From: Flawn Williams (fwilliams@npr.org) Date: 05-01-96 To: Victor Yiu (victor.yiu@psl-online.com) --------------------------------------------------------------------------- Organization: National Public Radio AROUND THE WORLD ON EIGHT D'S: A Case Study of the Sony TCD-D7 DAT Recorder by Flawn Williams (This is an update of an earlier article distributed at the 1994 Public Radio Conference workshop on Digital Field Recording.) SIX YEARS AFTER... I first started using DAT in January of 1988, taking a Technics SV-MD1 portable deck to Vietnam for a series of Alex Chadwick stories. Recording quality, tape duration, size of tape, and portability made the DAT an attractive choice over a Nagra IVs stereo reel-to-reel recorder for that project. The biggest drawback to using the DAT recorder was power consumption. A maximum of two hours recording could be done on an internal lead-acid rechargeable battery. (Other portables coming to market then, such as the Sony TCD-D10, were even thirstier for power.) For the Vietnam trip, this was solved by making a rechargeable outboard battery pack, with adequate capacity for eight hours of recording. Since our travel there seldom had us more than a day away from recharging power, this proved workable. Six years later, in early 1994, I started planning for a documentary recording assignment with reporter Phillip Davis, looking at desertification and land degradation in the Sahelian part of West Africa. We were going to Mali for the Science Desk, to do a major piece to be aired on Weekend ATC May 6. Much of this trip would entail living in remote villages for several days at a time, with the nearest power grid a hundred miles away. NPR's remote equipment pool has several Sony TCD-D10 Mark II DAT recorders, and an outboard battery system for them built by Bob Butcher of NPR Engineering Services. About eight hours of useable power are obtainable from a five-pound battery. But given the possible recording demands, I estimated I'd have to carry four or five of the batteries, since I wouldn't have much access to recharging power. That's a lot of weight when you're trying to stay portable... THE NEW KID ON THE BLOCK NPR had acquired one Sony TCD-D7, a small consumer-market DAT deck, so I decided to take a look at the feasibility of using a D7 for the trip. The size of the D7 (smaller than a Harlequin romance novel), the weight (under one pound with AA batteries), and other features made it an attractive choice for what would be a highly mobile assignment in West Africa. Linda Mack, a recordist for NPR in Washington, DC, had used her D7 on an NPR assignment in India in December 1993, and came back with high praise for it. Linda's experiences in India set me somewhat at ease with regard to power efficiency, as well as the deck's apparent low susceptibility to problems caused by heat and dust. But the deck's general ruggedness continued to be a scary issue, as I asked around and found several reports of problems. Chicago-based NPR reporter Ed Lifson had some good things to say about his D7, but it had also suffered several major breakdowns. Other early D7 users I talked to had varying degrees of love-hate relationship with their machines. the power miser The strongest point in favor of the D7 was its power efficiency. It's the first portable DAT recorder that can run for two hours or more on a set of four AA alkaline batteries. Previously only Aiwa models (also sold under the HHB and Marantz brands) had offered operation on AA's, and those decks required ten AA's for two to three hours of operation. All other decks have relied on undependable Nickel-Cadmium (NiCad) or other rechargeable options. The power efficiency of the D7 seems to go up if an external mike preamplifier is used to amplify mikes before recording through the D7's line input. This allows as much as five hours of recording time per set of AA's, substantially more than the three to four hours when using the D7's mike input. Some quick comparison tests showed that the D7's mike preamp added a noticeable amount of noise when recording a quiet interview using a dynamic mike such as a Beyer M58. But when using a higher-output-power mike, such as the Neumann RSM190 stereo condenser system, the sonic results improved markedly. The D7 has a 20 dB pad to help accommodate the hotter input levels from the condenser mike during livelier recordings. And the Neumann's battery phantom power supply is capacitively decoupled, which allows its balanced outputs to be connected directly to the unbalanced stereo mike input of the D7. Even better quality was obtained by using a high-quality battery-operated external mike preamp built by Sonosax, and going into the D7's line input. NPR had recently acquired some of these preamps after initial testing by Michael Schweppe on a Bering Sea recording assignment in the fall of 1993. The balanced XLR line outputs of the Sonosax tolerate unbalanced termination, and it also has a stereo mini output jack for convenient