Just when Tascam owners must have been starting to feel distinctly left out of the latest developments in MIDI control of tape machines, into the breach steps the LMTC Processor Controlled Synchroniser, from Micon Audio. This unit, with suitable software support, now offers them the same facilities as the Fostex system, allowing Tascam MSR/TSR series recorders, and the 644, 688 and 238 cassette multitracks, to be remotely controlled from within a MIDI sequencer.

THE CONCEPT

In a combined audio/MIDI recording system with a sequencer conventionally synced to tape, the multitrack machine, as the source of the timecode, has to be the master transport. The limited location and cycle facilities of most tape machines can, however, prove frustrating when compared to the speed and precision of MIDI in this respect. It is also self-evidently rather less than ideal to have to control what is essentially performing as a single integrated system from two different interfaces.

The obvious answer is to allow the software to also govern the tape transport, whilst maintaining the timecode-slaved status of the sequencer. This is done by creating a 'local off' mode for the software transport controls, where they no longer instruct the sequencer directly, but issue SysEx commands to be interpreted by the control interface. Using the on-screen play button (or the keyboard equivalent) in this mode simply instructs the recorder to start the tape moving, and the sequencer itself responds only when it subsequently receives input from the timecode track. Most of the sequencer's 'virtual transport' functions are supported, although naturally a delay for a tape-locating phase is incorporated in many operations. The facilities are extended to include track ready/record switching, and indeed just about any procedure that the tape machine itself will support.

The current versions of Steinberg's Cubase (V3.0) and C-Lab's Notator/Creator (V3.1) programs both incorporate control routines for the Fostex recorders, via dedicated M.ROS drivers, and 'Fostex Mode', respectively. The Micon LMTC will be supplied complete with the necessary software; additional M.ROS drivers in the case of Cubase, or a further program revision (V3.147) for C-Lab owners.

THE SYSTEM

The LMTC itself is a compact (173mm x 112mm x 30mm), stand-alone unit, built-into an extruded aluminium chassis, with just three MIDI connectors at one end, and a 9-pin D-Sub multi-pin at the other. The unit takes a DC supply from the recorder with which it is interfaced and therefore requires no separate power source. The LMTC multi-pin connects to a special split cable; one branch of this runs to the Accessory 2 connector on the rear panels of current-generation Tascam machines, and also carries a connection to the dedicated Remote Control port. The other branch, a twin cable terminated in phonos, sends and receives timecode via the line in and out sockets of the chosen track.

A further adapter is also supplied, as the cassette remote sockets use a 5-pin DIN connector, whereas the open reel recorders use a 25-pin D-Sub. The TSR8, furthermore, has the D-Sub, but requires one of the wires to be disconnected at the DIN plug — it's not quite as much of a lash-up as it sounds, and of course you only have to sort it all out once. Baud rate for the RS232 needs to be set at 2400 bit/S, and dbx noise-reduction set to off on the timecode track via the Tascam Sync facility.

THE CUBASE CONFIGURATION

Connection of the sequencer side of the system varies according to which program you are using, due to Steinberg's support of MIDI Timecode (similar to standard timecode but in a format that allows it to be incorporated into the MIDI data stream). With Cubase, MIDI Out on the Atari connects to MIDI In on the LMTC, communicating transport instructions as LMTC SysEx codes. The MIDI Out of the LMTC feeds MIDI timecode, generated from the SMPTE/EBU on tape, back into the MIDI In of the Atari, to drive the sequencer.

No further synchronisation is actually necessary, although it is quite probable that many people investing in this system may already have a tape sync unit. If the LMTC is used on its own in this way, its MIDI Thru duplicates the signal at the Atari's MIDI Out, and can therefore be used to drive the system's MIDI voice modules, but there is a problem on the input side — the Atari's MIDI In is occupied by the MIDI timecode input, and the LMTC does not incorporate a merge facility. There is nowhere in this configuration to connect a device to input data. A dedicated MIDI merger is cheap enough these days, but mixing MIDI timecode and note information on the same port can push at the limits of MIDI's bandwidth, and is to be avoided if possible. The apparent economic advantage offered, in this context, by the Cubase system's ability to read MIDI timecode is therefore somewhat illusory, for you actually still need a multi-port expander/merger, such as Steinberg's MIDEX, to create a viable system.

