Chapter 15

Timecode and synchronisation

The boundaries between audio and video operations are less clear these days, and subjects such as timecode which used to be almost universally the domain of the video engineer are now as pertinent to the audio engineer. Timecode is used widely in the audio post-production industry for synchronising machines and providing a real-time positional reference on tapes. It is used in video editing and in the editing of digital audio recordings, and it is used in hard-disk recording systems for compiling edit lists and for synchronisation. Many modern analogue tape recorders have timecode facilities as do professional digital recorders, some even being equipped with ‘chase’ synchronisers.

In the following chapter the basics of timecode and machine synchronisation are discussed, but omitting discussion of the many systems which have been used in the past (and are still used in certain cases) for the synchronisation of film systems. MIDI Timecode (MTC) is discussed in Chapter 14.

SMPTE/EBU timecode

The American Society of Motion Picture and Television Engineers proposed a system to facilitate the accurate editing of video tape in 1967. This became known as SMPTE (‘simpty’) code, and it is basically a continuously running eight-digit clock registering time from an arbitrary start point (which may be the time of day) in hours, minutes, seconds and frames, against which the programme runs. The clock information is encoded into a signal which can be recorded on the audio track of a tape. Every single frame on a particular video tape has its own unique number called the timecode address and this can be used to pinpoint a precise editing position.

A number of frame rates are used, depending on the television standard to which they relate, the frame rate being the number of still frames per second used to give the impression of continuous motion: 30 frames per second (fps), or true SMPTE, was used for monochrome American television, and is now only used for CD mastering in the Sony 1630 format; 29.97 fps is used for colour NTSC television (mainly USA, Japan and parts of the Middle East), and is called ‘SMPTE drop-frame’ (see Fact File 15.1); 25 fps is used for PAL and SECAM TV and is called ‘EBU’ (Europe, Australia, etc.); and 24 fps is used for some film work.

Each timecode frame is represented by an 80 bit binary ‘word’, split principally into groups of 4 bits, with each 4 bits representing a particular parameter such as tens of hours, units of hours, and so forth, in BCD (binary-coded decimal) form (see Figure 15.1). Sometimes, not all four bits per group are required – the hours only go up to ‘23’, for example – and in these cases the remaining bits are either used for special control purposes or set to zero (unassigned): 26 bits in total are used for time address information to give each frame its unique hours, minutes, seconds, frame value; 32 are ‘user bits’ and can be used for encoding information such as reel number, scene number, day of the month and the like; bit 10 can denote drop-frame mode if a binary 1 is encoded there, and bit 11 can denote colour frame mode if a binary 1 is encoded. The end of each word consists of 16 bits in a unique sequence, called the ‘sync word’, and this is used to mark the boundary between one frame and the next. It also allows the reader to tell in which direction the code is being read, since the sync word begins with 11 in one direction and 10 in the other.

This binary information cannot be recorded to tape directly, since its bandwidth would be too wide, so it is modulated in a simple scheme known as ‘bi-phase mark’, or FM, such that a transition from one state to the other (low to high or high to low) occurs at the edge of each bit period, but an additional transition is forced within the period to denote a binary 1 (see Figure 15.2). The result looks like a square wave with two frequencies, depending on the presence of ones and zeros in the code. Depending on the frame rate, the maximum frequency of square wave contained within the timecode signal is either 2400 Hz (80 bits × 30 fps) or 2000 Hz (80 bits × 25 fps), and the lowest frequency is either 1200 Hz or 1000 Hz, and thus it may easily be recorded on an audio machine. The code can be read forwards or backwards, and phase inverted. Readers are available which will read timecode over a very wide range of speeds, from around 0.1 to 200 times play speed. The rise-time of the signal, that is the time it takes to swing between its two extremes, is specified as 25 µs 5 ±µs, and this requires an audio bandwidth of about 10 kHz.

Figure 15.1   The data format of an SMPTE/EBU longitudinal timecode frame

Figure 15.2   Linear timecode data is modulated before recording using a scheme known as ‘bi-phase mark’ or FM (frequency modulation). A transition from high to low or low to high occurs at every bit-cell boundary, and a binary ‘1’ is represented by an additional transition within a bit cell

There is another form of timecode known as VITC (Vertical Interval Timecode), used widely in VTRs. VITC is recorded not on an audio track, but in the vertical sync period of a video picture, such that it can always be read when video is capable of being read, such as in slow-motion and pause modes. This code will not be covered further here.

