The Conversations: Walter Murch and the Art of Editing Film

by Michael Ondaatje

O: Did you edit that?

  M: No. But I did some of the preliminary culling of the documentary material used in it. And because there was so much material, I started taking representative pictures of each shot to help organize it all. That was how it began. I didn't develop the system in its fullest sense until Unbearable Lightness in 1986.

  O: Can you explain the system?

  M: One of the first things I do when I get a script is to break it down into sequences—the closest thing to these would be chapters in a book—which are my best guesses at what an audience seeing this film would understand as dramatic scenes.

  O: Like Almásy's plane crash into the desert at the start of The English Patient.That would make up one chapter or sequence?

  M: Exactly. Then you cut away from that scene to the train, with Hana, and you're off somewhere else. At that moment you're at the end of one sequence and the beginning of another. There are on average about thirty to forty of these sequences—or chapters—within a typical screenplay. And I number each of those sequences in the order in which they appear in the screenplay. But for practical reasons, since those train scenes may be scattered throughout the story, we shoot all the material to do with the train at the same time, because trains are so expensive.

  Then, as the film is shot, I select a short series of representative stills—two to five, usually—from each camera position, called a setup, and print them, with the setup number in the lower left corner. The crew may shoot anywhere from seven to thirty setups each day. The pictures are then stuck onto a piece of stiff board about two feet wide by four feet high—each sequence having its series of boards. In the end, when the film is finished shooting, there will be thousands of these pictures—about forty to each board on perhaps a hundred boards which I keep stacked together in sequence order.

  I can then hang the appropriate boards on the wall of my room when I'm editing a sequence, and have a visual display, organized numerically, of all the material that was shot for that sequence.

  O: And the boards themselves are organized by script sequence?

  M: Yes, but the photos on the boards are there in the order in which they were shot. So there is order and chaos mixed together.

  A photo board for Apocalypse Now. On every film since Unbearable Lightness, Murch takes a series of representative stills—two to five—from each camera position (setup), prints them with the setup number in the lower left corner, then sticks them on a foamcore board. Each sequence has its own series of boards.

  This means that the boards give me visual juxtapositions, vertically and horizontally—which is what I find so provocative. I see this image at the right edge of the first board, next to that image on the left edge of the next board. That juxtaposition may never have been intended by the script, but because there they are, next to each other, it tells me: This is possible. There's a little ticking bomb there that says: Photograph X could be, in the film, next to photograph Y—you never thought of that, but it's possible.

  There may be several “iconic” frames within each shot. Essentially I'm trying to answer the question: Why did the director shoot this shot? In still photography it is what Cartier-Bresson calls “the decisive moment.” One frame in particular, from the several thousand exposed for each shot, will catch my eye as being more representative of what the director was looking for. A glint in the actor's eye, a twist to the mouth, the way the camera was positioned with the light.

  As I'm assembling the film, I'll be trying to find the exact moment each shot reaches its optimal maturity: I want to hold every shot on screen long enough for it to deliver the goods, but cut it off at a moment when it also has the potential to lead to something else.

  Part of a photo board for The English Patient.

  Let's say that I've provisionally chosen a particular point to end the shot. I'll then turn around and look at that whole wall of photographs and—without my being consciously aware of it—my eye will leap to an image that answers the question I was asking at the end of the previous shot: Oh! It would be good to see that image next!

  O: So it's not a logical or rational jump.

  M: Well, there is a certain logic to it, ultimately, but at the moment of making the choice it is more spontaneous, like performing a musical improvisation.

  O: You're surprising yourself.

  M: Exactly. There's just the right amount of turbulence in this system. There is a pattern, and there will be sections where the photographs are all about the same thing, but sometimes they jump around. I like that. I want to introduce— within the relative rigidity of all my databases and notes and structures and numbers and everything—I want to include this element of randomness.

  O: This is William Burroughs as film editor.

  M: Right! I'm sure, if you wrote down, under each of the photographs on those boards, the dialogue that was being spoken, and then put it in that order, it would read like some of William Burroughs's books!

  A PEBBLE, A CRICKET, A WRENCH

  O: Like Fred Zinnemann, you've worked on documentaries as well as features. I was recently involved in responding to a friend's long documentary and I witnessed that traumatic and difficult stage where you need to radically restructure it, cut it down, reshape it, without giving it a false sense of speed or plot.

  M: You face that problem even with theatrical films. But the problems can be acute in documentary films—I'm thinking right now of the documentary section in The Unbearable Lightness of Being, where we had forty hours of documentary footage and had to fuse a tiny fraction of it with images of the film. How do you reduce the key moment in a nation's history, for which you have so many hours of material, into fifteen minutes? It was a question of time, simply spending time with the material and selecting striking images. Not just visually striking, but striking in all senses. Then finding ways to put those images together so they enhanced one another, both by resonance and by contradiction. If you find something very white, look for something very black, and put the two next to each other.