THE C-LAB CONFIGURATION

The C-Lab user's configuration requires the presence of the program's dedicated Unitor synchroniser, which also incorporates multi-port and merge facilities. The Unitor-based hook-up again sends the SysEx transport codes from the Atari to the LMTC MIDI In, but the output from the timecode track on tape needs to be paralleled to return both to the LMTC multi-pin and to the timecode input on the Unitor (the necessary splitter cable/adapter is not supplied). Input to the timecode track, as before, is from the 'send' side of the LMTC timecode cable. The C-Lab programs do not, at present, support MIDI timecode, and the computer remains locked to tape via the extremely fast Unitor system, which bypasses the normal 'Song Position Pointer/Continue message/MIDI-clock' routine, in favour of talking directly to the Atari processor.

FORMATTING

LMTC operation is slightly different for cassette-based multitracks and open reel machines. With a cassette transport it is necessary to allow the LMTC unit to generate its own timecode track. As it does so, it registers the system's non-linear relationship between the tach pulses used to drive the counter and the tape speed. On completing the 'stripe' pass, the cassette will rewind automatically in order to write the information gained as a data 'header' on the timecode track at the start of the tape.

'Formatting' a cassette, as it is referred to, is equally simple in either Cubase or Notator/Creator, activating the timecode generator causes the LMTC to automatically place the cassette transport in record/pause. Level checking can be performed at this stage (the LMTC outputs code at -7dBm, which is going to be fine on any Tascam machine, metering about -1VU). Now, simply pressing play on the recorder (not 'Start', as it says in the LMTC manual, which could all too easily be mis-interpreted as re-starting the timecode generator), releases the transport in record and activates the LMTC generating timecode from the start time, and at the same frame-rate, originally specified. This might seem a little confusing at first, for by the time you actually start recording code, your displayed generator-time will be considerably beyond the point at which you originally elected to start. What you need to bear in mind is that you are not actually recording code from the program's generator; you are recording code from the LMTC, which will have begun at the start time you specified.

Formatting is automatically terminated when you reach the end of the cassette, and the tape then rewinds to write the 'header'. Pressing stop during formatting produces the same result (the header is still written), but obviously striping will not be complete, which in the case of a cassette means that you can use the tape with the LMTC only up to the point where you stopped.

On an open reel machine, however, you can effectively start where you like; just locate the tape to the desired position, stick the timecode track manually into record/pause, and start the code generator, as before. Once again, pressing play on the recorder will activate the LMTC generating code from the start time originally specified. Pressing stop automatically rewinds to the beginning of the code.

'MEDIA-CHANGE'

The LMTC driver software within the two sequencer programs incorporates a 'Media-change' command to initialise the system and verify that the tape is correctly formatted for use. Cassette recorders are configured to perform a Media-change operation automatically under certain circumstances — if you power up with a cassette inserted, the system will initially read the header, execute Media-change, and then park itself dutifully at the start of the timecode. Media-change commands on open reel systems, or when changing tapes on a cassette system without powering down, are issued manually, via a simple dialogue box within the software. As the Media-change routine involves a forward wind of about four minutes, you are advised not to initiate one close to the end of a reel, but then I can't see why you would want to under normal circumstances.

CUBASE OPERATION

The LMTC will be supplied complete with the relevant software adaptation for either Cubase or Notator/Creator on disk. For Steinberg users this means M.ROS device driver files for 8, 16, and 24-track tape machines, designated LMTC1_8, LMTC1_16, LMTC1_24. Steinberg's INSTALL.PRG makes this very easy by setting the relevant filename extenders to either '*.DR' or '*.DRV' so that selected files (those with the complete extender) are loaded automatically.

The Sync dialogue box will then allow the LMTC adaptation to be selected via the SMPTE SYNC box. There are three further LMTC files: LMTC1_24.ARR gives you an Arrange window listing 24 'LMTC tracks' from which you can select tape tracks just by clicking on them, whilst LMTC1_24.ALL includes this, but also installs the MIDI Manager Part 'LMTCMIX' on track 1. Accessing the MIDI Manager reveals this to be a set of dedicated device controller objects conveniently arranged as a control panel.

