A few editors preferred to cut film electronically, transferring dailies to videotape and then laboriously assembling a cut by re-recording selected shots from deck to deck in the order desired. This "linear" editing suffered from one major drawback: since no videotape was physically being spliced, the editor was required, upon trimming, extending, or adding a shot anywhere but at the end of the cut material, to re-record all of the cut work that followed the change. Film, on the other hand, still retained a marvelous "non-linear" quality: one could add or extract material anywhere in the cut without affecting the footage that followed it. Computers facilitated a revolution in the editing room. By modifying home PCs such as the Apple Macintosh or DOS-based computers, Avid, Lightworks, and other companies ushered in the age of Digital Non-Linear editing. Their innovation: making it possible to digitize footage: originally onto optical disks, later onto high-capacity hard drives. Rather than cutting film or re-recording videotape, the editor merely creates files which tell the computer in what sequence to play back which pieces of digitized material. The editor can make changes at any point to any part of the sequence without affecting what follows. Non-linear has swept the editing world. From studio films to low-budget, even to student filmmaking, everyone seems to be editing on Avid or a competing non-linear system. The president of the Editors Guild wrote in 1996 that film-based editing was dead, and in a recent conversation with a technician at a major Hollywood lab, I was told that 80% of all films they process are telecined for non-linear editing. And yet, there are still plenty of flatbeds out there, and labs still print dailies on film if you want them to. (Mag stock is beginning to grow scarcer but, to our knowledge, will still be available for some time.) Cutting on film is still possible and still (at least compared to non-linear) relatively inexpensive. What advantages do digital non-linear systems offer over traditional film-based editing? The first is an enormous savings in time. Cutting film requires endless putting up of rolls, taking down of rolls, shuttling through rolls, rewinding rolls, etc. to find and view footage (or, on a Moviola show, unwinding of shots, viewing shots, re-rolling shots, etc., etc.) When the editor actually wants to make a cut, s/he must physically line up the out point on both the picture and the track to assure sync, make the cut, splice the new shot into the cut, and splice the dailies roll back together to close up the gap. Finally, when the editor wishes to extend a shot, whether by a a few frames or a few feet, s/he must find the necessary footage--which might be back in the dailies rolls, hanging from the trim bin, rolled up with unfiled trims or lifts, or (as inevitably happens, particularly during a crunch) completely and mysteriously AWOL. In non-linear, these problems disappear. The computer can instantly access and display any part of any shot at any time. Shots can be organized and re-organized without physical labor, by simply dragging shots into a new "bin" (the Avid term for a digital file folder). Cuts are made instantly; the editor need only choose an in and out point in the source material and an in or out point in the cut work (or vice versa), click a button, and voila. Extensions or trims can also be made instantly. A second advantage of non-linear systems is the increased ability to experiment. In film, if an editor wishes to re-cut a scene while retaining the option of returning to the previous version, s/he must make a dupe of the old scene, reconstitute the footage back into the dailies rolls, and recut. This is time-consuming and expensive, results in a large number of splices in the workprint, and requires another reconstitution and re-assembly if the original version is opted for. In non-linear, the editor simply creates a new file (or "sequence" in Avid parlance) for the alternate cut. Both sequences--the original and the recut--can then be viewed back-to-back if desired. One can theoretically generate any number of versions of a given sequence with no undesirable repercussions. A third advantage of non-linear systems is the ability to instantly create and view optical effects (such as fades, dissolves, wipes, titles, superimpositions, etc.). In film editing, one marks the workprint with grease pencil where an optical is desired, but the actual appearance of the effect can only be imagined. Later, the required camera negative is pulled so that an optical house may create the actual opticals; if the editor doesn't like an optical, it must be discarded or re-ordered, and the original expense of producing the optical has been wasted. Non-linear systems will create a viewable effect more or less instantly. Eventually, an optical house will still have to create optical negative with which to strike release prints; but the editor can first play with the effect on the computer until s/he is sure that the effect ordered is in fact wanted. Non-linear systems also have a number of disadvantages. The first consideration--for low-budget films in particular--is their cost. In Los Angeles, an Avid Film Composer with a relatively modest 18 GB of digital storage can rent for between $700 and $1,500 per week, while a flatbed and editing bench setup (including rewinds and splicer) should rent for considerably less. (Of course, editing on film requires full-time work by one or more assistant editors, and in low-budget, their salaries may or may not eat up the remaining difference in cost.) The second disadvantage is resolution. While some non-linear systems will digitize at "broadcast quality," systems geared toward feature film editing will generally digitize only at fairly low resolutions. At these resolutions, fine detail is lost and the image appears somewhat pixillated. This can make it difficult for an editor to assess the visual quality of a shot which s/he is considering for use in the cut. Higher resolutions consume more disk storage, which adds extra expense. Avid's Film Composer does not even offer broadcast-quality resolutions, although their Media Composer 8000 and 9000 models do offer both broadcast-quality resolutions and film-cutting capability. The third disadvantage of non-linear systems is that one cannot easily view cut material on a movie screen (or even the dailies, for that matter, since--on low-budget films at least--videotapes are frequently telecined directly from negative, and positive workprint is never made). In film-based editing, of course, one can drag one's cut reels down to the screening room any old time and have a look. In non-linear, even if a production has the budget to print dailies as well as telecining them, assistant editors must output cut lists from the Avid and manually conform the workprint to the editor's work on the Avid. In this way, cut work can be screened in the ordinary manner... but at considerable expense. A fourth disadvantage is the inevitability of downtime. Catastrophic breakdown of flatbed editors is rare; at most, typically, one might blow a lightbulb, which can be replaced in a moment. Avids, on the other hand, experience a variety of technical problems ranging from software conflicts to hardware incompatibilities to corrupted files to defective cards. Any one of these can stop editing dead in its tracks, and fixing it can in some cases take days. Even though Avids still save editing time in the long run, their occasional prolonged downtimes can cause intense frustration. Finally, non-linear systems have one last disadvantage. By allowing filmmakers to edit faster than ever before, and with their high rental costs, they are encouraging producers to put fewer and fewer weeks of picture-editing into post-production schedules. This raises a question: just because it's become possible to cut a feature film in five weeks, does that make it desirable? Traditional film editing enforces, through its physicality, a slow and deliberate pace, and produces a happy by-product: plenty of time for editor and director to reflect on the cut. Sometimes it's necessary to walk away from a cut for a little while, then to come back and screen it again with a fresh eye. Reflection over a period of a few months can be good for a film in progress. By encouraging producers to trim down the editing process, non-linear systems are unintentionally causing films to be rushed into picture-lock, whether they're completely ready or not. For better or worse, non-linear editing is clearly here to stay. How long film-based editing will co-exist with it is anyone's guess. Speaking strictly for myself, I think that the unbelievable speed and ease of non-linear editing more than makes up for its other deficiencies. After one has edited without those once-necessary evils of searching for trims or making one-frame extensions, returning to film-based editing seems like an impossible chore. But, having returned to the occasional 16mm flatbed project after editing on Avid, I found that it is possible to go back. There are certain tactile pleasures to editing on film which non-linear can't match: the feel of celluloid in your hand as you thread up a flatbed or make a splice, and the satisfying sight of cut material slowly growing in diameter around the take-up core. Of course, in the end it's all the same. As audiences view the finished film, they won't know by what technique it was edited; they will judge only whether the film touched them, moved them, entertained them, made them think. Putting together pieces of film to elicit these responses is the only thing that matters for the editor, whether s/he works on film, videotape, or non-linear.
|