Building a Film
Soundtrack
Sometimes reality is a poor substitute for artifice. Motion
pictures may be the defining art form of the Twentieth Century, and most people accept
that what we see up on the big screen -- or in our home theaters -- is largely fake. Live
action can be combined with animation or modified by various forms of trick photography
and exotic processing to create an environment and series of actions that never existed.
We know these are illusions, but they charm us anyway.
Except in a vague sort of way, however, most of us don't
realize that what we hear on a movie soundtrack is usually even more artificial than the
visual images. Only the tiniest bit of what comes out of the speakers has anything to do
with what was happening when the cameras were rolling. A movie soundtrack is built up
sound by sound, layer by layer, long after the pictures have been finalized. It’s all
part of a process called "post-production."
Even the simplest scenes require a lot of doctoring. It
might seem that something as straightforward as a couple of actors walking along a
sidewalk having a conversation might be easy enough to record, but that's not usually so.
For one thing, the natural sounds that accompany the action rarely record properly. They
either sound unnatural or are too loud or too soft. And there's always the risk of a jet
flying over an otherwise perfect take.
Part of the film audio engineer's art, therefore, is to
pick up as much of the live dialogue as he can, and eliminate as many of the other sounds
as possible, knowing they will sound better when added in later.
North American filmmakers like to use as much live dialogue
recording as they can. That not only makes matters simpler later, but it lends a sense of
spontaneity that other methods sometimes lack. But even if the director knows that the
sound recorded on the set will be unusable, either because of extraneous noise or the
difficulty of placing microphones properly for audio pickup without their being visible,
sound is almost always recorded when the actors are acting. If nothing else, it can act as
a guide when re-recording the dialogue later.
For decades, the standard device for recording dialogue has
been the Swiss-made Nagra open-reel tape recorder. Although hardly larger than a lunchbox,
this is one of the most highly regarded analog recorders in the world, and has
occasionally been used for mastering music recordings in preference to much larger studio
recorders.
For film use, the Nagra has one vital feature: it contains
a control track alongside the audio so the tape can be synchronized with the film during
the later editing process. Many of these machines are still in use, both because audio
technicians are fond of them and because they represent a sizable investment for the
largely freelance audio engineers who often do this sort of recording.
Nagra does have a digital model, which records four
channels of 20-bit digital audio, but these are still relatively rare because of their
cost. A third alternative is DAT, there being film-specific machines that can record a
synchronizing track along with the audio.
Dialogue is usually recorded using a microphone on a boom
held over an actor's head, just out of the picture. The microphones used are directional,
to focus on the performer's voice and to exclude as much other sound as possible. On
occasion, more than one microphone is used, sometimes to pick up a second actor, sometimes
to offer an alternative to the first unit. Radio microphones come into play for this
second purpose, although they can be plagued with interference, especially away from the
studio.
An image familiar to most of us is that of an assistant
director snapping a clapper-board shut at the beginning of a take. This is a vital step
for the later processes because it not only gives a visual identification of what's on
that piece of film (and an audio one -- the person who claps the board also reads its
information onto the tape), but also the clap itself gives both audio and visual markers
that can be lined up later in the editing suite so the sound and picture start together.
Clappers have mostly given way to devices that use lights and electronic noises, but the
principle is the same.
Once filming has been completed, the audio has to be
converted into a form that will be usable in post-production. Until very recently, that
meant dubbing it to magnetically coated 35mm film stock, using the control track to make
sure that the audio corresponded perfectly to the picture, frame by frame.
The two pieces of film could then be loaded into an editor,
lined up using the clapper information, and then run through the machine with audio and
image in perfect sync. By cutting the two in identical places, bits of different shots
could be assembled into complete scenes, then reels, maintaining synchronization
throughout.
These analog techniques do still exist, but they are
rapidly being taken over by digital technology. One popular medium uses Hi 8 camcorder
cassettes to record eight channels of 16-bit digital audio, which can be manipulated by
means of sophisticated audio editing computer programs. Multiple units can be linked so
that, ultimately, any number of tracks can be used to make a final mix
While the cassette system has become very common, it shares
with the older analog system some of the drawbacks of being tape based. That has been
addressed by a newer machine that uses removable hard drives as the recording medium,
again controlled by computer. These machines have huge storage capacity and the virtues of
random access.
