Walter Murch, ACE, has edited for more than six decades on more than 60 films and more Oscars, Emmys, Eddies and BAFTA awards than I can mention. His imdb page lists 66 films over more than 6 decades.

I sat in a screening room in San Francisco with Walter before watching the documentary he edited, Coup 53. In two previously published interviews, we discussed that film and follow-up questions about his book, “In The Blink of an Eye” and his book with Michal Ondaatje, “The Conversations. In the upcoming weeks, I’ll post an interview in which the questions will be asked by the members of the Blue Collar Post Collective.

This interview is about 7 of his films that we discussed after he gave me “homework” to watch five specific films. The films we discuss here include The Conversation, Julia, Particle Fever, Romeo is Bleeding, The Talented Mr. Ripley, Tomorrowland, and The Godfather.   The podcast of this interview was released in mid-February. This interview has some minor edits for clarification that make it slightly different than that podcast.

(This interview was transcribed with SpeedScriber. Thanks to Martin Baker at Digital Heaven)

This interview is available as a podcast.

HULLFISH: Apocalypse Now had a LOT of footage, but the comparison of most film-acquired projects to digitally-acquired project has meant a LOT more footage, generally. Does all that footage make it harder or easier for the editor?

MURCH: It’s easier for the director, largely. And it’s somewhat easier for the script supervisor, but it’s harder for the editor. I should just qualify by saying that the way Francis (Ford Coppola) shoots certain scenes — and The Conversation is one of them — is, he makes it as if it’s a documentary. So people have identities, they have dialogue. Sometimes they’re allowed to improvise. The scene is set and he just launches the actors into it and covers it with multiple cameras. And at the end of a take, he decides — this is all before videotap (in other words, no monitor) — so he has to kind of intuit what’s going on. He decides, “OK, I’ll do another take” or “I’ll move the cameras into a secondary position and ask them to do it again.” And this generates, as you can imagine, a huge amount of material.

So when I received the dailies on The Conversation from “the conversation” — which is to say the people walking around the beginning of the square — it was a little overwhelming for me. This is the first feature film that I had picture-edited. I don’t know what I had expected, but I wasn’t expecting a fusion of documentary and fiction.

So I had to do a grid map of all of the dialog and then all of the different cameras on X-Y coordinates and then — say this line of dialog is covered on this camera and this line of dialog isn’t and then I had a grading system of: “its covered and it’s very good” or “it’s covered and it’s a little sketchy.” So it was kind of like a Sudoku game.

But it was very valuable because the only films I had edited before that were documentaries, so I kind of flipped my mind into that mode. Francis did the same thing earlier on the wedding scene of The Godfather, which was simply a wedding, and everyone had their identities and multiple cameras rolled. And the same thing on Apocalypse Now with the attack on the Valkyries, which was much, much more daunting physically because the cameras were mounted in helicopters and you had much less control, so you would shoot ten minutes — which is all the camera could hold at that time — and then go back to base camp, and decide where and whether to move the cameras. So early on I got inoculated to a lot of footage for a scene and that kind of kept me in good shape editorially. So I was challenged early and I adapted my techniques to deal with that.

Because of the cost of film in those days, you would have two streams: there would be “print” and what would be called “b-negative,” meaning, “Don’t print this take, but hold onto it and maybe we’ll use it, but only if we get into trouble putting the scene together with the printed takes.” The ratio would probably be: out of seven takes, maybe three would be printed.

There’s a rule of thumb regarding that, which is that the next-to-last take is usually the best take, even though the director prints the last take. That’s due to the psychological interaction of actors and directors where the director is fishing for something and he sees the progress of the actors. If it’s a scene with multiple actors, he sees that it’s getting good; no, no, the fourth take wasn’t so good; now the fifth take is good. Sixth take is even better. Let’s do another take. And if the seventh take is not so good, what the director finds impossible to say is, “Okay. Print take six, but don’t print take seven.” Because the actors will say, “What was the matter with the last one?” So, you naturally print take seven and you say, “Great, let’s move on! Next setup.” And either the director will imply or leave it unstated that the sixth take — which is where they peaked — is the best one.