You can also operate the LMTC system without using the MIDI Manager facilities at all, setting tracks to record-ready just by selecting them in the Arrange window. The transport and master record controls then function much as normal in Cubase, but driving the tape transport. Cycle, including punch-in and out, is operative, but the normal fast-wind-with-monitor facility (fast wind entered from play) is disabled. Disappointingly, particularly in view of the extent to which Cubase seeks to emulate tape recorder style control, the transport does not attempt to track the Cubase song position in any re-location operations — the machine will not begin its wind until it gets the next play instruction and realises that it is in the wrong place. Nor does it follow the 'logical locate' sequence of the stop button (current position, left locator, head-of-track), or the programmable function key autolocate points.

This means, in practice, that the tape is almost permanently somewhere other than where you want it to be, so most operations begin with waiting for it to catch up with the current position. LMTC transport control in Cubase cannot therefore really be said to enhance speed and efficiency of working, although it can still offer much in the way of-precision and convenience.

MIDI MANAGER

Control via the MIDI Manager not only provides a more graphic interface, but also adds some additional facilities. 24 track-ready switches are represented, as well as another master record switch (the main record button is inoperative when the MIDI Manager is active). There are also further switches for the recorder's Rehearsal facility and monitor status, plus the LMTC Media-change command. MIDI Manager set-ups are easily re-configured, so if you are only using an 8-track system you can of course delete the unwanted controls. Sliders are incorporated allowing a negative (ie. earlier) offset of up to 635ms to be applied to bar/beat/clock locator settings used for dropping-in and out.

There are separate sliders for rehearsal and actual recording, for the inevitable delay is slightly different in each process, and there are also separate in-point and out-point offsets. Once set for your system, these are stored and, of course, as time-related not clock-related settings, remain valid at any tempo, optimising the drop-in performance of your tape machine — a valuable extra facility made simple to use by clear and intuitive presentation. Curiously, on a programmed LMTC drop-in, the system drops the tape out of record at the right locator whether punch-out is set or not. You can get round this easily enough by moving the right locator to the end of the piece if necessary, but surely the goal was to mirror normal Cubase operation as closely as possible?

A pre-roll period may be specified, ensuring there is sufficient time for the tape transport to get moving and the synchroniser to lock-up, when playing from a specific point. Again this is entered via an onscreen slider, allowing up to 15 seconds of run-up if necessary. Further objects can be freely defined to govern any SysEx controllable parameter that the multitrack will support, adapting the system to your own working methods. The Cubase interface to the LMTC is thus 'typically Steinberg': very graphic, very flexible, very helpful. It is not however, totally idiot-proof — mixing Arrange window and MIDI Manager control of track selection can get you into a bit of a mess, because tracks placed in record-ready in one mode can not be de-selected in the other, and the visual status of a MIDI Manager switch does not always reflect the status of the track. You do really need to glance at the machine itself before you can confidently put the system into record.

"Remote control of tape machines via SysEx allows almost complete integration of the audio and MIDI recording processes. I say 'almost' because of one big difference between the two systems: you can't 'Undo' an audio drop-in."

C-LAB OPERATION

The LMTC interface within the C-Lab programs seems almost unnervingly sparse in comparison. Shifted 'F' on the keyboard, which in the original V3.1 activated Fostex Mode now, rather more appropriately in the program update supplied with the LMTC, enters Tape Control Mode (TCM). The dialogue box for this has just TCM On and Off, plus the facility for issuing the Media-change command. Once set, Tape Control Mode is virtually transparent — Tape tracks are selected to ready simply by shifted left-clicking them; de-selection is accomplished by a shifted right-click (24-track operation requires the C-Lab to be run in its 32-tracks-per-pattern mode if you want to have access to all the tracks by this method).

You enter record on any track that is selected to 'ready' with the normal record icon or key, which can also be used for manual dropping-in and out during play. It makes no difference if a track has already been used for MIDI data, but confusion (and accidents) could perhaps be prevented by using a parallel pattern in another arrange level for tape track selection. Programmable dropping-in works via the locator settings and the Autodrop function, just as in MIDI recording.