In a perfect world, stringing all these bits of on-set
recording together would result in a complete dialogue track but, as noted, a lot of it is
unusable. It may have too much extraneous noise, or bad microphone placement, or a bad
reading of the line on the part of the actor. In extreme cases, it may be necessary to
replace a voice completely and use that of another actor.
The second step in the audio assembly is automated dialogue
replacement (ADR) or, as it used to be known, "looping". In analog practice, a
scene that needed dialogue repairs was cut up into short bits -- individual sentences,
often -- and the audio and video film spliced into loops that could display the image of
the actor saying the particular line over and over. As he listened to the on-set audio
through headphones, he would say the line repeatedly along with the film until an
acceptable version was recorded. The new sound would then be spliced into the dialogue
track to replace the old.
ADR studios offer a variety of microphones to match those
used in the field, and acoustics can usually be controlled to some extent to achieve a
reasonable match. Again, digital technology is now used for these functions, but the
process is similar, if somewhat more easily accomplished. The same techniques, it might be
noted, are used for dubbing movies into foreign languages; the originating film company
supplies an M&E (music and effects) version of the film, with everything except
dialogue, and the foreign studio inserts its own script.
Except for the salvageable live dialogue, almost everything
that goes into a movie soundtrack is added after the fact. The exception has to do with
musicals, where the musical numbers are recorded in advance, the actors miming along as
they are replayed on the set.
The on-set and ADR engineers take considerable pains to see
that that the dialogue track has nothing but dialogue on it, but there are sounds you
expect to hear with certain actions -- opening a door, picking up a coffee cup, walking
across a room -- and these are supplied at the next stage, which is called
"foley" after the man who first practiced it.
A foley studio is very similar to an ADR studio, but its
purpose is to add incidental sounds rather than words. Again, the picture is cut up into
short segments that are shown on a screen, and the foley artist tries to match appropriate
sounds to the action on the screen, wearing a variety of clothes (shoes, for instance) and
using a wide range of props. A given scene may use numerous different such sounds, each
carefully laid down separately. Multi-track digital recorders can be used for this, and
they are useful for minor timing adjustments (as when a click is a fraction of a second
late), but many engineers prefer to use 24-track analog tape recorders at this stage
because of the speed and quietness that the recording function can be punched in and out.
Foley artists often try to improve parts of an otherwise acceptable sound effect.
Only after all the foley sounds for a scene are assembled
is the analog tape dubbed to digital (or, where it is used, to 35mm magnetic film). As
with dialogue and ADR, the result is a series of tracks in perfect sync with the
corresponding pictures.
Over the years, an immense body of recorded sound effects
has been built up, and these are extensively used in films. In many cases, that's because
they exist and there's no real point in creating an effect if an acceptable one is already
available. Which one is acceptable is not always obvious: in several effects samplers I've
heard, some of the sounds are so similar that I could hardly tell them apart, yet
presumably a sound designer has to choose one over another.
Some sounds are used because they're conventional. A real
gun doesn't sound like much when you record it, but there is a selection of bangs that we
accept as gun sounds and they are what is used. Ditto with punches: slugging someone
doesn't make all that much noise (except, perhaps, to the sluggee), but the resounding
thwomps used in movies have the illusion of reality. And one commentator some years ago
pointed out that, if you listen carefully, if a movie has a night scene and there are
frogs audible, they will be California tree frogs no matter where the movie is set because
that's what viewers expect to hear.
Percussive sounds -- gunshots, doors slams, explosions --
used to be dubbed to 35mm magnetic film and then simply spliced into the effects track at
the appropriate frame. Now, computer software lets the engineer select the sound from a
menu and paste it into the track at the appropriate moment. These sounds may be in stereo,
or even surround sound, but a lot are mono; it's up to the mixer later on to place it in
the sound stage and add whatever acoustic environment he feels is necessary.
At this stage, continuing ambient sounds are added as well.
Things like wind, traffic, crowd noises and tree frogs are laid down, again to be
fine-tuned later in the mixing process. Often, the sound designer will customize the
effect by combining more than one, sometimes surprisingly. In one movie, the sound of a
train being coupled together stood in for a ship breaking up.
After this stage, there may be dozens of separate effects
tracks, foley and recorded, all synchronized but not combined in any other way. Some
sub-mixing may be done at this point if certain combinations are obvious, but generally,
that is left to the final mixing stage.
