The credits for John Ford's My Darling Clementine (1946) include Wyatt Earp as technical consultant but only one person responsible for all of postproduction sound (the composer). The credits for Lawrence Kasdan's Wyatt Earp (1994) list the names of thirty-nine people who worked on postproduction sound. The difference is not simply a matter of expanding egos or credits.
"An older film like Casablanca has an empty soundtrack compared with what we do today. Tracks are fuller and more of a selling point," says Michael Kirchberger (What's Eating Gilbert Grape?, Sleepless in Seattle). "Of course a good track without good characters and storyline won't be heard by anyone."
With soundtracks much more dense than in the past, the present generation of moviemakers has seen an exponential growth in the number of people who work on the sound after the film has been shot. What do all those people add both technically and esthetically? "When I started out, there was one sound editor and an assistant," says picture editor Evan Lottman (The Exorcist, Sophie's Choice, Presumed Innocent). " As editor for a big studio picture in the early Seventies I usually cut the ADR [dialog replaced in postproduction--EW] and the music as well." Today an editor on a major feature would either hire a supervising sound editor who gathers a team of sound specialists, or go to a company like C5, Inc., Sound One, or Skywalker that can supply the staff and/or state-of-the-art facilities.
Sound is traditionally divided into three elements: dialog, music, and effects (any auditory information that isn't speech or music). Although much of the dialog can be recorded during principal photography, it needs fine tuning later. And almost all other sound is added during postproduction.
How does sound get on pictures? The following is a rough sketch of the procedure for a major Hollywood feature production. But it is not a blueprint; exact procedures vary tremendously with the budget and shooting schedule of the film. Blockbuster action films, for instance, often devote much more time and money to sound effects than is described below. The process certainly does not describe how the average film is made abroad; few other cultures have such a fetish for perfect lip-synching as ours--so even dialog is recorded after the shoot in many countries.
This article can only begin to suggest how digital technologies are affecting post-production sound. For one thing, there is wide variation in types of systems; for another, digital sound techniques are evolving faster than alien creatures in a science fiction movie.
Even the sound recorded live during principal photography is not wedded physically to the image and has to be precisely re-linked during postproduction. It is usually recorded on 1/4" magnetic tape (though there are alternatives) and marked so that it can be ultimately rejoined with the picture in perfect synchronization.
On the set the location recordist (listed as production mixer) tries to record dialog as cleanly and crisply as possible, with little background noise (a high signal-to-noise ratio). A boom operator, usually suspending the microphone above and in front of the person speaking, tries to get it as close as possible without letting the microphone or its shadow enter the frame.
An alternative to a mike suspended from an overhead boom is a hidden lavalier mike on the actor's chest, which is either connected to the tape recorder via cables or wired to a small radio transmitter also hidden on the actor. But dialog recorded from below the mouth must be adjusted later to match the better sound quality of the boom mike. And radio mikes can pick up stray sounds like gypsy cabs.
While on the set, the sound recordist may also ask for a moment of silence to pick up some "room tone" (the sound of the location when no one is talking), which must be combined with any dialog that is added during postproduction (with reconstructed room reverberation) so that it matches what is shot on the set. (We don't usually notice the sound of the breeze or a motor hum, but their absence in a Hollywood product would be quite conspicuous.) The set recordist may also capture sounds distinctive to a particular location to give the postproduction crew some sense of local color.
Theoretically, the first stage of sound editing is "spotting," where the editor(s) and possibly the director go through each second of the film with the supervising sound editor in order to generate a list of every sound that needs to be added, augmented, or replaced. This practice has fallen prey to demands for early previews, which have wreaked havoc on postproduction schedules.
Dialog editing is mostly a matter of cleaning up production sound. The work can be as detailed as reusing a final consonant of one word to complete another where it had been obscured or removing an actor's denture clicks.
Some of the dialog heard in the completed film was not recorded on location. Shooting silent (MOS) is much easier than having to achieve perfect quiet from the crew, the crowd watching the film, or airplanes and birds passing overhead. Even with the compliance of onlookers, nature, and ubiquitous car alarms, however, miked dialog may be unusable because it picked up extraneous noises such as a squeaky camera dolly or clothing rustle.
Despite these difficulties, directors almost always prefer production dialog, which is an integral part of the actors' performances, to looping (rerecording speech in post-production). Although there is a trend in looping sessions toward using booms and the original microphones to mimic the situation on the set, it is nearly impossible to duplicate all the conditions of the shoot. Orson Welles found that out after shooting the festive horseless carriage ride in The Magnificent Ambersons. Because the scene was photographed in an ice plant with tremendous reverberation (which wouldn't be heard outdoors), the dialog of all six characters had to be looped. When Welles heard the original looping, he rejected it because the voices were much too static; they didn't sound as though they were spoken by people in an automobile. The sound-man's low-tech solution was to redo all the lines with the performers and himself seated on a twelve-inch plank suspended between sawhorses. For a week, says James G. Stewart, "As we watched the picture I simulated the movement of the car by bouncing the performer and myself up and down on the plank."
It is tough, however, for actors to match later the emotional level they achieved on the set. Ron Bochar, who supervised the sound on Philadelphia, describes the powerful scene where Tom Hanks is responding to an opera recording as a case in point. Ideally the aria and the dialog would be on separately manipulable tracks so that the dialog could be kept intelligible. But Hanks wanted both the freedom to move around and the ability to hear and react to the singing of Maria Callas. As a result, both his dialog and her aria are recorded on the same track and the dialog is less than ideal. But everyone involved agreed that the live performance was preferable to looping the scene. "That's one of those things about 'mistakes' that get put in because you are forced to or they just happen," says Bochar. "They turn out to be things that you could never re-create. You'd ruin the scene by making it cleaner."
Today, one of the first jobs of dialog editors is to split spoken lines (usually from different camera--hence microphone--angles) onto separate tracks. Doing so, says Kirchberger, "makes them as independently controllable as possible, so that we can later 'massage' them in such a way that they fit together seamlessly." This is not to say that filmmakers can't do creative things with dialog.
Robert Altman, most notably, developed with rerecording mixer Richard Portman a technique for creating his unique multi-layered dialog style. During the shoot Altman, who allows a lot of improvisation, mikes each of his simultaneous speakers on separate tracks (sixteen for Pret-a-Porter). Later the rerecording mixer can raise and lower the relative volume of each track to create a weaving effect among the various actors' lines.
Dialog can also be edited to affect characterization. Suppose the director wants to make an arch-villain more domineering. A mixer could raise the volume of his voice and adjust the tonal qualities to make him sound larger than life. It's the aural equivalent of someone invading our space by standing too close to us. The picture editor could enhance the villain's sense of menace by regularly cutting to his voice before we see him. Because he seems to lurk just beyond the edges of the frame, the viewer will feel uneasy about his potential reappearance whenever he is not present.
Next month, we look at ADR.