I told this story to (director) Fred Zinnemann when I was working on Julia and he was 70-years-old at the time and was very traditional. He shot a low amount of footage. And Francis, in The Conversation and other films, would — if the actors went off, he would keep going and maybe sometimes print or shoot 15, 20 takes. And his justification of that was that: first, the actors are good, then they got confused, and we just keep going and out of boredom, then finally they come up with something unique, and I want to get that unique thing.

So with digital, none of that really applies. There are all kinds of different methodologies that different directors use. But in my experience, almost everything gets printed. You have to cope with everything, but they indicate that there are different streams — that was the good take and this is the not so good take, but they get “printed” and the editor has to deal with them, whereas in film those takes would just not get printed, and so you would have a clean deck, as far as that goes.

But the unique thing about digital is the idea of resets within a take. So the director calls “action,” you’re going along and then at a certain point the director will say, “Okay, keep rolling but go back two lines and start again from there and make it a little angrier.” Then you go forward to line 10 and then he’d do the same thing. You go back to line eight and then move forward. So it’s kind of like cross-country skiing across the scene.

HULLFISH: Great analogy.

MURCH: The most number of resets that I’ve coped with was around mid-thirties, 32 different resets. And that’s difficult to keep track of. I think there’s technology now that copes with that, but this was five or six years ago.

HULLFISH: The Conversation had a very conservative use of close-ups, I thought. Is that something that you feel is true — just in your style editing?

MURCH: Yeah, I think so. It went through different kinds of geological epochs in the shooting. The conversation itself (the conversation that starts the film) is shot documentary-style. Nobody really knew what the cameras were going get. What Francis was after were serendipitous moments between the actors and the actors relative to the background people. The people were not actors they were just ordinary people. They weren’t even extras. The actors we were following were in the world of reality which added this wildcard aspect to it all.

Then the section where Gene Hackman is analyzing the tapes is very two-dimensional. It’s just a medium close-up of him. And then looking at a kind of flat image of the tape recorder and the VU meters and the knobs.

And then there is the three-dimensional space of the party where Gene Hackman invites everyone back to his laboratory and they have a little party and that’s staged in a deep space.

It’s not abnormally free of close-ups, but it certainly isn’t over close.

HULLFISH: With some inexperienced editors I’ve seen, they tend to go to a close-up right away or they use them a lot. There’s an overuse of them and then that removes their power.

MURCH: Exactly.

HULLFISH: So I thought in your case, you were very careful about where you chose to put those close-ups.

MURCH: Right. When I was editing Julia, my assistant editor had worked with Fred Zinnemann before, but it was the first time I had worked with him and my assistant made a point — whispering to me — “Don’t go to close-ups too soon, because Mr. Z — which is what everyone called him — hates that,” for exactly the reason you pointed out — that the close up is kind of the nuclear button on the scene. When you go to a close-up, you want it to mean something.

And of course, that makes it difficult. If you have matching problems in a wider shot then where do you go? I remember in the scene in Julia, where they’re at Alberts, which is this cafe, so it’s a scene between Jane Fonda and Vanessa Redgrave. This was a real cafe in Strasbourg. After dailies, the script supervisor said, “It’s a nightmare of mismatching. You’re gonna have your hands full with this one.” But I found a way to cope with it. But, their hands were not always in the same position when they said a line and I wanted to hold off using the big close-ups until the apex moments of the scene. So you have to choose your moments carefully.

Films are full of mismatches. It’s just like any magic trick: it’s how visible those moments are or are NOT for the audience.

HULLFISH: You mentioned Fred’s desire not to go to close-up too early. In the book, “The Conversations,” you talked about your preference to cut before action begins, not mid-action. Do you find that certain directors tell you, “I love cutting mid-action” and then you have to change? Or do you have to have that discussion that you don’t like to do it and then debate why?