Virtually every film has music in it to some extent,
whether the occasional bit of incidental mood-setting or vital to the atmosphere of the
film. There are two sources of music: existing recordings and a score specially created
for the movie.
Nostalgia movies are notable for peppering their scenes
with old records, and the technology for this isn't much different than for sound effects.
Ditto "production music", which is generic music that producers can license and
use in their pictures -- generally low-budget efforts -- without having to commission or
record their own. Both types can simply be dubbed to whatever production medium is being
used, and mixed in when appropriate.
Blockbuster pix tend to have original music tracks, and
these are often recorded in a handful of special studios that allow the conductor to watch
the action on a screen as he directs the orchestra. Like dialogue, this is recorded on a
medium in sync with the appropriate picture. Other music, not so action-specific, might
well be recorded in a conventional studio.
Either way, the final group of tracks exist, and the
assembly into a single composite sound track can begin.
To this point, the efforts of the various sound people,
from on-set boom man through foley artist and sound effects specialist to music recorder,
have been putting together a set of building blocks. These are the elements of the final
track, aligned in terms of time, and recorded as well as technology allows, but still
basically raw. Blending them into a coherent whole takes special talents and some very
distinctive equipment.
First comes the mixing theater itself. Some of these are
quite modest, for television shows and smaller-scale productions, and these sometimes have
projection video monitors rather than movie projectors. For mainstream films, however, the
viewing rooms are as close as possible to full-blown theaters -- some even have seats and
can double as screening rooms.
It's important that the room in which the sound is mixed
have characteristics as close as possible to real theaters so that the mix can be made in
a real-life acoustic environment. Many are THX certified, which is only sensible as the
whole reason for THX certification in theaters is to make them as similar as possible to
the mixing rooms where the creative decisions are made.
Across the back of the room is usually an immense mixing
console, enabling the engineers to control numerous audio sources. Newer rooms have much
smaller digital "boards" that do as much, or more, but in much less space.
Traditionally, the heart of the mixing process, at least in
the hardware sense, was a wall full of "dubbers," surely some of the most
remarkable machines in all of audio.
A dubber is a large tape deck that plays the 35mm magnetic
film produced in the earlier post-production phases. In the control room adjoining the
mixing theater, a number of these machines are linked together, and to a projector, so all
of them start and stop at the same time. To watch this process is impressive, as dozens of
machines spring into action at the same instant.
Each dubber is loaded with one of the raw tracks --
dialogue, ADR, foley, music -- and they are all set at the corresponding beginning frame.
Each output is fed to an input on the mixing console in the theater, where the mixer can
control its level and various other attributes, and can direct it to one of the channels
of the surround sound final mix.
A touch of a button starts all the dubbers, the projector
and the final recorder, and traditionally mixers have tended to work in short spurts, fine
tuning the mix by trying it out, rolling the tapes back, trying again, until they're
satisfied. The process is called "rock and roll".
As with other functions in the building of a sound track,
this final mix has been increasingly digitized. The DA-88 Hi 8-based digital decks are
now, as one engineer told me, as common as dubbers after about three years of
availability, and these are now being supplanted by the MMR 8 disc-based recorder, which
offers greater flexibility and 24-bit resolution. Because there's less sense of finality
to each element of the mix, creating longer stretches at a time has become the rule, and
rock and roll is less a part of the process.
The mixing engineer has a number of visual guides to aid
him. For instance, long diagonal lines are sometimes drawn on the film itself in grease
pencil, or scratched into the emulsion, to show the beginnings and ends of fades. When the
line projected on the screen reaches the far side of the screen, the fade should be
complete. Dark scenes are sometimes temporarily lightened so that the mixer can see
details better, details that determine what the sound does but don't necessarily have to
be apparent to the audience. This echoes the use of "slash prints" -- black and
white versions of scenes -- which are sometimes used earlier in the post-production chain
because their high contrast enables engineers to see detail more clearly.
The final mix is the real creation of the sound track. If
it doesn't work at this stage, all the work that has gone before will have failed.
Ultimately, line by line, scene by scene, reel by reel, all
the audio elements are blended into a final six-channel surround mix, which has been
created in an environment that emulates that in which moviegoers will ultimately
experience the film.
...Ian G. Masters
ian@mastersonaudio.com
|
