CHAPTER 39

EDITING THE FIRST ASSEMBLY

 

THE PHYSICAL PROCESS FOR FILM

MACHINERY

For those unused to handling film, the best type of editing machine is undoubtedly a flatbed table editor such as the Steenbeck (Figure 39-1). Easy to thread, it keeps sound and picture in constant sync but allows either to be moved in relation to the other. Sound and picture are simply spliced into the left side of the film's passage through the machine.

SOUND TRACK EVOLUTION

Until the fine-cut stage, the sound track is dialogue only and is assembled into a single track. This is a compromise, for within a single sequence several mike positions may be cut together, and the sound may vary in level and acoustical quality. Because the priority at this stage is to achieve a correct dramatic balance, the simplest assembly method is used to allow rapid editing changes during the lengthy and experimental business of achieving a fine cut. Many table editors permit two or more sound tracks, but editorial changes become slower when you have to adjust more than one sound track every time you shorten a cut or transpose two sequences.

Once a fine cut is achieved, dialogue tracks are split apart into separate tracks to allow the appropriate control in the sound mixing stage over each mike position in dialogue. Sound effects (FX), music, and atmosphere tracks will be laid as appropriate. Some sequences may need post-synchronizing (also known as looping or automatic dialogue replacement [ADR]). Original dialogue tracks recorded near an old house next to a noisy expressway, for instance, may not be very intelligible and will have different amounts of traffic noise on each camera angle.

image

FIGURE 39-1

Steenbeck flatbed table editing machine.

A mix chart is made of all tracks laid, especially because the sound editor will probably have laid extra tracks to allow for experimental alternatives. Editor, chart, and tracks then go to a sound studio where specialists make the final mix under the director's and editor's guidance. The final mix, which does so much to condition where spectators direct their imagination, determines much of the film's force with an audience.

PICTURE EVOLUTION

The final-cut workprint picture will eventually go to a conformer, who very carefully cuts the camera original negative to exactly match what is by now a very tired and beaten up workprint. The conformed negative sections are joined by cement splices, each of which requires an overlap. While the workprint was simply cut and butt-spliced with an adhesive tape splicer, cement splicers need a portion of overlap and therefore lose part of the next frame of camera original wherever a splice is made. And so, when editing film workprint, you must never use adjacent sections of film without dropping out three frames between them. These unused workprint frames will ensure that an allowance of frames exists in the camera original to effect cement splices with their overlap requirements. Keep the three frames from the workprint (in a “cutlet” can) in case during editing you later decide to reconstitute the two workprint shots back into one.

CAUTION: Forgetting this means certain disaster. When cutting workprint, never use adjacent footage without a minimum of three frames separation. This allows a conformer to overlap-splice the camera original to match the butt-spliced workprint. Nothing else will work.

In the lab, the negative is conformed to the workprint not as a single strand of spliced negative, but as two checkerboard rolls, each containing every other shot, with black spacing between shots that produces a checkerboard appearance in the synchronizer. The procedure, also called the A/B roll printing process, uses at least two rolls called Roll A and Roll B, with possibly a C Roll if titles or other matter are to be superimposed. Checkerboarding allows:

  • Resetting the printer light and color filtering during the black spacing ready for the next shot instead of these changes taking place during the first frames of the shot
  • Black spacing that hides the unsightly overlap portion of 16 mm cement splices, which once used to appear in the first frame of every new shot
  • Picture overlaps so that cross fades, or dissolves, can be made during printing

A photographic sound negative is prepared from the magnetic master mix, ready to combine with the picture. Next, the A and B rolls of negative are timed or graded (color- and exposure-graded), then passed through a contact printer where negative and positive stock are passed under a light during emulsion-to-emulsion contact with each other to make a contact positive print. Occasionally there will be additional picture rolls, should titling or subtitling require it. The A/B roll printing process allows a print with no splice marks showing and low-cost dissolves and fades. In fact, the print stock always passes at least three times through the contact printer as follows:

First pass: Roll A picture
Second pass:

Roll B picture (alternate shots printed in Roll A's spaces left by its black spacing)

Third pass:

Sound negative (sound photographically printed on the edge of the film).

During the multiple passes, the machine can be programmed to make fade-ins and fade-outs. In the overlap portion one pass fades out, then at the next pass the other scene fades in, producing an inexpensive and reliable cross fade, otherwise known as a dissolve.