  That's where you start. What you're constantly trying to find, as you distill it, is ways that these images can go together at deeper and deeper levels. What will immediately attract your attention—unless you're very lucky or very wise—is something on the surface. What makes you pick up a particular pebble? You're walking on the beach and thinking about the political situation in the Congo and you pick up a pebble. Why did you pick up that pebble? Something made you do that.

  It's similar…. Why do I select that shot? Something about it made me select it. Once having selected it, a process of organic crystallization begins. I can think of no higher tribute to a film than that—that you sense simultaneously that it's crystalline and organic at the same time. Too crystalline and it's lifeless, too organic and it's spineless. The human body is made of amorphic crystals—our DNA is an amorphic crystal which provides just enough structure to make it persist in a world that is trying to undo it. Yet it's random enough to be adaptable.

  Sound cue sheets for the mix of Apocalypse Now, drawn by Murch.

  One of the reasons I lobby for the increased collaboration of everyone who can have a voice on a film is that through collaboration you add facets to the work. The work is going to be seen by millions of people, over many decades and under very many different circumstances, and even though the film is a fixed thing, you want it to be multifaceted so that different people will see different things in it, and come away rewarded.

  The best, easiest way to get that multifacetedness is to allow the collaboration of lots of people, as well as Chance, which is sometimes Fate in costume. Each of those moments of collaboration, each contribution by someone other than the director, adds a slightly different perspective to the work, some chisel mark slightly at an angle to the central vision. And each of these moments, these facets, has the potential to make the work “sparkle” in a creative sense, and make it more accessible to a greater variety of people over a longer period of time.

  If, instead, the film i
s a single, monolithic vision, the viewer has no option other than to submit to it on its own terms, bow down to it, or turn away from it.

  O: The theatre director Paul Thompson, who works on and directs collaborative plays in theatre, speaks of the importance of that collaborative balance as essential in adding to the “thickness” of a scene.

  M: Yes. Where he says thickness, I would say density, but it's the same idea. You trytoget things to where there's substance to them. When you hit them, they hit back. At the same time, there must be clarity. I am always striving for a clear density. If the scene is clear but insubstantial, I think about what I can do to, in your friend's words, thicken it up. If it has density but opacity, I try to add clarity.

  O: You've written about needing cricket sounds for Apocalypse Now, and how you built the overall sound up from one singular cricket, reproducing it so that the sense of crowdedness was there, but keeping the sounds all in the same key….

  M: We wanted a hallucinatory clarity, which you don't get when you go out in the field and record a thousand crickets. You get something, of course, with that kind of recording, and it's something that I use all the time in films: a shimmering curtain of sound. But at this particular point in Apocalypse we didn't want that. That was too real, too ordinarily real. We wanted something that was hyperreal. We got it by recording individual crickets very close, then electronically multiplying them till we had a thousand crickets. It's as if they each had their own little radio mike on. Then we had a thousand tracks of crickets.

  Murch with Babe, who is wearing a microphone designed by Alan Splet to capture snorting sounds for The Black Stallion.

  O: And their pitch was the same.

  M: Yes. One cricket is very much like every other cricket. That allows you to superimpose them all on top of one another and not get chaos. If you had thirty different kinds of insects, recorded them all separately, and then played them all together, multiplying that by a thousand, it would be something else. It would not have harmonic unity to it.

  O: We're getting back to that issue of a “point of view”—even in sound. If I'm in the tropics and I hear a certain bird shrieking, my ear will remove all the other, minor noises, I will be able to focus on, to pick up that bird shrieking. You, as a sound editor, are making that selection for the viewer.

  M: Yes. That's an ideal example! On the first feature film that I mixed—Francis Coppola's The Rain People—there was a scene with a woman—played by Shirley Knight—in a telephone booth alongside the New Jersey Turnpike. In advance, I thought, I'll record lots of traffic and it'll give the sense that you're beside the freeway. I did this, but very soon realized it was counterproductive. To make its point the recording had to be played at such a level that it ruined your ability to concentrate on the dialogue in the telephone booth. I didn't want to give up the idea of locating the woman in a place—beside the high-way—but I wanted to hear what she said, because that's what the scene is about: she's talking to her husband and telling him the reasons she's run away from home.

  I discovered that if I used what you might call a precipitant sound, something we associate with a specific environment but that is itself distinct, then the other sounds come along automatically. What I did was record somebody dropping a wrench fifty feet away—as if fifty feet away from the booth, in the garage of the service station where the woman in the movie is stopped. It was important that it was far away, and that it was a certain kind of wrench dropping on a certain kind of polished concrete. If you've ever been in such an environment, you know what that sounds like and you know that such sounds are commonplace in service stops near big highways.

  That little sound was able to bring along with it, imaginatively, all the traffic. But the traffic sound exists in your mind. I spent a lot of time trying to discover those key sounds that bring universes along with them. I tend not to visualize but auralize, to think about sound in terms of space. Rather than listen to the sound itself, I listen to the space in which the sound is contained….