All the normal transport functions, including Cycle, are available, as in Cubase, although Start now functions like the normal Continue mode (play from present position). The big difference in the operation of the C-Lab interface is that the system is constantly updating the tape position in response to any sequencer location operations, thus imposing the minimum delay whenever a play command is issued.

The programmable Atari function key autolocate facility proved particularly valuable for rapidly accessing specific points in a song, providing, in fact, a much faster method of getting around in TCM than using the triple-chevron icons to step through the Arrange entries. Using the chevrons to move further than one entry causes the tape transport to briefly try to enter play at each arrange entry on the way to the one you want, rather than winding straight to the location as it will do under an F-key command. The standard double-stop routine sets both sequencer and tape to the top of the track, and the transport will also directly track the main bar counter display, or the Song Position Line with Graphic Arrange Mode (GAM) activated (but not GAM 'scrub' mode).

Pre-roll, or pick-up time, is set to the desired length via the normal count-in facility (a bar/beat figure which will thus vary in absolute time with musical tempo). The recorder's Rehearsal function is accessed simply by activating record with-the right mouse switch rather than the left. You can then audition the programmed drop-in as Input Monitor switching, without actually entering record. The delay, both in Rehearsal switching and actual dropping-in, has to be dealt with by directly setting the locators to the necessary offset. This is best done with the 'Position in ms' option activated, allowing the offset to be specified as a time figure — once established for your machine, the necessary negative delay will then remain consistent, irrespective of tempo settings, as in Cubase.

REALTIME MIDI GENERATOR

If you want to go beyond the facilities already described, or indeed control some of them in a different, perhaps more graphic, way, you will need to utilise the user-definable elements of the program's MIDI Manager equivalent, the Realtime MIDI Generator (RMG). One of the few things the gloriously half-translated LMTC manual gets right is to have included a comprehensive listing of the SysEx commands required to control all the available Tascam functions; while Steinberg supply a pre-programmed LMTC MIDI Manager set-up on their disk, C-Lab users are left to program their own.

RMG allows up to 64 fader and 64 switch icon definitions; it is beyond the scope of this piece to venture into a beginners' guide to setting up your own SysEx control elements, but provided you are not the type that is frightened off by the sight of anything in hexadecimal, you will find it perfectly logical, even if you haven't ever had sufficient incentive to look at that part of the program before.

I will, however, point out the one thing that catches out even the most accomplished users who try busking it on this task without reading the manual: 'P-User-event' numbers 9 to 12, which correspond to the on-screen switches, are permanently tied in pairs to the other user-definable set, numbers 5 to 8, hence why they are never actually listed in the definition line. When you access P-User 5, P-User 9 corresponds to the switches, whilst 5 is assigned to the faders. The rest are the same, 10 with 6, 11 with 7, etc. Easy enough when you know, but there is nothing on-screen to even hint at it.

Such operational advantage as there is to be gained from defining RMG SysEx control elements lies, in my opinion, not so much in the provision of a graphic interface but in the fact that it is the only way to access the some of the less basic functions. I certainly would not want to use RMG as a 'control panel', for although you can still access the transport controls, an open RMG window does deprive you of most of the rest of the screen, obscuring the Arrange listing, the active Pattern display and the Track Parameters box — the MIDI Manager window can at least be re-arranged to include only the parts you need to see, and re-sized to be no larger than necessary. Having said that I am still inclined to wonder why anyone would want to enter either MIDI Manager or RMG to put a track into record ready, when you can do it so easily from the main screen.

The most useful definitions I found, indeed the only ones I actually made use of in the end, were to create an RMG switch for the Tascam's Insert (drop-in monitor mode) facility, and (with RMG definitions being P-User events, and therefore available for inclusion within the event list) a stop command to be inserted into the event listing to automatically stop the tape at the end of the song.

IN USE

Remote control of tape machines via SysEx allows almost complete integration of the audio and MIDI recording processes. I say 'almost' because there remains a level on which the inherently different properties of the two systems must always be recognised; drop-in accidentally on a MIDI track and you can activate the 'Undo' function. Try that one with audio! Multitrack hard disk recording truly does offer the prospect of transparent integration with MIDI recording, but until such systems are more widely available, this is about as close as most of us are going to get.

