MIDI By Example (Part 7)
Another piece of the MIDI puzzle
If you've just discovered the limitations of FSK as a means of syncing tape to sequencer, it could be time to seek a little professional help...
In Part 6, last month, we looked at ways of using an FSK-to-MIDI converter to get a drum machine or sequencer to play in time with a multitrack tape recorder. There were, we discovered, two main problems with using such a system. Firstly, the song sequence has to be completed before recording the sync tone to tape - you cannot change the structure of the song afterwards. This is because the transmission of MIDI Clocks from the sequencer is translated into a sync tone and any tempo or time signature alterations made after the recording of that tone will alter the playback timing of the sequence. The second problem, though not perhaps so difficult to live with, is the fact that the sync tone for each song has to be recorded to tape individually - you can't record tape sync to the entire reel in one go. This is again tied in with the nature of FSK.
The solution to both problems lies in the use of SMPTE code rather than FSK. Widely regarded as an all-round professional synchronisation system, SMPTE is an acronym for the Society of Motion Picture and Television Engineers - the organisation responsible for its development. It differs from FSK primarily because it is a measure of absolute time; when using a SMPTE-to-MIDI converter, the SMPTE clock commences at a Start time determined by you - the actual time then being recorded to tape using a tone similar to that of FSK. This comprises two frequencies an octave apart, one denoting a '0' and the other a '1'. On playback, the SMPTE-to-MIDI converter recreates the timing information and passes it onto the sequencer. As the translation is a one-way process, the sequencer doesn't have to be in playback mode when the code is recorded (or 'written') to tape and this means that you can subsequently alter your song as much as you wish - as long as the Start time is kept the same, the multitrack recorder and sequencer will stay in sync.
There are two different types of SMPTE-to-MIDI converter (see Figure 1). The first is very similar to the Smart FSK type in that it creates MIDI Start, Clock and Song Position Pointer messages when the time code is played back from tape. As the sequencer is not playing back at the time when the code is recorded to tape, it is invariably necessary to enter any tempo or time signature changes via the buttons of the SMPTE-to-MIDI unit itself - a rather awkward and time-consuming process. The second type of converter generates a different kind of MIDI message called MIDI Time Code - or MTC for short - which is effectively a translation of time into a series of MIDI events. This allows us to use the tempo 'map' of the song sequence as a means of control and so alleviates the need to enter tempo and time signature changes as additional data. If your sequencer gives you the option, always use MTC.
In operation, the SMPTE-to-MIDI converter is fairly straightforward; Figure 2 shows the basic connections.
When using SMPTE, it is often more convenient to set a Start time of 59 minutes 45 seconds and to record time code on the entire length of the tape, using an outer track lo reduce the possibility of signal bleedthrough from adjacent tracks. The Start time for the first song is then set to one hour, allowing 15 seconds from the commencement of the time code. Subsequent songs then have their Start times set according to where the previous song ended. When adopting this method, it's a good idea to leave about thirty seconds between songs so that any change of mind about the length of fade outs (for example) can be accommodated after the event. The level of tone on the tape should still be around -3dB.
One of the problems of using a SMPTE-to-MIDI converter (...there's always a problem) arises if you want to continue recording to the sequencer while running in sync with the multitrack. The MIDI In of the sequencer has to accommodate both the MIDI Out from the SMPTE-to-MIDI converter and the MIDI Out from your synth or whatever instrument you're playing. Fortunately, most converters of this type have a built-in merge facility whereby the MIDI data arriving at the MIDI In is combined with the MIDI sync information being created internally (Figure 3).
A typical set-up is shown in Figure 4. The SMPTE-to-MIDI converter is connected to the multitrack as before; when the time code is being played back, the MIDI information from the keyboard (notes, pitch bend, aftertouch and so on) is merged with either MTC or MIDI Clock data and transmitted from the converter's MIDI Out. Care has to be taken to ensure that not too much keyboard information is generated, otherwise the accuracy of sync-to-tape will be compromised through the late arrival of MIDI sync data at the MIDI In port of the sequencer.
Feature by Vic Lennard
Previous article in this issue:
Next article in this issue: