5.1 FOR FILM COMPOSERS

5.1 FOR PRODUCERS

SURROUND FOR ENGINEERS-THE MAGIC PENTAGON

Simple Scenario: Syncing a MIDI Sequencer to Video with the Video as MasterTimecode Source

You might be a composer who wants to get started putting music to picture, for example. If the video or film production company have supplied a videocassette tape with SMPTE timecode recorded onto one of the audio tracks then the situation can be fairly simple. In this case you can just use a SMPTE-to-MIDI Timecode (MTC) converter. This will convert the timecode playing back from the VCR into MIDI messages which can be used to control your MIDI sequencer. When the sequencer receives the MIDI Timecode messages it will start at the correct point and stay in sync with the video frames when you press 'Play' on the VCR. In this scenario, you may be using a stand-alone SMPTE-MTC converter with a hardware sequencer, or, perhaps more frequently, a SMPTE-MTC converter which is built into a MIDI Interface for your Mac or PC which will feed the MTC directly to your MIDI sequencer software. You may even be using a stand-alone device with a built-in SMPTE
reader - such as a sampler, recorder or mixing desk - in which case the sampler will fire samples at specific SMPTE times, the mixer automation with be synchronized with the picture, and the recorder will record and play back in sync with the picture. This is probably the most straightforward syncing situation you will come across and should pose very
few problems in practice.

The whole idea of using SMPTE/EBU timecode came about because TV stations needed to lock their video tape recorders together in perfect synchronization throughout each 24 hours of operation. This is why there is a separate timecode 'address' for each second of the 24 hours. Any particular SMPTE number can be regarded as corresponding to a physical location on a video or audio tape, so it can be thought of as the 'address'of that particular location. Timecode was developed to comply with standards set by the Society of Motion Picture and Television Engineers (SMPTE) in the USA and the European Broadcast Union (EBU) in Europe. Video consists of picture 'frames' and there are 25 of these each second in Europe (30 in the USA). So each second is further subdivided into 25 (or 30) frames. And because of the way the code is designed, each frame can actually be further subdivided into 80 subframes or 'bits'. Just to muddy the waters a little here, some equipment actually displays subframes as divisions of 100 - which is a little easier for us humans to deal with than 80ths of a frame - but the underlying sub-divisions are still 80ths.
Another complexity is the variations of timecode used in the USA and some
other countries for different TV standards, with colour video actually running at a rate of 29.97 seconds for example, but we won't delve any deeper into this here as we rarely have to deal with this in Europe. It is also worth noting that some people refer to timecode running at 25 fps as EBU timecode on the grounds that this is the European standard while they refer to timecode running at 30 fps as SMPTE timecode, the American standard. Most people just call it 'SMPTE' or 'timecode' - whatever the framerate.
 

The timecode information recorded within each audio or video frame is called the 'timecode word' and each of these 'words' is divided into 80 sections called 'bits'. Each word contains a single, unique timecode address corresponding to a particular video frame or a particular digital audio 'frame' or physical location on audio tape. In the case of digital audio, the samples are further grouped into 'frames' of data having a certain length so you sometimes encounter the term 'audio frame' in this context.
 

A timecode address can be 'burned in', ie mixed with the original video signal to form a new image containing the timecode address superimposed on top of the original video, and this can often be positioned wherever you like on the screen. Boxes which can do this, such as the Digital Time Piece, contain a character generator which produces the numbers you see on the screen. It is important to make sure that the unit generating the video signal containing the visible timecode is 'genlocked' to the video recorder onto which it is being recorded. With burned-in code you can always see which frame the video is at - slowing or stopping to read it as necessary.
 

This contains quarter frame or full-frame timecode and MIDI 'cueing' or 'transport' messages to tell the system to stop, start or locate. Full-frame messages are sent during high-speed spooling where regular updating of each frame would mean sending too much data for the system to cope with. Quarter-frame messages are sent when the system is running - to
reduce the demands on MIDI bandwidth. A single frame of timecode contains too much information to be represented by a standard 3-byte MIDI message so it is split into 8 separate messages. Four of these are sent for each timecode frame so it takes eight of these 1/4-frame messages  to represent one complete SMPTE frame address. As such, MTC itself can only resolve down to quarter frames while SMPTE can resolve to 1/80 of a frame (1 bit). You may be worried that MTC is not accurate enough but it turns out that this
is not really a problem - because all MTC receivers will interpolate incoming timing data to whatever accuracy the designers of the equipment have chosen. That's why some MIDI sequencers let you specify the SMPTE times of events down to 1/80 or 1/100 of a frame.
 

MIDI Machine Control uses System Exclusive MIDI messages to remotely control MMC-equipped tape machines and VCRs. All the basic  functions such as Stop, Start and Rewind are available in all MMC systems - while some implementations offer much more detailed control.

This is a digital audio sync signal usually available on a BNC-type connector. This BNC (British Naval Connector) is the same as the one often used for composite video connections.
 

This is a word clock sync signal used by Digidesign and some other manufacturers to sync their digital audio workstations. It runs at 256 times the rate of standard wordclock so it is sometimes referred  to as 'Word Clock x 256'.
 

The timing signals within a AES digital audio stream can be supplied via a separate XLR connector which just carries timing signals - no audio - between some professional digital audio equipment.
 

Timing signals are carried within the S/P DIF data stream, but can be supplied via an S/P DIF connector with no audio data present - as on the Digital Timepiece, for example.
 

ADAT recorders send and receive digital audio sync via their opticalconnectors but can also send and receive transport control messages using aspecial multipin connector. These can be used to connect to a BRC unit or to several of the popular synchronizers from Mark of the Unicorn and others.