You are here

MIDI Synchronisation Explained

Tips & Techniques By Paul White
Published July 1994

For the benefit of those adding a MIDI sequencer to their multitrack tape facilities, Paul White explains the various synchronisation options.

Multitrack tape recorders can only provide a limited number of tracks which, in turn, places a limit on how much you can record while still keeping the individual parts separate — or it did until MIDI sequencing came along. With MIDI you can have, within reason, as many tracks as you have MIDI instruments, and with many modules offering 8‑ or 16‑part polyphony, even a modest system can provide more tracks than a professional multitrack tape recorder. Of course, MIDI sound sources aren't always what you want, but most modern recordings make use of at least some electronic sound sources, and by getting your sequencer and multitrack tape recorder to join forces, you can have the best of both worlds.

You could simply create your backing track using a sequencer and then mix this to one or two tracks of your multitrack tape machine, leaving the remaining tracks free for vocals and guitar solos, but the sequenced parts, once mixed, can't be remixed. Secondly, whenever anything is recorded to tape, some quality is lost, especially when working with cassette.

A far more flexible option is to have the sequencer running in sync with the multitrack tape recorder. If you work in this way, all the sequenced sounds can be fed into the final mix without ever having been recorded to multitrack at all. If the sequenced parts come from a single, multitimbral MIDI sound source, then a cassette multitracker with six or more input channels will be able to handle both the taped music parts and the sequenced MIDI parts. If, however, your MIDI system comprises several different sound modules, you'll need a multitracker with lots of mixer inputs or even a completely separate mixer. So far the concept is pretty simple and the advantages equally clear, but how does synchronisation work?

Drum Machines With FSK

Many drum machines come with a tape sync facility built in. In general, they employ what is known as an FSK (Frequency Shift Keying) system which records a series of electronic tones onto a spare track of the tape machine; if you're into analogies, these work very much like the physical sprocket holes in a cine film. These tones are related to the tempo clock of the drum machine, so as the tempo is increased, the electronic sync track follows.

To sync such a drum machine to tape, you must first program your entire drum part, complete with any tempo changes you might want to include, then play back the piece, at the same time recording the sync code output from the drum machine onto one track of your tape recorder. Some tape machines have dedicated sync ins and outs, while with others you have to use the audio ins and outs. Note: sync code is normally recorded onto the highest numbered tape track, which will be track four on a 4‑track machine. If the machine has a special provision for handling sync codes, this will invariably be found on the highest numbered track.

To run the drum machine in sync with the tape machine, the tape output from the sync track you have just recorded must be plugged into the Sync In socket on the drum machine, the drum machine switched to Tape Sync, and the tape started from the beginning. The drum machine will automatically start when it receives the sync code from the tape machine and should stay in time with the tape until you stop the tape machine. This allows normal recordings to be made onto the remaining tape tracks.

FSK sync is very simple, but whenever you stop the tape, you have to wind it right back to the beginning again to establish sync, which can be very frustrating. Nevertheless, if you are happy to work in this way, you can add a sequencer to the system simply by setting the sequencer to External MIDI Sync and connecting the MIDI Out from your drum machine to the MIDI In of your sequencer, as shown in Figure 1. Unless your sequencer has more than one input and has a MIDI merge facility, you won't be able to record new parts onto the sequencer while you are locked to tape, because the sequencer's MIDI In is taken up by the MIDI Out from the drum machine. This might not worry you, but if you want the flexibility to work on your MIDI parts while listening to the tape parts you have recorded, you'll need to add a MIDI merge box to the system, in order to combine the MIDI output of your master keyboard with the MIDI Out of the drum machine (see Figure 2).

Once the tape tracks have been recorded, the tape outs are mixed with the outputs from the sequenced MIDI instruments and drum machine, as shown in Figure 3.

Smart FSK

There is now a refinement of the original FSK sync code, called Smart FSK; this still works by recording the code to tape in the form of tones, but it is designed to work with MIDI SPPs (Song Position Pointers) so that the tape machine can be started anywhere in the song and the sequencer will always find the right place. Because the code is still generated from the sequencer's MIDI clock (which is related to the tempo of the original sequence), any tempo changes that have been programmed will be reproduced accurately when the sequencer is sync'ed to tape. However, you can't go changing the tempo of your sequence after the code has been recorded to tape.