connection to the D7. All these outboard gizmos would make for a rather gangly contraption piled into my camera bag. There are, after all, spiffy portables from Sony, Fostex, and a few others, that combine high-quality preamps, phantom power, and other control advantages into a single package. But their power consumption is still hefty, and the initial price tag of these machines has kept them out of reach. NAGGING DOUBTS ABOUT RELIABILITY So, with the proper accessories, the D7 could do great field recordings. But how would it hold up in field use? For me, the most obvious potential danger of the D7 lies in its use of stereo mini jacks for audio input and output connections. These jacks have shown themselves to be prone to fatigue in earlier cassette and DAT recorders. And in the D7 the mini jacks are mounted directly on a printed circuit board in the machine, making repair and replacement of these jacks more difficult. Linda Mack's India experience had been free from failures. But other users of early D7's were reporting short battery life, failure to go into record, failure to load tape properly, malfunctioning mike input jacks, and other problems. A few people within my circle of contacts had already convinced Sony to replace their first (or even their second) D7. CHECKING OUT THE POWER DRAIN But I acquired a backup D7, and set out to compare the two units for power consumption. On both decks this turned out to be about four hours recording time on four alkaline AA batteries, using a dynamic mike plugged into the mike input jack with full amplification, and somewhat over five hours recording time using the line input. Both tests were run with the display light left on. The D7 also has a coaxial power input jack, which accepts 6 volts DC from external sources such as Sony's AC adapter, a car cigarette lighter 12 volt adapter, etc. I fabricated an external D cell pack, using a $1.99 four-D-cell holder from Radio Shack. To ensure connector compatibility I used the cable from my Sony AC power supply; I installed an inline connector so that I could detach the cable from the AC supply and connect the cable to the D cell pack. With four alkaline D cells, I got in excess of 16 hours of recording time while using an outboard mike preamp. The D cell pack weighs 1.5 pounds, and is about the same size as D7. The D cell holder can also use rechargeable D cells, so environmentally sensitive users can get hundreds of cycles out of one set of batteries and about eight hours record time per cycle with a D7. And while I wouldn't recommend trying to run the D7 on conventional dry-cell AA's, which can't deliver as much current as alkalines (Sony says it'll run, but for less than two hours on a set), dry-cell D's in the outboard pack will work fine. The recorder and power pack together weigh less than, and are about the same size as, my original Technics portable. But the D7, at that size and weight, can record 16 hours to the Technics' two hours! Using an external rechargeable 6-volt lead-acid battery (Panasonic LCR6V10BP, 10 amp-hours, about three pounds), I got well in excess of 24 hours record time on the D7. But for the West Africa trip I decided to go with the D cell pack, since it offered the best combination of long-life power and light weight. Plus I figured that, even in power-starved rural Africa, there might be D cells available if I chanced to need them.... GETTING AUDIO IN AND OUT Next came the question of audio connections. To lessen the torque on the stereo mini jacks, I used a right angle Sony stereo mini connector, found on their AVK-715M cable. This cable is a two-foot-long stereo mini (right-angle!) to twin RCA configuration; I clipped off the RCAs, and attached two XLR female plugs to connect the mike preamp output to the D7's line in jack. I also tried using a right-angle adapter on my headphone plug, but found that the D7's molded case design makes a right-angle adapter awkward there. So a regular stereo mini connecting cable was used for my Sony MDR-7506 headphones. The D7 drove these well, even when connected through a switch box that allows a mono sum or single output channel to be routed to both ears. Jarring is a fact of life for any recorder being carried over my shoulder or being bounced around in an off-road vehicle. To ease the impacts, I fashioned a foam rubber cradle for the D7. This gave about one inch of padding around the four sides of the deck, plus a couple of inches under the bottom, but left the top accessible for operation and monitoring. I punched holes through the side walls of the foam rubber cradle for all necessary cables. The cradle allowed me to position the D7 in my open-topped camera bag at an optimal angle for visibility and access, as well as providing mechanical protection. I dispensed with the useless fake-leather Sony case. RESULTS FROM MALI I'm relieved to report that no breakdowns or operational failures were noted on the trip to Mali. I didn't need to use the spare D7 that I had brought. I