Recording timecode

Timecode may be recorded or ‘striped’ on to tape before, during or after the programme material is recorded, depending on the application. In many cases the timecode must be locked to the same speed reference as that used to lock the speed of the tape machine, otherwise a long-term drift can build up between the passage of time on the tape and the measured passage in terms of timecode. Such a reference is usually provided in the form of a video composite sync signal, and video sync inputs are increasingly provided on digital tape recorders for this purpose.

Timecode generators are available in a number of forms, either as stand-alone devices (such as that pictured in Figure 15.3), as part of a synchroniser or editor, or integrally within a tape recorder. In large centres timecode is sometimes centrally distributed and available on a jackfield point. When generated externally, timecode normally appears as an audio signal on an XLR connector or jack, and this should be routed to the track required for timecode on the tape recorder. Most generators allow the user to preset the start time and the frame-rate standard.

Timecode is often recorded on to an outside track of a multitrack tape machine (usually track 24), or a separate timecode or cue track will be provided on digital machines. The signal is recorded at around 10 dB below reference level, and crosstalk between tracks or cables is often a problem due to the very audible mid-frequency nature of timecode. Some quarter-inch analogue machines have a facility for recording timecode in a track which runs down the centre of the guard band in the NAB track format (see ‘Mono, two-track and stereo formats’, Chapter 6). This is called ‘centre-track timecode’, and a head arrangement similar to that shown in Figure 15.4 may be used for recording and replay. Normally separate heads are used for recording timecode to those for audio, to avoid crosstalk, although some manufacturers seem to have circumvented this problem and use the same heads. In the former case a delay line is used to synchronise timecode and audio on the tape.

Figure 15.3   A stand-alone timecode generator. (Courtesy of Avitel Electronics Ltd)

Professional R-DAT machines are often capable of recording timecode, this being converted internally into a DAT running-time code which is recorded in the subcode area of the digital recording. On replay, any frame rate of timecode can be derived, no matter what was used during recording, which is useful in mixed-standard environments.

In mobile film and video work which often employs separate machines for recording sound and picture it is necessary to stripe timecode on both the camera’s tape or film and on the audio tape. This can be done by using the same timecode generator to feed both machines, but more usually each machine will carry its own generator and the clocks will be synchronised at the beginning of each day’s shooting, both reading absolute time of day. Highly stable crystal control ensures that sync between the clocks will be maintained throughout the day, and it does not then matter whether the two (or more) machines are run at different times or for different lengths of time because each frame has a unique time of day address code which enables successful post-production syncing.

The code should run for around 20 seconds or more before the programme begins in order to give other machines and computers time to lock in. If programme is spread over several reels, the timecode generator should be set and run such that no number repeats itself anywhere throughout the reels, thus avoiding confusion during post-production. Alternatively the reels can be separately numbered.

Figure 15.4   The centre-track timecode format on quarter-inch tape. (a) Delays are used to record and replay a timecode track in the guard band using separate heads. (Alternatively, specially-engineered combination heads may be used.) (b) Physical dimensions of the centre-track timecode format

Synchronisers

Overview

A synchroniser is a device which reads timecode from two or more machines and controls the speeds of ‘slave’ machines so that their timecodes run at the same rate as the ‘master’ machine. It does so by modifying the capstan speed of the slave machines, using an externally applied speed reference signal, usually in the form of a 19.2 kHz square wave whose frequency is used as a reference in the capstan servo circuit (see Figure 15.5). The synchroniser is microprocessor controlled, and can incorporate offsets between the master and slave machines, programmed by the user. It may also be able to store pre-programmed points for such functions as record drop-in, drop-out, looping and autolocation, for use in post-production.