MURCH: The only time that’s happened was with Brad Bird on Tomorrowland. Otherwise, nobody’s ever mentioned it to me. With Brad, I think it was because he’s an animator and animators are — by nature — they think in shots and how do you join one shot to the other?

Almost naturally, they think of ending the shot with an action with which they will join up to the next shot. So the fact that I didn’t do that got his attention and he wanted always to have matching action. I tried to explain sometimes what I was after, but that didn’t really interest him.

HULLFISH: And we’re in the job of servicing our director, right?

MURCH: My rule of thumb — especially if it’s an idea that you really believe is important for the film — is to say your idea. What you’re being paid for is to contribute ideas, not just sit there and “tell me what to do.”

If the director doesn’t like it and you still really think it’s important, wait a while and then find a good opportunity to once again say, “You know, that thing we were talking about last week? Now that we’ve…” And if the director says, “no,” you say, “OK.”

If it’s REALLY important to you, try it a third time. If the director still says, “I don’t want to do that,” then shut up, because you don’t want to become a pest. On the other hand, if something is really important, find a diplomatic way to get the idea across. And everyone has different methodologies for that.

HULLFISHLet’s talk about Particle Fever. I loved this documentary. I loved the cut as Nima finishes erasing a chalkboard, then there’s a big slash drawn on the board and that motivates the cut to the next shot.

I thought of your theory from “In the Blink of an Eye” that the shot has a life and after a certain moment, that life is over-ripe and it’s time to move on.

https://www.youtube.com/watch?v=BbbGETO8G9g

MURCH: The closest analogy is what happens in music. That gesture was kind of like the blare of a trombone or something. At the moment, what you generally want to do is cut just before it overstays its welcome. If somebody is moving towards a door and about to go through the door, unless you’re making the point of the door itself, you don’t let them go through and then cut. But you don’t want to cut too soon either, so you have to feel the moment where it is inevitable that the person will go through the door and not do a Colombo — which is to stop and turn and say something else.

It’s kind of like the moment when a plane is taking off. The moment where the wheels are up. You’ve run out of runway and you have to take off. And that’s the moment to cut.

HULLFISH: To go back to The Conversation: I don’t know whether this is something you remember specifically, but I thought that there were numerous times that there were empty frames either at the beginning or the end of the shot — not that that didn’t mean the shot was over — but that they were absent of people.

MURCH: That was part of the aesthetic of the film in that Francis wanted this unsettling aspect — sort of the DNA of video surveillance — to infect the visual style of the film as if there wasn’t somebody behind the camera and that there was a motion detector attached to a servo motor and as long as the actor is moving in the shot, you hold. If the actor leaves, you wait, wait… It’s been programmed and if he doesn’t come back in the shot, then you go looking for him. He exited left, so I will pan left and do it very robotically. Those pans don’t have a human behind them. They are, but the operator was told to make it look artificial.

You see that throughout the film, particularly at the beginning when Harry’s in his apartment.

HULLFISH: Back to Particle Fever again. There are lots of pre-laps. What does a pre-lap get you? (To start audio from the next scene while you’re still on the previous scene or vise-versa.)

MURCH: It’s a way of blowing a little smoke across the moment of the cut in that the moment that the new sound enters in, you’re asking the audience, “Can you make sense of this?” And so you alert the audience to a new development.

Again, you have to judge the right moment. Then at the right moment, you cut to the reality which is producing that sound, whatever it is. It’s the opposite of what in screenplays is called a smash cut. With a smash cut you just want to hit (he smacks his fist into his palm) with full force at the moment of the cut. Whereas a pre-lap is a way of bleeding some of the DNA of the incoming shot into the outgoing shot. And so you’re making sometimes a poetic simile between: “What is this person thinking?” You end on a close up of a man, and you hear a line of dialog and you’re looking at the face and a woman says, “What did he mean?” Cut. And now it’s the girl. And she says, “I told you to go to the store BEFORE dinner” or something. But by implication, what did he mean? Is something that the character in the outgoing shot is thinking or imaginatively hearing.