did the whole two weeks of recording (about 26 hours of tape) with two sets of D cells and two sets of AA's...and the second set of each type had substantial life left at the end of the trip. (I also used 14 9-volt cells to power the Sonosax mike preamp, which consumes a pair of batteries every four hours or so.) The D7's meters are difficult to judge peak levels accurately by. Be conservative, go for most peaks at the -12 dB dot; the meter is very unreliable above that. Watching meters on the DAT machine back in the studio while playing tapes recorded on the D7, I'm seeing levels 3-4 dB hotter than I'd normally produce with other portables. The D7 has a headphone limiter/ hearing protection system, called the AVLS, which is best left bypassed. For starters, it's confusing to operate: the same output jack doubles as headphone and line output, and a three-position slide switch selects line out (fixed level), headphone out (variable, with protective limiting), and headphone out with no protection. But the function only changes if the switch is moved while nothing is plugged into the jack! And the protection circuitry produces obnoxious drops in level. It takes a long time to recover from peaks. The headphone level is under software control, in twenty steps from off to maximum. The setting defaults back to 8 out of 20 (though my normal preference was up at 15) when the machine powers down to sleep mode after a few minutes of idle time when using AA batteries. INTERNAL VS. EXTERNAL POWER This power-down and consequent return to default settings doesn't happen when an external DC supply is engaged. So my external D cell pack kept the headphone levels from being a hassle, at the expense of slightly higher power drain in stopped mode than when using AA batteries (not a big problem). Response time when resuming recording after a pause is also faster when running on external power. When a plug is inserted into the external DC jack, this disconnects the internal AA batteries. If you're in the middle of a tape when this change of power occurs, the display mode reverts back to its Counter function, and resets to zero; the most recently recorded PNO number information is forgotten. Current A-Time is reread from the tape during the power-up cycle. Rewinding the tape back to the last recorded PNO, and then fast-forwarding up to the end of recorded information, restores this information and lets you keep on numbering with higher PNO's and correct Absolute Time. Plugging in an external DC supply, of whatever source, also defeats the battery strength meter on the D7's display. So when working with external D cells there is no gauge of remaining power available. But since the internal AA's are not being drained when running on external power, when the D's finally do run out it's quick and easy to unplug the external pack and resume recording with the internal AA's. There will be that momentary disruption, though, so if you need uninterrupted recordings make sure to start with fresh batteries. The D7 never completely shuts down as long as power is available to it. Its display reverts back to a clock setting or shows NO TAPE. A set of internal AA batteries will discharge gradually over several weeks to run this display. If all power is removed from the deck for more than a few minutes the clock and calendar settings will be lost, as there is no separate battery for the clock. AUDIO CONTROLS There are no separate gain controls for left and right inputs, just a combined stereo control. When using an external preamp or stereo mike, or a mono mike feeding both input channels, this is actually an advantage...but for split-track dual mono recording straight to the D7 it's a problem. The D7 also has two of the most listenable automatic gain control algorithms I've heard on any portable recorder, in addition to the manual record level option. This is a real plus for reporters doing interviews in pressured surroundings, but not used for the stereo recordings we did in West Africa. One note, if you use the AGC: when recording in noisy surroundings using a sensitive mike and automatic gain control, engage the 20 dB pad! Otherwise the AGC pumps too hard. The AGC circuit is AFTER the pad. No peak limiter is provided. TRANSPORT FEATURES The D7 has given me a higher percentage of erroneous tape loads than other portables. The solution is to eject the tape and then reinsert it. But at least the D7 shows ERROR on its display when it doesn't load tape properly! Some older DAT portables had a rare but alarming penchant for not loading tape properly against the head drum; not until the tape was rewound and playback attempted did you discover that nothing had been recorded! The D7 writes and reads Absolute Time, also known as A-Time, which is running time from the start of the tape in hours, minutes, and seconds. In a different subcode address the D7 also records Year/Month/Day/Time-of-Day, using an internal clock/calendar memory. Having Time-of-Day stamped in the subcode