COMPOSITE PRINTS

The resulting composite print (sound and action married together on one print) shows the result of adjustments made for inequities in color or exposure of the original negative. This first print is called the answer print because it demonstrates the viability of these changes. If perfectly acceptable, it becomes the first release print. Usually further changes are required, and these are incorporated in the next answer print. The first acceptable print becomes the first release print.

THE PHYSICAL PROCESS FOR VIDEO

HARDWARE AND SOFTWARE

Linear Editing: This method, and the large tape players and recorder that it requires, have practically vanished, but for anyone still using the method, a rundown follows. The most common format for offline video editing machines was 3/4" U-Matic, but 1/2" VHS editing and later Hi8 became standard. When you shoot in one format and want to edit in another, you simply dub (electronically transfer) from one to the other before editing. Though there is some quality loss, especially in picture from one generation to the next, this is a minor consideration because offline editing leads eventually to final conforming by way of an online postproduction edit.

Nonlinear Editing:  The professional standard is presently the Avid company's software, which uses a Macintosh computer, digitizing and other hardware, and a large hard drive as the center of its operations (see www.avid.com). The Avid Film Composer edits digitally from a 24fps original (meaning 24 frames per second) and will minimize pull-down problems. Video acquisition is moving to 24 P (which is 24fps high definition video) in progressive scan format because it allows a simple transfer to film. This frame rate, originally chosen for being fairly low and flicker-free, is likely to remain standard for future theater projection systems because it constitutes the entire history of film to date.

Many excellent competitors to Avid exist, such as Media 100 (see www.media100.com) and Final Cut Pro for Macintosh computers (see www.apple.com/finalcutpro/), which handles Digital Video (DV) at normal or high definition for a fraction of the cost of Avid. For IBM PC there is the superbly ergonomic Lightworks and its sister Heavyworks (see www.lwks.com), which were designed by film editors instead of computer people, Adobe Premiere (see www.adobe/motion/main) and a host of other systems.

Because everything in picture and sound is retrieved by timecode numbers, you will need a good logging system to keep track of your materials. This may be a separate logging database, or it may be incorporated into your editing software. Once clips have been arranged on the timeline and all the minute decisions made to produce a fully effective film, the system will list what you have done as an EDL, ready to program the online machine. The most advanced systems can edit at broadcast quality, but usually editing is carried out at a lower resolution so a large body of dailies can initially be stored on the hard drive. As high definition television (HDTV) takes over, all existing video equipment will become obsolete.

EDITING MECHANICS

Linear or cassette-to-cassette editing is quite simple in principle, and the mechanics do not take long to master. Essentially you pick sections from cassettes played by the source machine and transfer each chosen section to the record machine. This involves picking an out-point to material already cut, and then determining in-and out-points for the new section to be added on. When you press the record button, you see it all on the monitor: the last seconds of previously compiled material and then the new segment cut onto the old.

Timecode-driven machines give frame-accurate editing, and each preview is the same. With control-track editing (where the machine syncs to a count of the sine wave control track), editing decisions are plus or minus a few frames. Worse, as you preview, the decision point may drift even further.

Nonlinear or digital editing (NLE) offers great advantages in speed of execution and in sound quality. Virtually any imaginable experiment with your footage can be tried in seconds once you know how to use the program. The best editors are said to be those who learned their aesthetics from film, but film editors seem to unanimously approve of nonlinear's advantages. Optical effects, such as slow or fast motion, freeze frames, titling, fades, or dissolves are easy to specify in video form, but they become very expensive and difficult to specify when you execute them in film.

SOUND CONSIDERATIONS

NLE programs of any sophistication have multiple sound tracks so you can split them and lay them in parallel according to their origin and purpose in the movie. Movie sound postproduction has always included provisions for mixing multiple tracks together, but until recently they were rolls of film. Now they are tracks originating from a diagram on the monitor screen. Sound should mostly be seamless, even when recording conditions produce components that do not cut together seamlessly at all. In a dialogue scene, for instance, there may be four different microphone positions, each having a slightly different level, coloration, and resonance. Played without any adjustments they sound truly awful and fragment the scene for the audience. Plainly they must all come to sound like one track. This is done by placing all the close shots on one track in the sound timeline, all the two-shots on another, all the long shots on a third, and so on. Then you can set levels and tone equalizations for each mike position to make all the tracks sound acceptably similar. You will, of course, have to allow for sound perspective because we expect sound to vary in quality according to how close or distant we are from the sound source, such as people speaking.