  O: As in the way we hear that bell in the distance, subliminally, while the English Patient eats a plum—so we become conscious of the landscape between his bedroom and the bell, which seems half a mile away.

  M: And by implication, yes, all the birds and the insects that live in that world. And, by contrast, it is very different from the racket you've heard in the film up to that point, of the convoys going through the mountains and planes being shot down out of the sky and trains jiggling back and forth. You're now in an environment quiet enough to allow you to hear a distant bell. The bell brings a whole raft of associations along with it. There are religious connotations and geographical and cultural connotations, in addition to the purely spatial. The crickets we were talking about earlier: Where are they? They're nowhere. They're in Willard's head. They are spatial, but it's a mental space.

  “THE BLUE LOOKED DEAD”

  M: The chemistry of soundscapes is mysterious and not easy to predict in advance. You go into the final mix of a film knowing that certain things are possible, but not knowing exactly how they're going to work themselves out. I remember in THX 1138 we were trying to achieve a great contrast between two environments: the silence of the white limbo prison and the chaos beyond it. In the silence of the prison you hear only distant cooing sounds—which I recorded at the Exploratorium, here in San Francisco—ambiguous machines very far off and so remote that you can't really tell what they are. And occasionally thunder, on a whim, just because the space is so big.

  After many hours of walking, the characters THX and SRT—Robert Duvall and Don Pedro Colley—get to the edge of the prison, this white space. The light falls off, and magically there's a door…. They open it out of curiosity and are sucked into a corridor of people rushing like cascading water, hundreds of people, just rushing.

  To create that sound I went out and recorded similar environments: people coming out of football matches, roller derbies, marathons, lots of people moving very fast. I recorded waterfalls, sewage pipes, rushing air—all those kinds of sounds. As I'd done in The Rain People, I piled them all on top of one another, thinking more and more and more equals more. I was going to conquer this once and for all! When we played everything together, however, a funny perceptual thing happened. The needle of the recorder went way off the meter. We couldn't see the needle anymore, the sounds were so powerful. But our impression was that it wasn't very loud. It was frustrating after all that work. How does 1 + 1 + 1 + 1 + 1 = 2?

  There was something about the chemistry of those sounds: they were all rushing sounds but none of them had edges, so that the ear couldn't seize them. Objectively measured, there was lots of energy, but subjectively, because there were so few edges to the sound, it wasn't particularly loud.

  Then I remembered—these are the mysterious things: why did I think of this?—that a few months earlier I'd been at the Academy of Sciences at two in the morning, recording footsteps. For some reason I'd put the recorder at one end of the African Hall, and stood at the other end and just shouted incomprehensible, guttural speech. It echoed in a beautiful way: the recorder was 150 feet away, the sound bounced off the marble and glass surfaces, in the dark, at two in the morning.

  I remembered that track and thought, I'll add that one to the mix. The effect was instant, overwhelming loudness. I had to pull the faders almost all the way down, and it was still too loud. Even though the meter was just sitting there quivering slightly, in the middle.

  Again a paradox. We were experiencing great loudness, but electrically there was not very much energy. In retrospect I realized that what I'd done was provide one of those precipitant elements, like the wrench dropping, like the bell in the distance. This voice provided edges because—ack ack ook ook ark—it had starts and stops to it. Those edges collected all the energy of the sounds around them and delivered them in a startling way. So I was able to play the scene—electrically speaking—quite softly. But your impression is that it's very loud, because of the presence of
this sound that precipitates the energy of the sounds around it.

  The lives of people who orchestrate symphonic music must be full of such discoveries, handed down as observations … such as, Oh, if you combine the oboe with the violin you'll get this effect that far outstrips the amount of energy you're putting into it, because of the synergy between these two unlikely elements. Cooking is full of these discoveries too. Cooks are always trying to find unusual substances that, if you put them together, seem to excite the taste buds in new ways, because of the talent of that one contradictory taste to precipitate the elements of another: He added crushed olives to the cake icing! If you remove the precipitant, you're removing a very small element, but it makes what's left kind of bland.

  It happens in the chemistry between sound and picture as well. A certain sound colour will make you see colours in the picture in much more vibrant ways.

  There was a crisis with Apocalypse Now, back in '78, when it turned out that we didn't have the rights to Georg Solti's recording of “The Ride of the Valkyries,” which is what we'd been using. Decca Europe refused to give it to us. We were very close to finishing the film and the fear was that we wouldn't get the rights, and then what would we do? We triaged the situation: continuing to petition Decca for permission, making arrangements with the San Francisco Symphony to record the music again, and trying to find an old recording that was close enough to Solti's that maybe we could get the rights to. That last task fell to me.

  I went to Tower Records and bought all nineteen recordings of “The Ride of the Valkyries.” I sat with a stopwatch and a metronome, figuring out which recordings followed roughly what Solti had decided to do metronomically with the music. In the end, I eliminated all but one—Erich Leinsdorf conducting the L.A. Philharmonic. I thought, It's not quite the same, but it's close enough and maybe, if we're lucky, I can make some adjustments to the picture….