Several manufacturers now produce this kind of sync unit, and some include a MIDI merge facility so that new sequencer parts may be recorded while the sequencer is synced to tape. Figure 4 shows how a system can be set up using a smart FSK sync box. In this example, the drum machine is running as a slave from the sequencer's MIDI Out. Note that the sequencer must send and recognise SPPs for smart FSK to work. Modern sequencers invariably work with SPPs, but some older systems may not be so obliging.

SMPTE

Originally designed to sync soundtracks to films, SMPTE is now also commonplace in MIDI studios. Unlike Smart FSK, SMPTE is based on real time measured in hours, minutes and seconds, with further subdivisions to accommodate individual frames of TV and film material. Because SMPTE code is independent of tempo, a whole tape can be recorded or 'striped' with code before any recording or programming starts. Think of it like printing the inch markings on a ruler.

Because there is no tempo information inherent in the timecode itself, a conversion has to be done somewhere along the line, either by the computer used to run the sequencing software, or by the microprocessor inside a SMPTE‑to‑MIDI sync box. The starting tempo of a piece of music and the SMPTE location, plus any subsequent tempo changes, are stored in the form of a 'tempo map' which must be used alongside the sequencer data whenever the sequence is run sync'ed to tape. Some modern sequencers can create a tempo map automatically and store this along with the sequence data, which makes life much easier.

So, why go to the bother of using SMPTE? Unless you are planning to work to picture, SMPTE offers little more than a smart FSK system. Both allow you to start at any point in a song and both will look after your tempo changes, smart FSK doing so automatically. In SMPTE's favour, it does allow you to stripe the tape first, whether you have programmed any sequence or not, while smart FSK requires that you have programmed your sequence, at least as far as its length and tempo goes, before you start work. Furthermore, if you do decide a tempo change is in order, you have to re‑stripe with a new FSK code, whereas with SMPTE you simply have to create a new tempo map. But in practice, these differences are fairly minor.

MTC

A MIDI‑specific variation on the sync code theme is MIDI Time Code, which is essentially a MIDI interpretation of the traditional SMPTE protocol. Most of the modern sequencers can lock to MTC, and because the code comes in via MIDI, the hardware requirements can be simpler. For example, the Alesis BRC remote controller outputs MTC so, with no more than a MIDI lead, you can sync an ADAT plus BRC system to any modern sequencer. Again this system requires that you create a tempo map, but as I've pointed out, the better modern sequencers handle this for you automatically.

Another a dedicated sync code goes by the name of Direct Time lock or DTL, which is used exclusively by Mark of the Unicorn's Performer sequencing package. Hardware support for this format is limited and it is less popular in the UK than in its native USA.

Time Code And Noise Reduction

Tape noise reduction can affect the reliability of time codes, so it is best to switch it off on your sync track if at all possible. Multitracks with a dedicated sync input and output usually bypass the noise reduction and EQ controls automatically. Dbx noise reduction systems have the greatest effect on time code, whereas Dolby C or Dolby S don't usually cause any problems.

You may need to experiment with the code level, since if recorded too high, it may leak over to the adjacent tracks. On the other hand, if it's recorded at too low a level, it might not read reliably, causing the sequencer to 'glitch' or even stop.

In order to use a sync unit with a multitrack tape machine, it is vital that the machine either has a dedicated sync in and out or that you can access the output from your sync track directly without it having to be added into the stereo mix from the multitracker — time code is very unmusical!

More About SMPTE

SMPTE stands for the Society of Motion Picture and Television Engineers. Strictly speaking, the code covers only the American TV format of 30 frames per second (fps) and film at 24fps, the European equivalent being the EBU code, which includes 25fps TV. In practice, these are all rolled into one under the title SMPTE/EBU, but the EBU part is often dropped for convenience. Apart from the more common 24, 25 and 30fps formats, the standard also includes 'drop frame' which is used when converting film to TV. The system gets its name because whole frames of picture are periodically discarded to eliminate cumulative timing errors which would otherwise cause noticeable sync problems. Drop frame is not used in audio‑only applications.

It is normal to set the SMPTE format to the local TV standard, which in Europe means 25fps. In the USA, this would be 30fps. Because individual frames of picture are too coarse a measure for audio, additional resolution is gained by dividing individual frames up into smaller units of time.