HULLFISH: There’s a series of jump cuts in the documentary of guys watching reloading Mac browsers to watch the collisions. Do you remember building that sequence? It’s all these people all over the world. They’re waiting for the Large Hadron Collider to run for the first time. And you did this great little montage of jump cuts between people.

MURCH: This was an unscripted documentary. It was shot over a six or seven-year period as the Large Hadron Collider was being built. And we had access not only to the footage that Mark Levinson directed but all of the archival footage that CERN had collected in all of the years of their existence since the late 1940s. So there were about five hundred hours of material that we had access to.

At a certain point in the evolution of a documentary like Particle Fever or like Coup 53, the decisions that you make editorially with the director are, I would say, almost identical with what you were dealing with in fiction film.

The film is perhaps still too long, but you’re trying to compress it and to tell the story in the clearest and emotionally engaging way that you can do it in the shortest amount of time and you will make discoveries on how to do that.

Getting to that point, though, is very different because with a fiction film that is shot with multiple takes, you obviously have a script and editorially you have an abundance of interpretation and a paucity of events. Meaning the only events are what’s in the screenplay.

The screenplay may have — let’s say — 125 scenes. You don’t have another 125 scenes that you can go to. What you have is what’s in the screenplay, but you have many different interpretations. Some of them with small differences, others with large differences in performance. And for a scene, the editor might have 50 different line readings of a line of dialog from seven or eight different camera positions. And how do you thread that needle? That’s the challenge?

If that line of dialog is said with THAT attitude, what’s the best interpretation of the NEXT line of dialog? And should we be on the person who is speaking or should we be on somebody listening? All of those kinds of questions.

Whereas the opposite is happening in a documentary, you have a paucity of interpretation. Things generally only happen once. But you have a multiplicity of events because you could have five or six hundred scenes potentially for a documentary that you squeeze out of five hundred hours of material.

And that’s the question. What are the building blocks of the film? So you are writing the story, basically deciding those questions. And then because you only have one line reading of that line of dialog, you have to find a way to make that particular line reading — in the context of the larger film — work at the optimum value that it can achieve. And you try to minimize what seem like mistakes and maximize the potential of everything else.

HULLFISH: Another scene that I loved in Particle Fever was APPARENTLY the opposite of your lesson that you only take a scene so far and then once it’s expired its use then you move on.

There’s a great scene that ends with a guy missing his exit. Do you remember? And obviously it’s a very funny moment. It’s a great place to end. But you’d think, “Oh, we don’t need to have the guy miss his exit,” it’s a mistake, but it does reveal something.

MURCH: Yeah. He misses his exit while he’s talking and we stay on him because it’s a character moment. It’s the classic absent-minded professor. Here’s somebody who’s so excited about finally getting to see the Higgs-Boson announcement that he misses his exit.

It was real. It wasn’t concocted. The camera person on that shot was David’s wife. She was just there and it really happened.

HULLFISH: That’s very funny.

MURCH: So those are the nice, spontaneous moments that you want to make the most of. To get the best character revelation. It’s a light moment at a kind of heavy period of the film where it’s finally coming to fruition.

HULLFISH: That’s an interesting thing to talk a little bit about: trying to regulate the tone of a documentary so that you aren’t at 11 or at zero or sad or happy too long. Talk to me about your sense of that as you’re a little further into cutting a documentary film — feeling that.

MURCH: These are sort of what you might call the peristaltic motions of digestion of the audience. You are, in a sense, almost force-feeding the audience because they can’t control the speed of the film. And if you feed people too fast, they feel like the stuff is being rammed down their throat and they’re going to choke.

