Appendix B

Edit decision lists

As stated earlier, EDLs are one of the options of output from a nonlinear system. This is still a popular way of working. That is, a Sow-cost nonlinear system performs the off-line edit and an EDL is generated and sent with the source tapes to the on-line linear tape edit suite for finishing or auto-conforming. All professional nonlinear systems offer the ability to generate EDLs. Some use stand-alone software packages and others have integrated applications. Whichever is available, certain guidelines need to be followed. If not, there is the risk that the on-line edit will become an expensive and embarrassing exercise.

If working on a programme which is going to be edited as an off-line project and subsequently going to be auto-conformed in a tape-to-tape suite, then preparation for the on-line starts at the point that you start the off-line. The reason is quite simple – you are going into a tape-to-tape suite, so all your media, that is graphics, voice-overs, music, must be on timeeoded tapes. Any media arriving on floppies, audiocassette or ¼″ tape should be compiled onto a timeeoded source tape before loading into the nonlinear system for the off-line edit. If this is not done then the EDL created will be unusable by the on-line edit suite.

Another point to remember is that most on-line edit controllers do not like tape names with alphanumerics. While some can handle alphanumerics, all can cope with simple numerics. So, wherever possible, organize source tapes with a simple numbering system. In fact some older controllers will only support a two-digit numbering system, i.e. 01, 02, 03, …, 99. To play safe and keep things simple, label all tapes with a chinagraph pencil and where necessary create a basic look-up table of the original tape name and the numbering system that has been used. As with all things, preparation is everything!

Assuming the above preparation has been done we can now look at what an EDL is and examine some of the limitations. An EDL is little more than a list of text statements indicating which shot from which tape was used and where it is on the programme tape. When it comes to effects, the dissolve and, maybe, some wipes might be supported in the on-line suite. However, digital effects and multiple layers of video will not be.

On-line edit suites frequently have equipment from a number of manufacturers, from the edit controller and vision mixer to the digital video effects device (DVE). As a consequence the level of communication between all this equipment will be at a very primitive level. This means that EDLs of necessity have to be very basic.

Before making an EDL there are a number of questions that need to be asked. For example:

It is important to know what edit controller is going to be used, as edit controllers from different manufacturers are basically incompatible. So the list must be made in the right format for the edit controller in question, be it Sony, Ampex, Paltex, CMX, GVG (and this list is by no means exhaustive!). In addition it might be worth knowing what level of software the edit controller has installed. One area where many first-time off-line editors get caught is the floppy disk size. Quite a few edit controllers still only accept EDLs on a double density disk (800 kB). Trying to get hold of a double density disk can sometimes be difficult – most floppy disks nowadays are high density disks (1.4 MB), which cannot be read by some edit controllers. If faced with this problem, it is possible to take a high density disk and mask off the hole which does not have the write-protect tab. The Mac or Wintel machine will see this disk as a double density disk. You will need to format the disk as a 800 kB disk and then save your EDL to the disk.

Knowing what vision mixer is used might be helpful as some do conform to SMPTE wipe codes. If not, be prepared to add comments to the EDL for the on-line editor.

Sometimes on-line editors will request a particular sort mode for the EDL, based on the length of the programme and the source tapes. The different sort modes are as follows.

Sort mode A

This is the simplest sort mode to understand, but not necessarily the fastest to perform in the on-line edit. The A sort mode is linear, so that when the auto-conform begins it records the first shot of the show, then the second, and so on until the programme is complete. Any effects, for example dissolves, are performed as they are encountered from the EDL.

Sort mode B

This sort mode might be preferred when the programme length is longer than the source tapes. It is primarily of value for long form shows. During the auto-conform a source tape is loaded and it fast forwards and rewinds, finding the required shots and dropping them in the appropriate place on the record master tape. While potentially faster than A mode it can be confusing to watch, as the record master will initially have a lot of black holes, which can be disconcerting for both client and the editor. However, if the EDL is ‘solid’ then there are speed advantages.

Sort mode C

In this mode the record tape shuttles up and down, grabbing the shots it requires as the source tape steps through from beginning to end. This process is typically used when the source tape is longer than the record master, for example when compiling a movie trailer, or a promotions menu where three drama series are being previewed in a one-minute trailer.

Sort modes D and E

These are variations on the B and C sort modes. Here the effects are all performed as the last edits of the auto-conform. It is frequently not a popular choice unless the on-line edit suite is very well lined up, because the D and E modes do not track which videotape player made an original edit. This means that at the end of the list when the effects are created, the edit controller may well ask for a source tape to be loaded into a different playback machine from the original edit. If machines are not all perfectly matched and the tape line-up performed to the same exacting level, the match frame edit required for the effect will not be invisible.

Modes B to E are sometimes referred to as checkerboard editing, due to the record master having lots of black holes at the beginning of the process. Nevertheless, for a solid EDL, time can be saved.

Further considerations

As nonlinear systems support multiple vision and audio tracks, it may be necessary to create several EDLs because the on-line edit controller will not support multiple tracks in the EDL In the case of multiple vision tracks, an EDL for each additional vision track will need to be created. In the on-line suite these additional EDLs will be auto-conformed first to create a series of sub-masters, which can then be played in during the on-line edit as required. For audio, up to four audio tracks can usually be supported, although older controllers may only support two audio tracks.

There is a need to be careful when creating motion effects. Nonlinear systems typically support a greater range than can be supported by a single pass on a videotape deck. There will also be the need to create a duplicate reel list. This list is required for all those dissolves, etc. that occur between two shots which exist on the same source tape. For a tape-to-tape suite to perform a simple dissolve in this situation a copy of one of the shots must be on a separate tape. If a digital suite is to be used then the duplicate reel list might not be required as most digital tape decks support what is known as pre-read. This facility allows a digital recorder to play back before the record head (or pre-read), so that the digital recorder can replay the end of the last shot and record the same information as the dissolve to the new source takes place. However, once this is done it is a major task to change the transition.

The most important thing to remember is that changes should not be made in the on-line – on-line edit suite time is not cheap. To maximize the benefit of the automated process of auto-conforming on tape the EDL must be well prepared, bearing in mind the limitations of on-line tape-to-tape linear editing.