The LMTC system works; indeed it works well, with precision and as much speed as the transport used will allow. It actually proved to be impressively robust in operation, resisting all my attempts to provoke it into misbehaving, although there are some persistent anomalies; in Notator/Creator, when setting-up a section to be cycled by selecting an Arrange list entry via the triple-chevrons, with the tape machine in Insert mode, no off-tape signal will be heard the first time round the cycle. The second and every subsequent pass will be fine, and no other condition of setting-up will replicate this problem. I suppose the simple solution is not to do it that way, for there are several other methods of arriving at the same end result.

It is also definitely not a good idea, in either program, to mix local and remote control of track-ready switching; I am not just referring here to the potential status confusion caused by conflicting 'flip' conditions, but something potentially more damaging. If you select a track to ready from the screen, but then decide to de-select it and do so at the tape machine itself, entering any other track into ready via screen control will also restore the first track to ready. You could all too easily find yourself recording over tracks that you thought were safe.

It is in this area that my principal reservations about the system arise (though they apply equally to the Fostex system). Neither the Cubase nor the Notator/Creator interface makes the 'audio track ready' condition obvious enough, in my opinion. A display which is not used in any other context, such as tracks flashing in inverse video, is needed for clear differentiation between audio and MIDI recording. I am also not convinced that Tape Control Mode in the C-Lab implementation should have been allowed to alter the normal MIDI recording procedure in any way; Space Bar Record is a function I normally use in a different way, for different purposes to the main Record function — I do not like being forced to make an unnecessary departure from procedures that have become instinctive.

Cubase is better here; if you are on a MIDI track, things are pretty close to normal. I would have liked to have had an additional option to set the tape control mode of either interface to a 'transport remote only' mode, putting transport controls into 'local-off' but leaving everything else as normal for timecode sync operation — having now used the system for some time, whenever I have audio material of any value on tape I find myself wanting the security of manual control of the tape machine's record functions, with the computer 'locked-out', whilst maintaining the extremely useful transport facilities. I am rather uncomfortable with this; I feel I shouldn't want it, and should trust the system. The fact that I don't says to me that the software is, at present, not quite doing its job as well as it should. Perhaps we can anticipate some further revision here.

On the hardware side, there can be no complaint about the LMTC unit itself: it's robust, totally functional, and was reliable throughout the extensive test period. The cabling is practical, rather than elegant; however, I would still rather have seen a separate cable for the remote connector rather than the short loop-out from the Accessory multi-pin.

CONCLUSION

The Micon Audio LMTC plugs a very obvious gap in the market, offering Tascam owners comparable machine control facilities to Fostex owners. Once the Fostex system was up and running, a similar setup for the other half of the market was obviously never going to be too far behind, and can now be seen to match its rival point for point. The control software for the LMTC, in each case, reflects the values of the host program — Cubase very graphic, Notator/Creator more economical, but both equally powerful in their own way. I think perhaps Steinberg have the edge when it comes to presentation, but the C-Lab system steals all the marks for efficiency of operation by keeping the tape always so much closer to the action.

It obviously makes sense for SysEx communication with peripheral devices to be extended to tape machines. The control precision and graphic information display capabilities of the computer make it the natural heart of combined operations. The ultimate destination of ever-increasing integration is undoubtedly a system in which no distinction is necessary between different types of recording, and all data can be manipulated in a similar fashion from the one interface. However, to view products such as the LMTC (and Fostex equivalent) as a further step down that road is misleading, in my view. Tape has inherent limitations which it will impose on any other medium with which it is partnered. What the system does provide is an extremely worthwhile extension of the control facilities of a combined tape/MIDI system. I have no hesitation in recommending that all owners of present generation Tascam machines and either of the 'big two' sequencers should check this out; but beware, it is not something you would want to be without once you have tried it.

STOP PRESS The LMTC now also works with Fostex multitracks, exactly as described here with Tascam hardware.

Further information

Micon Audio LMTC Processor Controlled Synchroniser £264.38 inc VAT.

International Networking Associates (Contact Details).