Late last year, an AMPS Technical Subcommittee consisting of myself, Nick Lowe, Eddie Joseph, Chris Munro and Anthony Faust, met to discuss the 24/25 Frame Non-linear Post Production problems. Briefly, the main problem arises because the UK Television System runs at 25 fps and Telecine transfers from 24 fps Film to Video have to be either speeded up to 25 fps (a frame for frame transfer) or have an extra 'phantom' frame inserted electronically by the Telecine every second to maintain the correct picture and sound speed. Unless careful numerical databases are kept, the latter method can cause problems with picture editing, particularly at the Negative Cutting stage. Non-linear workstations can digitise a speeded up frame for frame transfer and then switch to work correctly at 24 fps. However, the sound is usually complicated by the fact that it must first be 'speeded up' from 24 to 25 frames to match the Video Transfer. The method normally used is to transfer the sound at normal sync speed to an Audio workstation using the original Timecode as a sync reference, then electronically 'bending' the sound by 4.16666666% in the workstation to bring it up to 25 frames to match the video. The sound and picture can now be synced up to the clapper boards and laid back to the Video Tape for the Picture Editor. Again, unless original sound Timecode is logged against the Video Timecode and complex calculations are applied to the database, it is difficult for the Sound Editor to return to the original tapes. For this reason, DAT 'clones' are usually made from the Audio Workstation, using the (new) Video Timecode. These are usually done at the 'wrong' speed and the Sound Editor must make the necessary adjustments before he can use them. At this point, my brain hurts... Anyway, the Subcommittee meeting started with a pint of beer and a general discussion on whether AMPS should report and comment on various post-production routes or recommend a preferred route. The conclusion was that AMPS should recommend the route best suited for sound quality and post-production, based on the experiences of the membership. Firstly, it was agreed that the synchronisation of sound and picture is always better when carried out by a trained Assistant Film Editor. The point was made that, traditionally, when the syncing up is finalised in the cutting room, it is never touched again, therefore edge and rubber numbers are the absolute reference back to the original sound and negative. However, syncing up | done by others, notably some Telecine operators, is not as accurate and causes problems in the EDL and FlexFile databases if adjustments have to be made later. In order to synchronise rushes directly on the non-linear editing system, it is necessary for the Assistant Film Editor to have time on the system each morning to carry out the task of transferring the sound to the system. This should be done digital to digital and thus maintain the original sound levels and quality. However, no other editing work can occur during this period if only one editing workstation is available. It is recommended that a second workstation without data storage be added to the editing room, networked to the main system with access to the data storage. This should be considerably cheaper to hire than another complete system. This would allow the syncing up of rushes simultaneously without interruption of other work in the cutting rooms. The additional workstation also has the added production advantage of the Assistant Film Editor having independent and simultaneous access to the editing process for such processes as logging and making playoffs to videotape of cut sequences for review outside the cutting room. It also gives the Assistant Film Editor hands-on experience of the non-linear editing system. The main advantage of direct digitisation to the same system as the picture is that the database information relating to picture and sound is kept together in a common format. Therefore no EDL anomalies occur at a later date, not to mention the frustration of having to sort out some of the cross-platform problems that still occur. Using high quality digital to digital audio transfers means that there is no need to re-conform later and the Sound Editor can work from the already loaded material, saving considerable time and money. Any transfer and synchronising system that 'distorts' the original sound should be discouraged. For example, non-standard sample rates in Digital such as 48048 Hz to force the pull down required for NTSC synchronising in Telecine or the technique of using an audio workstation to speed up the sound by 4.166666% to synchronise 24 fps sound to 25 fps videotape transfers in PAL systems. This latter method was discussed in detail and it was pointed out that the resulting loss of Timecode information integrity caused many problems in post-production. Although software is available to make the necessary adjustments, it is felt that this is a further complication to the already complicated databases required for EDLs. |