And it’s at that point that they either check out of the film or actually get up and leave. So you need to have these moments where the digestive tract kind of relaxes for a moment and you can have a moment to digest what you’ve been fed. And humor, such as it is, is a very good digestive agent.

It indicates to the audience that the filmmakers are kind of on their side without being overt about it. It’s saying, “We know you need a break right now. And so we’re gonna give this to you.”

You’ll see moments of that in Coup 53 — which is also a heavy film because of the subject matter. But there are light moments within it. And we knew that from the beginning. When we did the card structure for the film, we had a purple card that just had the moment “grace” on it, like a grace note in music, and we would just almost arbitrarily put these things here in speculative moments.

Not all of them wound up at exactly that place, but it was a reminder to us — Taghi and me — that this was important. Otherwise, it gets really too dense.

HULLFISH: And what ended up being used for those filler card moments of grace?

MURCH: Moments where there’s just no dialog and thoughtfulness. Something very heavy has happened, and we’re just looking at somebody’s close up. We just stay there. Things like that.

HULLFISH: I’ve talked to other editors about the use of shoe leather — which, of course, is a derogatory term for us editors — but sometimes it’s that shoe leather that’s that moment of grace.

MURCH: Right. Exactly. If you look at sea anemones, it opens up and then it’s waiting for something to float by. So a little fish comes by and it grabs the fish and then it closes and it’s now digesting the fish. Then when it’s digested, it opens up again. And the trick with a film is that the logic component of a film and the emotional component of the film are sometimes on different tracks and they have different moments of digestion.

Sometimes you can digest an emotion while you’re receiving information. The danger is trying to load up both of them at the same time. So you have to kind of alternate how you feed the information.

HULLFISH: Let’s talk about Romeo is Bleeding. You asked me to watch that. Out of your considerable filmography, why that one?

MURCH: It’s a good example of when an editor comes in to help a film in trouble in post-production. There was a traffic jam of conflict between the director, the producer and the author of that film, and as usually happens, the editor got caught in the middle. So the studio shut down the film, looking for a new editor to come in and calm things down. I was at loose ends at the time and intrigued by the script by Hillary Henkin, who wrote Wag the Dog. Gary Oldman – a great actor – played the lead, Jack Grimaldi. And it was beautifully shot by Dariusz Wolski (it was his first credit as DP – he was later to do Pirates of the Caribbean). The music was composed by Mark Isham.

https://www.youtube.com/watch?v=OHft_cw-RBI

Also, I really liked the films of the director, Peter Medak (The Ruling Class, Let Him Have It, The Krays) though I hadn’t worked with him before. So when I took over the editing of Romeo, I had to identify what the trouble was and then find a solution for it and calm everything down.

The problem seemed to mainly concern a disagreement about the character of the female lead, Mona, played by Swedish actress Lena Olin. Lena, who I had worked with on Unbearable Lightness a few years earlier, was cast as a last-minute replacement for Ellen Barkin as Mona deMarco, a Mafia assassin. The part required an Italian-American accent, which Lena attempted but just could not master. My suggestion, which was accepted, was to make Mona into a Russian mobster – think Little Odessa – and as a result, she became Mona Demarková. Her accent – a crazy blend of Swedish trying to be Italian-Amerian – turned out to fit this new character perfectly.

Unfortunately, Romeo was not a success at the time, but it has its adherents now, more than 25 years later. Lena, in particular, has been singled out for her powerful and amazingly athletic, menacing performance.

HULLFISH: In The Talented Mr. Ripley talk to me about the dynamics of the climactic murder scene on the boat.

MURCH: It was a difficult scene to shoot: it took place in a rowboat out in the ocean and there were weather problems, the sun on one day, no sun on the other.

And the fatal hit itself is a complicated thing to pull off: Dickie’s head is bashed in with an oar. Makeup and VFX and everything. We tried to make everything as intense as possible.

https://www.youtube.com/watch?v=h4e-Si4oGEw

This is where Ripley (Matt Damon) reveals to Dickie (Jude Law) that he wants to live together with him. This is crazy from Dickie’s point of view. The murder itself comes out of the rage and anger and disgust that Dickie feels, and Ripley reacting violently to this. So you just try to make it as intense as possible, physically and psychologically.