Now you can view the film under optimal sound conditions during the edit and make dramatic and other decisions based on a close knowledge of their potential. Having multiple tracks available also allows you to lay in wild (that is, non-sync) sound like doorbells, car horns, or various atmospheres, exactly where they belong. They, too, can now play their part. Often a sound effect helps to narrate the story, and editing pace will be affected by its presence.

You can also lay in specially recorded presence track when you background from one angle to another. There may be a match in the foreground voice recording but a mismatch in amounts of ambient sound (background noise inherent to the location). Remember that you can always add presence to quieter tracks, but cannot subtract it from the louder ones.

If you have the right music at hand while editing, this too can be laid in. Now you can make compositional decisions concerning scene rhythm and duration that would be quite uncertain if taken on a theoretical basis. This so often happens with feature films, which record their music last.

MINIMIZING GENERATIONAL LOSSES

Copying Limitations: In analog sound, repeated copies and premixes in analog recording (that is, non-digital audio and video) led to deterioration from generation to generation that was audible as an increased hiss level and diminished fidelity. Likewise, between generations of analog picture, losses were also disturbingly evident as increasing picture noise (picture break-up particularly in the color red), color shift (in which red in particular moves across the screen, like bad newspaper printing), and an overall deterioration in color fidelity and acuity (sharpness of detail). Anyone who has ever copied a VHS tape knows this kind of image degradation intimately.

Digital copying, however, can run to 30 or 40 generations before any discernible deterioration shows. This is because waveform information is recorded as a stream of on/off or binary pulses that are very robust compared with the fine gradations of voltage that determine analog recording. With re-recording, these quickly lose their integrity, while the same waveform made up in graph fashion by the binary zeroes and ones of digital recording is very enduring.

Digital Copying:  In practice, most consumer and prosumer (high-end consumer) camcorders now employ Firewire technology. This ingeniously uses a single cable, called an i.LINK 1394 DV in/out cable, to link together camcorders, recorders, or computers, much as USB cables do with computer peripherals and MIDI cables do with computerized music. A Firewire link not only sends picture, sound, and timecode information in either direction, it can also transmit deck control information from your computer to your camcorder or record/replay deck. The ability to copy back and forth, virtually free of generational loss, is a quantum leap for digital technology.

Compression:  All is not roses, however, because digital recording usually involves different types and degrees of compression, which involves discarding some original data. This is done by a timebase that compares each new frame, be it picture or sound, and records the difference rather than all the information in the whole frame all over again, as analog recording does. Compression is accomplished by using one of the industry-determined compression algorithms called codecs. Those that involve more loss than others are called lossy codecs. Lossless recording and re-recording is (like most things in the electronic world) just around the corner and promises ever better quality, provided the computer industry can produce ever faster and more capacious computers at a price mere mortals can afford. Under optimal conditions, projected HDTV is difficult to distinguish from 35 mm projection and will obviously just keep getting better.

TIMECODE: NTSC DROP FRAME AND NON-DROP FRAME

Timecoding provides every frame of picture and sound with its own elapsed time identity in hours, minutes, seconds, and frames, and every kind of video editing operation utterly depends on it. However, because of the primeval nature of American 525-line television running on a 60 Hz electrical supply, although NTSC (National Television Standards Committee) appears to have a 30 frames per second (fps) rate, it really runs at 29.97. This becomes cumulatively significant over time, so an electronic band-aid solution was devised called drop frame timecode. This arrangement drops a frame every so often to keep the displayed timecode in sync with actual time. For this reason camcorders and editing software offer the (initially bewildering) choice between drop frame and nondrop frame recording. Just remember to stay consistent and know which one you chose. The irreverent will tell you that NTSC really means Never The Same Color.

LINEAR ONLINE AND OFFLINE EDITING

The window dub is a copy of the camera original with its timecode burned in as a window (see Chapter 38, Figure 38-4). The window dub is edited in place of the precious camera original to preserve it from undue handling. The timecode window copies through to subsequent editing generations and is sometimes the last recognizable detail. Most importantly, it lets you dub from one generation to another, knowing that when cutting is complete you can transcribe an edit decision list (EDL) from the screen and then use it to drive a computerized (online) editing setup to reconstruct a pristine copy from the camera original tapes.