Chase synchroniser

A simple chase synchroniser could simply be a box with a timecode input for master and slave machines and a remote control interface for each machine (see Figure 15.6). Such a synchroniser is designed to cause the slave to follow the master wherever it goes, like a faithful hound. If the master goes into fast forward so does the slave, the synchroniser keeping the position of the slave as close as possible to the master, and when the master goes back into play the synchroniser parks the slave as close as possible to the master position and then drops it into play, adjusting the capstan speed to lock the two together. A full-featured chase synchroniser is pictured in Figure 15.7.

Figure 15.5   Capstan speed control is often effected using a servo circuit similar to this one. The frequency of a square wave pulse generated by the capstan tachometer is compared with an externally generated pulse of nominally the same frequency. A signal based on the difference between the two is used to drive the capstan motor faster or slower

In fast wind modes, a chase synchroniser will tend not to read timecode, since the tape is not normally in contact with the heads and the timecode reader may not be able to read code at wind speeds, so it reads tachometer pulses from the tape machine’s roller guide, transferred over the remote interface. The synchroniser will be programmed so as to count the correct number of tach pulses per second for each machine (they tend to differ considerably) or it may be able to work this out automatically during the first few seconds of operation. When the machine goes back into play it reads timecode again and adjusts its estimation of its position, which should be fairly close to that worked out from the tach pulses. The synchroniser then uses the difference between the master and slave timecode values, plus or minus any offset, to speed up or slow down the slave in order to lock it closely to the master.

Figure 15.6   A simple chase synchroniser will read timecode, direction and tachometer information from the master, compare it with the slave’s position and control the slave accordingly until the two timecodes are identical (plus or minus any entered offset)

Figure 15.7   A modular chase synchroniser with serial bus control facilities. (Courtesy of Audio Kinetics UK Ltd)

Such a synchroniser could be used to lock two multitrack recorders together, for example in order to increase the number of available tracks, or it could be used to slave a quarter-inch machine to a VTR for laying off or laying back stereo sound tracks in video editing. It should act as an almost invisible link between the machines and should require little attention. The initiation for chasing should not need to come from the user; the slave should start to move as soon as it sees timecode move from the master. Some chase synchronisers of this sort will even work if no remote connection is made to the master, simply chasing the timecode presented to its input (which could have come from anywhere). Systems vary as to what they will do if the master timecode drops out or jumps in time. In the former case most synchronisers wait a couple of seconds or so before stopping the slave, and in the latter case they may try to locate the slave to the new position (this depends on the type of lock employed, as discussed in Fact File 15.2).

Occasionally a machine may be fitted with an internal chase synchroniser which locks to a timecode input on the rear of the machine. It may also have a built-in timecode generator.

Full-featured synchroniser

In post-production operations a controller is often required which offers more facilities than the simple chase synchroniser, such as the example pictured in Figure 15.8. Such a device may allow for multiple machines to be controlled from a single controller, perhaps using a computer network link to communicate commands from the controller to the individual tape machines. In some ‘distributed intelligence’ systems, each tape machine has a local chase synchroniser which communicates with the controller, the controller not being a synchroniser but a ‘command centre’ (see Figure 15.9). The ESbus is a remote control bus used increasingly in such applications, designed to act as a remote control bus for audio and video equipment.

The sync controller in such a system will offer facilities for storing full edit decision lists (EDLs) containing the necessary offsets for each slave machine and the record drop-in and drop-out points for each machine. This can be used for jobs such as automatic dialogue replacement (ADR), in which sections of a programme can be set to loop with a pre-roll (see Fact File 15.3) and drop-in at the point where dialogue on a film or video production is to be replaced. A multitrack recorder may be used as a slave, being dropped in on particular tracks to build up a sound master tape. Music and effects can then be overdubbed.

Figure 15.8   A full-featured sync controller: the Audio Kinetics Eclipse. (Courtesy of Audio Kinetics UK Ltd)

In locked systems involving video equipment the master machine is normally the video machine, and the slaves are audio machines. This is because it is easier to synchronise audio machines, and because video machines may need to be locked to a separate video reference which dictates their running speed. In cases involving multiple video or digital audio machines, none of the machines is designated the master, and all machines slave to the synchroniser which acts as the master. Its timecode generator is locked to the house video or audio reference, and all machines lock to its timecode generator. This technique is also used in video editing systems.

Recommended further reading

Ratcliff, J. (1995) Timecode: A User’s Guide. Focal Press