But then you want to — this is one of those grace moments at the end — you want to find a way to relax after the murder and find some sort of grisly poetry in the aftermath of the violence. As horrible as that is, it can be horrible and transcendent at the same time.

HULLFISH: So the anemone releases?

MURCH: Right.

HULLFISH: Structurally that scene happens almost perfectly in the middle of the movie. Do you remember trying to have it land like that or was the structure already set.

MURCH: That was how it was in the screenplay. It varied maybe five percent one way or the other, depending on the overall structure.

Anthony’s first assemblies were always long (Director Anthony Minghella). The first assembly of The Talented Mr. Ripley was four and a half hours. So we had to find a way to cut at least two hours out of the film to get it down to a releasable length.

https://www.youtube.com/watch?v=Axur_6pjtH8

The fact that it is right in the middle and the fact that Jude is so charismatic is also a problem for the film because: “Wait a minute! I loved looking at that guy and now he’s dead? And we’re left with Ripley?” Matt’s a wonderful actor, but Ripley is a kind of creepy guy. So the whole film — like The Conversation — is told strictly from Ripley’s point of view. It’s a single point of view film, which is a technique that both Francis (Ford Coppola) and Anthony used to compel the audience to identify with a character that they would normally tend to resist.

If you’re making a film about somebody who is not the normal kind of hero of a film, then making the film have a single point of view is a good strategy. In a sense, you’re forcing the audience into a Stockholm syndrome. The audience have no alternative. There’s no moment where two of the other actors go off and discuss Ripley or discuss Harry Caul. “What is it with that guy?” That just doesn’t happen. There’s no relief from the single point of view. Everything you’re looking at is either Ripley or something that Ripley is looking at.

HULLFISH: As if you’re stuck on a desert island with one other person…

MURCH: Right.

So after Jude dies, not only does the film still stick with the single point of view, but now Dickie (charismatic Jude Law) is gone. The film recovers, but it takes a while. It is a shifting of gears where the audience has to deal with the sudden loss of Dickie.

It’s what Hitchcock did in Psycho where Janet Leigh is killed 40 minutes into the film. Her death is not exactly at the halfway point, but almost. Psycho is about 100 hundred minutes long, so it’s a similar kind of thing.

https://www.youtube.com/watch?v=zwqi3vqvDUg

HULLFISH: Talk about the cut — if you can remember — from when Peter Smith Kingsley was describing Tom Ripley’s characteristics to Tom. Then cut to him walking back in the stateroom as the audio of Peter continues.

MURCH: The scene — as shot — went all the way through killing Peter. And after Peter is strangled, the shot tilted up to the porthole window, then it went out the window And for various reasons, that did not work. I mean, it was shot perfectly well, but emotionally, it was hard for people to deal with.

So using that flash forward to Tom entering his cabin, while keeping the dialog in the past was something that I came up with one night to try to help that moment so that we see the two guys together, Ripley dealing with some very complex emotions. He’s distressed, but he’s trying with his voice to be happy. And we’re looking at his face. And at the same time, he’s tightening a cord around his hand, which is ominous.

And then you cut to a doorway, Ripley comes into his next-door cabin, and we’re still hearing the dialog with Peter. And then the audience realizes what has happened. So it’s a post-lap. And it’s devastating, but not brutally so because you don’t actually see the murder itself. It takes the audience a moment to catch up with events: “What the hell just happened?” And that slow realization both softens and in some strange way intensifies the moment because now — on the other side of that cut — the audience realizes that Ripley just killed the one person in his life who seemed to understand and love him for who he was. All because of this fluke of meeting Meredith (Cate Blanchett), he had to kill Peter.

So it was a very complex thing.