EDITING CONCEPTS

Basic editing can be self-taught using common sense, but both basics and the further reaches of sophistication can be best learned after some detailed analysis of finished films (see Chapter 5, Project 5-2: Editing Analysis). Most editing programs come with an instructional project of some kind, but invariably they exist not to teach you editing but to teach you how to use the software. Most young people, having ingested an average of 18,000 hours of television while growing up, take to editing like ducks to water. If you carry out the hands-on shooting and editing projects (Chapter 6: Shooting Fundamentals) and edit your results, you will learn a huge amount both about editing and shooting. Much film education is Yow learning, as in “Yow! Why didn't I shoot a [fill in the blank] shot?” During editing the chickens come home to roost.

What follows is an overview of the conceptual process of editing, with particular attention to what you can expect to find as you move stage by stage through editing a project.

THE FIRST ASSEMBLY

Putting the material together for the first time is the most exciting part of editing. Don't worry about length or balance at this stage. Adapting experimentally to the problems you encounter is everything. To make a first assembly, work on one scene at a time and put the film together in whatever order is convenient, as follows:

  • Run all the material for the scene to refresh your memory. Sit through out-takes, false starts, aborted takes, everything because there may always be something in there that you'll need.
  • Referring to the editor's marked-up script, figure out how the coverage might best be assembled. At this stage, use a lot of the master shot and leave aside close ups or double-cutting (repeatedly cutting between, say, two speakers when a single angle would adequately relay the action). Nor at this stage should you use any overlap cutting (where a speaker's outgoing dialogue, for instance, overlaps a shot of the listener before the listener replies).
  • Assemble the simplest version that is faithful to the script without trying to cure any questionable changes in the actors' pacing.
  • If you have two versions of something and both seem equally viable, include both so you can choose later in the widest possible context.

See the whole film as soon as possible in a long, loose form before doing detailed work on any sections. Of course, you will be longing to go to work on a favorite sequence, but to fix nagging details would be to avoid confronting the film's overall identity and purpose.

FIRST ASSEMBLY VIEWING

NO INTERRUPTIONS

Run the first assembly without stopping and without interruption of any kind. Make no notes because this will take your attention from the screen. You want to take in the film as an audience would.

WHAT DO YOU HAVE?

The assembly viewing will yield some important realizations about the character, dramatic shape, and best length of the film. You will also have a handle on all the performances and know what kind of overall control you may have to exert. Fundamental issues are now out of the closet. You will sense an overall slowness, that some scenes include unnecessary exposition, or hang on beyond a good ending point. You may find you have two endings, one false and one intended, or that one character is unexpectedly stronger than another. A sequence you shot in miserably cold conditions by a river at night turns out to stall the story's advance and must be dropped.

The first assembly is the departure point for the denser and more complex film to come. As a show it is long and crude, yet despite its artlessness, it can be affecting and exciting.

RUN THE FILM A SECOND TIME

Now run the film again to see how your original impressions stand up. Following further discussion with your editor, make a list together of major aims for each sequence, arranging them strictly by priority.

DIAGNOSTIC QUESTIONING

To question the imprint the film has made and predict a likely audience response, you must view the film as if seeing it for the first time. At this point you are dealing with the film in its crudest form, so the aim is to elicit your dominant reactions.

After seeing an assembly, rapidly list the memorable material. Refer to the script and make a second list of whatever left no particular memory. Why does this other material not deliver? The human memory discards quite purposefully what it does not find meaningful. All that good stuff you could not recall was forgotten because it simply did not work. This does not mean it can never work, only that it is not doing so at present. Here are some possible reasons, each suggesting a different solution.

Problem Possible Solutions

The writing is poor in comparison with other sequences.

Cut the whole scene? Shorten? Rewrite and reshoot?

Acting is at fault. Dramatic rhythms are too predictable or actors are not in character or in focus.

Help, but not a cure, is available in further editing. Very often reaction times are wrong and convey the wrong (or no) subtext.

Rebalancing these can help.

Scene outcome is predictable.

Scene structure is at fault? Too long or too slow? Maybe the scene is in the wrong place?

Two or more sequences make a similar point.

Repetition does not advance a film's argument unless there is escalation in dramatic pressure, so make choices and ditch the redundant.

Dramatic intensity plummets. A useful analogy is the idea of a rising or falling emotional temperature. To see material in its context is to see correct relative temperatures more clearly.