is great for knowing exactly when something occurred. And A-Time will continue to be necessary as long as the regular timer/counters in all DAT decks are so inaccurate! You cannot write additional PNO's after recording, or erase PNO's, or renumber them, without an expensive external accessory (a step backwards from the D7's predecessor, the TCD-D3). But the same RM-D3K accessory gives you SPDIF coaxial and optical digital input and output, infrared wireless remote controller, and timer record/playback start capability (though it still requires an external AC timer switch for timer operation) for about $200. There's a Hold function on the D7, which locks out all the deck's buttons. This can help to prevent dropping out of record accidentally. But I didn't use it much, since Hold also blocks the ability to press the Record button to add new PNOs during recording. For doing translated interviews, putting in PNOs at foreign answers and translated answers speeds access for logging and dubbing. One major advantage of the D7's Hold function, compared to that on the Sony TCD-D10, is that if you try to push a button to change something on the D7 while its Hold button is engaged, the display will change to "HOLD". On the D10, the Hold button is unobtrusive, there's no display to tell you it's been switched on, and you can't even power up the machine if the Hold button is engaged! HALF SPEED OPERATION The TCD-D7 offers a choice of regular "Standard Play" DAT recording speed (48 kHz sampling rate, 16 bit linear resolution) or what they call "Long Play". In the latter mode, the tape moves at half its normal speed, and the recording is made at a 32 kHz sampling rate using 12 bit non linear resolution. That translates into roughly the same frequency response and signal-to-noise ratio as a well-maintained FM radio station...still very listenable quality, quite acceptable for voice recordings. The biggest advantage of this Long Play function is that recording time on any given DAT tape is doubled. A DT-120 "two-hour" tape, for example, can hold four hours of "Long Play" audio. (No advantage of battery life is gained, though.) There are some quirks to contend with if you want to use half-speed operation. The biggest is that most DAT machines will not play back tapes recorded in this mode. They'll display Absolute Time, and see the PNO's on the tape, and even tell you that the tape is encoded at 32 kHz...but they won't slow down to play it properly. So you'll need to play back half-speed tapes on your D7 or find another machine with Long Play mode support. (Of the five other models of DAT machine I work with regularly, only one will play back Long Play tapes.) The Time-of-Day subcode data remain correct in half speed mode. But the Absolute Time subcode data correspond to full-speed time, not half-speed! The Absolute Time clock counter during Long Play playback will move twice as slow as a normal clock, only showing thirty seconds of elapsed time for each minute of Long Play recording. If you have the RM-D3K accessory box, you can make a digital transfer from a half-speed tape playing back on the D7 to another DAT recorder with a SPDIF coaxial or optical digital input. The signal travelling from deck to deck is still in 12-bit 32-kHz form, but it will be accepted and recorded AT NORMAL SPEED on the receiving deck. This regular-speed cloned tape should then play back on any DAT deck. There's also the option of making an analog copy onto a normal-speed tape. The only other potential stumbling block would be trying to make a digital transfer from a half-speed DAT to a workstation or other device that does not accept 32 kHz sampling rate. Many of them do, but if yours won't then do the transfer through an analog connection. RANDOM NOTES Beware of wrapping or coiling a mike cable around the TCD-D7. The green light used to illuminate the display panel has been known to induce a high-pitched whine into the mike cable. For best protection against induced noise, operate with the D7's light off. Given the number of problematical D7 machines that public radio users have already encountered, an extended warranty for the machine would seem a good investment. The Sony Partnership offers a two-year parts-and-labor warranty at 800-767-7669 for $120, and a four-year package for $270. The warranty must be purchased within 90 days of the purchase of the machine. (The regular warranty on the D7 is ninety days labor, and one year parts.) Also, given the cost, size, and weight of this latest model, acquiring and carrying a spare DAT recorder is not as impractical as it once was. And a spare could be your savior, as well as being a dubbing deck: even warranty repairs or replacement of a consumer-market machine will not happen overnight. Flawn Williams is the Chicago Bureau Engineer for National Public Radio, and a trainer for NPR's Training Office. If you have any questions or anything to add to this report, you can contact Flawn at fwilliams@npr.org or flawn@aol.com.