HULLFISH: There’s so much effort and so much talent in writing a script as you know — so much development process — and yet all the answers don’t happen with that script.

MURCH: And that’s one of the great mysteries of writing a good screenplay: it has to command the attention and emotions of the reader and be logical within a certain framework, but you can’t quite predict what’s going to happen, moment by moment, but when it happens that seems to be the right thing. Deja-vu, in a sense.

And yet this same screenplay has to somehow — in its DNA — acknowledge the fact that it doesn’t have all the answers to all of the problems that might happen during shooting, but it has to provide the raw material for answers to those unknown problems, at a whole different level of creativity – not with words on paper, but with images and sound in time.

A really great screenplay has to be almost indifferent to what happens to it. It’s like those men who make animals out of twisting balloons. They can make a giraffe and then suddenly something twisty happens and the giraffe becomes an elephant. But it’s still the same balloon.

And in a sense the screenplay is the “balloon” that is able to manifest itself as a giraffe or an elephant, depending on necessity, and not care. The danger — in a paradoxical way — is a screenplay that is too perfect, which HAS to stay exactly the way it is to be perfect. And the cinematic world just isn’t that way, most of the time. It frequently requires this ability to transform itself. Which is another kind of danger!

Every film is a mixture of different strengths and weaknesses. Nobody expected Jude Law to be as charismatic as he was. We knew he was a good actor, but we just didn’t know what was really going to happen. It turned out very different from the first Ripley film (Purple Noon, 1960) with Alain Delon.

In that film, Dickie is not especially attractive. It’s Alain Delon, Ripley, who is the beautiful one. And in a sense, what Anthony did in casting, is he flipped the relationship there. And so Talented Mr. Ripley has to cope with that twist.

HULLFISH: Most shots are relatively short, but occasionally, an editor uses a much longer shot — 60 seconds or 90 seconds. Does the potential of that edit point keep growing at every moment until you make that edit?

MURCH: Yes. Absolutely.

HULLFISH: Is it more dangerous to have to walk that tight wire?

MURCH: I like to say that the editor is making 24 decisions a second, whether to cut or not: “no, no, no, no, no, no, no, no, no, no, no, no, no. YES.”

If the shot is going to be long it has to be well-written and well-acted and have a complex evolution within the shot to allow it to sustain itself for a minute or more as a single shot. It will have its own internal dynamics built into it.

If the director is going to do that, he or she must put a lot of horsepower into that decision. Under those circumstances, it’s not up to the editor to say, “I want to cut this shot at the midway point” unless something disastrous happened in the shooting, and you have to then figure out what you’re going to do.

Every good scene has its own spine, and each of the moments internal to the scene is like a vertebra and you’re just watching it go from vertebra to vertebra and then toward the coccyx — toward the end — you do have to make a decision. “Shall we let it go all the way? Or can we cut a little sooner than was originally intended?” This is something you’ll only find out when you put the whole film together and see that undigested single shot and how it works within the context of the whole film.

HULLFISH: In “In the Blink of an Eye” you described a shot as having branches. You can’t really cut along the limb, but you can cut anywhere that there is a branch that shoots off.

MURCH: Yes, that would definitely apply – what I call “nodal” editing. Just like the branch of a tree has a linear grain, and then suddenly a ‘knotted’ grain where a new branch emerges. But a master shot that lasts for a long time is something that must have its own unique internal quality that the director and the actors have developed through staging and rehearsals. It’s almost a tree unto itself. And usually, there’s not a lot of extra coverage because the filmmakers have committed to this decision. And so there may not be anything an editor can cut to.

HULLFISH: Just in general, what makes you want to play a scene on a reaction shot — specifically quite a bit of a scene on a reaction?

MURCH: The quality of the acting: the thoughts that are visible moving across the character’s face. You’ll begin to feel this when you watch dailies.

It also will depend on the characters having their own arc within the film. Is this a moment when that character realizes something about his relationship with the other characters, something that’s going to change everything? That would be a reason to cut to that character in a reaction shot.