If your film is raising the temperature, then inadvertently lowering it before the intended peak, the viewer's response is seriously impaired. The transposition of one or two sequences will sometimes work wonders. Naturally, this can only be done if the scenes are not locked into a fixed time development.

The viewer is somehow set up by the preceding material to expect something different.

We read film by its context; if the context gives misleading signals or fails to focus awareness in the right area, the material itself will fall flat.

These suggested areas of examination are just a few involved in the kind of dramatic analysis that a playwright or theater producer uses when rehearsing a new work. Just as a playwright routinely rewrites and adjusts a work based on audience feedback, so the filmmaker makes a vast number of adjustments, large and small, before admitting that a work is finished. First you dig for your own instincts by feeling the dramatic outcome of your material. Later, when you have a fine cut and the material becomes showable, you will call in a few people whose reactions and tastes you respect. You will probably find quite a bit of unanimity in what they tell you. Because a filmmaker has only trial audiences until the work is finished, assessments are hard to make and are certainly not objective.

While still in this assembly stage you and your editor begin asking basic questions:

  • Does the film feel dramatically balanced? If you have a very moving and exciting sequence in the middle of the film, the rest of the film may seem anticlimactic.
  • Does the film seem to circle around for a long while before you feel it start to move?
  • When is there a definite feeling of a story unfolding, and when not? Asking this will help you locate impediments in the film's development.
  • Which parts of the film seem to work?
  • Which parts drag, and why? Some of the acting may be better than others. Sometimes the problem is that a scene is wrongly placed or repeats the dramatic contours of a previous one.
  • Which of the characters held your attention most, and which the least?
  • Was there a satisfying alternation of types of material, or was similar material indigestibly clumped together?
  • Which were the effective contrasts and juxtapositions? Are there more to be made?
  • Sometimes a sequence does not work because the ground has not been properly prepared, or because there is insufficient contrast in mood with the previous sequence. Variety is as important in storytelling as it is in dining.
  • What metaphorical allusions did you notice your material making? Could it make them more strongly? That your tale carries a metaphorical charge is as important to your audience as a water table is to pasture.

RESOLUTIONS AFTER SEEING THE FIRST ASSEMBLY

Once you have seen and discussed the whole ungainly epic, you and your editor can make notes of far-reaching resolutions about its future development. These may involve performances, pacing, parallel storytelling, structure, or overall meaning. Remember that your editor is massively uninterested in whatever you originally intended. That is now irrelevant, and you must come to terms with it, consigning all other intentions to history.

PRIORITIZE

When you tackle the problems that inevitably show up in any cut, arrange them by hierarchy and, as in building a scene during rehearsal, deal only with the largest faults. If the film's structure is awry, reorder the scenes and run the film again without making any refinement to individual scenes. If there is a serious problem of imbalance between two characters who are both major parts, go to work on bringing forward the character presently deficient. Correct only the major problems after each running.

LENGTH

Look to the content of your film itself for guidance over length and pacing. Films have a natural span according to the richness and significance of their content, but the hardest achievement in any art form is having the confidence and ability to say a lot through a little. Most beginners' films are agonizingly long and slow, and the advice of professionals on a film's proper length is usually painful but valuable. If you can recognize early that your film should be, say, 20 minutes long at the very most, you can get tough with that 40-minute assembly and make some basic decisions.

Remember that nonlinear editing (NLE), unlike other editing forms, can preserve every one of your cuts. You can always look back at earlier versions to decide whether there is anything you still like about the film's earlier incarnation.

STRUCTURE

Most of all, you need to find the best dramatic structure to make the movie into a well-told tale. A good screenplay does not guarantee the best experience for an audience because during production and editing other criteria come into play. These result from the emotional changes and development brought by the cast and the filming itself. These are what an audience will experience, and to these you must address yourself as they become apparent. The director is always the one most encumbered by the film's history.

LEAVE THE EDITOR TO EDIT

Having decided what the next round of changes should be, leave the cutting room until summoned back. Not all editors or directors can work this way but it is important to try. The reason is simple: The editor loses objectivity while correcting the many problems, and so will any director who remains present. But a director returning with a fresh eye to see a new cut can tell the editor, whose eye is now conditioned by the work, which changes are actually working. The director is always a surrogate audience whose keenest tastes must be satisfied.

..................Content has been hidden....................

You can't read the all page of ebook, please click here login for view all page.
Reset
3.137.217.220