Film is shot at 24 fps in the camera and is projected at 24 fps in the cinema. However, in PAL video, the telecine transfers the film to video at 25 fps. This is necessary in order to maintain a one-to-one relationship between frames on the film and the videotape. However, as a result of this, the film runs a little faster (a little over 4% faster) on the videotape. By itself this pitch difference is quite perceptible. In musical terms, it is nearly half a semi-tone higher in pitch.

Speeding up sound - expensive and time-consuming
When this video footage is edited on a video editing system, the very pacing of the film during the editing process - which is without doubt where the film really comes together - is seriously affected. While some people might not think that such a small change in the speed would affect a film's pacing, one need just realise that this tiny change in speed would result in the typical Indian feature film being nearly 600 feet longer in the film version than the video version!

The second problem that results from editing at 25 fps when the material is shot at 24 fps is with the sound. The sound is recorded on location either on a Nagra or a DAT recorder in sync with the camera running at 24 fps. When the film is speeded up to 25 fps for the transfer to video, the sound must also be similarly speeded up in order to sync to the picture. Speeding up the sound in this manner and that too precisely by 4.1667% is a difficult operation that requires specialised equipment or audio software. Needless to say, this operation will be expensive and time-consuming.

Final Sound Quality
After speeding up the sound in some manner, it must be digitised into the video editing system and synced to the picture for further editing. Now, if this is live sound that will be used on the final film soundtrack, one must be aware that the quality would have been seriously compromised by the speed change. This sound can possibly be used on a mono sound track but would not be good enough for use on a stereo or digital sound track.

If the live sound is only a pilot track and the film must be dubbed, then a work print or a video output of the film must be made for the audio post-production. If a work-print is made and then used for dialogue dubbing, all recording will again be at 24 fps. Now, in order to take this dialogue back into the editing system for the final trim of the film, even this audio must be speeded up thereby reducing its quality as well. That will make it difficult to have a high quality dialogue track for the final release on a stereo or digital format.

If a video copy is used for the audio post-production, then all work will happen at 25 fps. This means that the audio quality will be preserved when it is taken back into the non-linear editing system. However, after the final mix, when the audio is transferred to optical negative, it will run slower by 4%. This will result in songs that are slower in tempo and lower in pitch as well as dialogues that are bassy and seem to be dragging. Again, this will most certainly affect the final product's reception by the audience.

Why Match back software?
Considering the problems in sound sync and picture pacing, why did someone write software for matching EDLs to negative edge numbers? The answer lies in the fact that such problems do not exist for film transferred to NTSC video. When a telecine converts film to NTSC video, it still runs at 24 fps and repeats frames on video to achieve an effective rate of 30 fps. Thus, the film pacing and the sound sync are not much of a problem. It must be mentioned however that these repeated film frames on NTSC video (called pulldown frames) are a problem to identify and remove when creating a cut-list from the EDL. These pulldown frames are the reason why NTSC based cut-lists are not frame-accurate.

If the film is shot at 25 fps for transfer to PAL video, then the sound need not be manipulated for syncing to the video. However, such 25 fps projects would still play slower than the original in a theatre and in addition, they will consume 4% more film raw-stock. For these reasons, films are shot in 25 fps only if they are intended for television or advertising use.

When match-back software claims to be 24-25 and 25-25 compatible, it simply means that the software can handle film shooting speeds of 24 fps or 25 fps transferred to video at 25 fps. It does not mean that there is any other solution available for handling the problem of sound sync or film tempo.

A film editing system that can work at a true 24 fps speed
It will be clear from the above discussion that in the Indian environment, it is not advisable to use match-back software to edit feature film projects. The solution to editing film on a non-linear editing system is a film editing system that can work at a true 24 fps speed. The only such system available in the world today is the Avid Film Composer or Media Composer with the Film Option.

On the Avid, the telecined material at 25 fps is digitised into the system. From this point on, the Avid plays back the material at 24 fps. The sound can thus be fed into the system at its original speed and it will be in perfect sync with the picture. When the Avid outputs a video copy for audio post-production, it plays back at an effective frame rate of 24 fps as well. Thus, audio post-production can be done in perfect sync with either a film work-print or an Avid video output with equal ease and with complete interchangeability.

This discussion has tried to examine the issue of match-back software in an unbiased manner. The fact is that Avid also has its own match-back software called MediaMatch and an option on Xpress systems called Matchback which are equally inappropriate for editing feature films in India. The purpose of this document has been to prevent users from unknowingly going down a path fraught with perils.