HULLFISH: I’ve talked to editors that discuss reaction shots in terms of empathy. That they want to see how a character feels about a question or a statement, which can be a reason, right?

MURCH: Right. Sure.

HULLFISH: But you’re putting it in the context of story — “Does that person have a reason later on in the story that you need to know this moment?” Or acting performance — “This is such a great performance I want to be on him because he’s so much stronger than this other person.”

MURCH: Imagine the situation not in a film, but in life. You’re sitting with two other people and they’re having a conversation. The first person will talk. Then the second person. A, B, A, B. But you will not just look at the person who’s talking, and only when they stop talking, look to B.

In “In The Blink of an Eye” that’s what I called the “Dragnet” style of dialogue cutting. Where we’re always on the person talking, and we cut to the other person only when he has his lines.

But pay attention to what happens in the real world: it’s a much more fluid thing, as soon as you ‘get’ the idea of what the first person is saying, you will look at the other person to gauge their reaction. So that would produce a post-lap cut – the first person is still talking when you cut to the reaction shot of person B, the listener. And then you might stay on person B while person A starts talking, which would be a pre-lap cut.

HULLFISH: Sound is also a specialty of yours. In The Godfather, there’s a great scene with sound design where Michael goes to an Italian restaurant to avenge his father’s death. There’s great use of a train in the distance when Michael kills Solozzo and McCluskey.

https://www.youtube.com/watch?v=ppjyB2MpxBU

MURCH: It’s an elevated train. The script was not written with that in mind, but Francis wanted to hold any music off until after the murder. Interestingly, that moment after the murder was where the film was originally going to have an intermission.

Godfather is the first long film not to have an intermission, but when we were still working on it, in 1972, the plan was to have an intermission as soon as Michael runs out of the restaurant. But Bob Evans, the head of Paramount at the time, nixed the idea “We don’t want to let the audience off the hook!” A good decision. His other good decision was to allow the film to be almost three hours long, overriding the contractual obligation to have it be no longer than two hours and twenty minutes.

So Francis and Nino (composer Nino Rota) had designed a big operatic moment at that moment and Francis didn’t want to dilute that by having music underneath the previous restaurant scene. And yet, it’s a fairly long scene and almost half of it is in Italian with no subtitles, which is a risky thing to do, because it throws the audience back to watching body language and voice tone to figure out what’s going on — unless they understand Italian.

So it seemed to need something to underpin Michael’s emotions during the scene. And that’s when I had the idea of using an elevated train sound, as a kind of mechanical string section. I grew up in New York, not far from where that restaurant was supposed to be. And I knew that part of the Bronx was a rat’s nest of many lines of the elevated train system. The off-screen train sound comes and goes perhaps five times during the scene, getting louder each time. Finally just before Michael stands up and pulls the trigger, the sound screeches with a metallic braking sound. Very Bernard Hermann.

HULLFISH: And it’s great emotionally. It’s like the teakettle in so many scenes in movies, where, as the tension goes up, the squeal of the kettle intensifies.

MURCH: Yes, exactly.

The final interview with Murch will be coming in a few weeks. It is a series of questions from the curious minds at the Blue Collar Post Collective Facebook group. Stay tuned!

To read more interviews in the Art of the Cut series, check out THIS LINK and follow me on Twitter @stevehullfish or on imdb.

The first 50 interviews in the series provided the material for the book, “Art of the Cut: Conversations with Film and TV Editors.” This is a unique book that breaks down interviews with many of the world’s best editors and organizes it into a virtual roundtable discussion centering on the topics editors care about. It is a powerful tool for experienced and aspiring editors alike. Cinemontage and CinemaEditor magazine both gave it rave reviews. No other book provides the breadth of opinion and experience. Combined, the editors featured in the book have edited for over 1,000 years on many of the most iconic, critically acclaimed and biggest box office hits in the history of cinema.