do so. Suffice it to say that all three units attest to the great engineering skills of their designers. Eighty plus companies labored diligently for more than three years to establish a workable R DAT standard. (The "R" stands for rotary head, as distinct from S-DAT, a format that employs stationary heads and may someday be employed for professional DAT gear.) They have obviously done their job well.
Whatever the future holds for DAT recorders, the technology and ingenuity, as well as the many features of this new recording for mat, must be admired and respected. As I understand it, these first units are selling in Japan for the equivalent of $1,200 to $1,500.
If and when they show up on these shores, their price will probably be considerably higher. Still, if the audio enthusiasts in this country want DAT recorders badly enough, you can bet that they'll buy them and, in so doing, will establish a demand for them.
Once the market broadens, it'll just be a matter of time before DAT recorders will be come affordable for the less affluent members of the audio fraternity.
MITSUBISHI
Car Player Prototype
----- A play-only unit. Mitsubishi's DAT machine will be for car use.
When you consider the amount of digital electronics involved in a DAT recorder, it is nothing short of amazing that Mitsubishi (and no doubt others in the near future) will be producing a car DAT player that fits a standard DIN or ISO dashboard cutout. The unit supplied by Mitsubishi was clearly a hand built sample. I was warned not to treat its mechanism too harshly, but in my lab it worked beautifully. It was powered up via a DIN cable connected to a Mitsubishi car power amp (Model CVX-5, rates at 70 watts per channel in stereo mode); in effect, I had on my lab bench a complete car sound sys tem driving a pair of monitor speakers.
When the Mitsubishi is first powered up, a display area at the right of the tiny front panel illuminates but remains blank. As soon as a tape is inserted, the time display appears, showing minutes and seconds. The initial reading is 00:00, but seconds and minutes begin to register as soon as play starts. Also displayed is the program or selection number and a long arrow symbol containing the word "Play." If the repeat-play functions are called up, the display also shows either the word "All" (for repeat of the entire tape) or "1" (for repeat of the current selection).
Controls are similar to those on the home DAT recorders except, of course, that there is no "Record" button. "Bass," "Treble," front-rear "Fade," and concentrically mounted volume and balance controls supplement the tape-motion controls. A repeat button, when pressed, sequences through single, full, or no-repeat commands. Taking a cue from some car radios and even a few car cassette players, Mitsubishi has included a "Scan" button. Press it, and the tape advances from one recorded selection to an other, stopping and playing the first 10 S of each before advancing to the next. Pressing the "Play" button at any time disengages "Scan" and allows you to listen uninterrupted to the selection of your choice. Considering that a DAT cassette may contain as much as two hours of recorded material, this is a very useful feature, allowing a driver to audition many selections without having to repeatedly take his hands off the wheel.
When the "Stop/Eject" button is pressed, the tape is first released from its wrap around the spinning head drum and fully retracted into the cassette shell while the head drum ceases its rotation. The cassette is then ejected partway from the cassette slot-just enough for you to see its rear edge in the slot. A second push on the same button releases the tape cassette still more, allowing you to retrieve it from the slot. At least on this prototype, I was told to push tapes only partway in for loading and then to press the "Play" button. When I did so, the loading mechanism took over, gently positioning the cassette further inside the player and "threading" the tape for its 90° wrap around the spinning head drum. This unit may be a prototype, and a hand-made one at that, but its mechanical action was smooth and reliable and the sound quality reproduced via the accompanying amplifier was excellent. If production units are as good, this product should enjoy an excellent reception when it becomes available.
Mitsubishi was kind enough to supply a wired DIN adaptor. Its use allowed me to measure the DAT player directly without having to go through the accompanying car amplifier, though I did use that amp for my listening tests.
I had recorded a frequency response sweep from my EIA test CD to a DAT cassette using the digital-to-analog-to-digital approach on one of the home DAT recorders. I felt that this type of test tape would be adequate for checking the playback response of the Mitsubishi; as you can see from Fig. 5, it was. Response varied by no more than ± 0.25 dB from 20 Hz to 20 kHz. For signal to-noise tests, I had to use a tape that I had made (with no signal applied) on one of the home recorders; therefore, the plot of Fig. 6 may well show that deck's noise in recording as well as the residual playback noise of the Mitsubishi. In any event, this car player measured over 100 dB for most of the 1/3-octave points analyzed; overall composite noise, A-weighted, was 86 dB. That's not at all bad for a "prototype."
Fig. 5--Playback frequency response. Mitsubishi car stereo prototype.
Fig. 6--A-weighted SS analysis. Mitsubishi car stereo prototype.
Summary
After spending just over a week with these four units, I have concluded that DAT represents the next step in the quest for a true high-fidelity home recording system. Of course, those of us who have wanted to record audio digitally, at home, have been able to do so for some time by linking a VCR to a PCM processor. Somehow, though, doing so has always seemed to me to be a sort of "add on" to what is basically a video-oriented technology. After all, when you re cord digital audio onto videotape, you're still wasting some of the tape's capacity by re cording horizontal and vertical synchronizing video pulses and merely substituting PCM code for picture signals. In DAT we have a system that is designed expressly for recording and playing high-quality audio-a sys tem that takes into consideration the special needs of audio, including the right levels of error correction and detection. Being dedicated to audio, the system also provides for all of the convenience features that digital data can utilize, such as various programming, searching, and scanning modes. Personally, I can't wait to measure DAT units built for U.S. distribution and sale. I'm eager to do so for two reasons: First, so I can report my findings to you, and second, so I can pick the right model to purchase for myself.
(Audio magazine, Jul. 1987)
Also see:
Sony DTC-75ES DAT Recorder (Equip. Profile, Nov. 1990)
Sony TCD-D3 DAT Walkman Recorder (Jan. 1991)
Akai AD-93 (DAT) Digital Audio Tape Recorder (Jun. 1988)
Harris XD-001 UH DAT Recorder (Equip. Profile, June 1989)
DECIDING ON DAT: Witness of the Persecution (Mar. 1988)
Nakamichi 1000 DAT Recording System (Nov. 1989)
= = = =