ProRec Interview: Walter Murch

PLEASE NOTE: This article has been archived. It first appeared on in June 1998, contributed by then Senior Editor Joel Braverman. We will not be making any updates to the article. Please visit the home page for our latest content. Thank you!

Walter Murch is known as the film editor and sound mixer for some of the best movies of our times – the Godfather, Apocalypse Now, The English Patient are just a few. Recently a popular Pro Audio magazine interviewed Mr. Murch in a special audio-for-video section. However, the topics discussed did not really go into technique and technology, but focused more on aesthetics. I wanted to find out more about what technology he uses when mixing for picture.

Walter Murch

Walter told me that he is the only person who actually does the editing AND the sound mixing on feature films, an approach he developed at Zoetrope which he helped found with Francis Ford Coppola and George Lucas. I dropped in on Walter at his farmhouse in Marin County, California where I asked him too many questions over a cup of tea. He had just returned from Cannes a couple of days before with his re-edit of Orson Welles “Touch of Evil”.

At work mixing Apocalypse Now

I noticed a picture in a recent interview of you in a small studio – is that your personal studio?

That was up in the barn and I was editing what you might call a Directors Cut of “Touch of Evil”, which Orson Wells directed forty years ago and any project that I take on, particularly short term projects I can just do them up in the barn, renting whatever happens to be the available and appropriate technology for that particular film. So what I had there was an Avid, which is a film editing machine, which has up to 24 tracks of sound that run along with it, but you can only actively work on 8 at one time. If you get a track the way you want it, you can make it one of the ‘Sleeper’ tracks, sort of ‘demote’ it to playback only, and then move another track up into the active area, so you can play back 24, but you can only actively work on 8 tracks. Everything just ran through a Mackie Mixer which was also feeding audio from CD’s and cassettes and DAT machines, and a DA88 which is an 8-track recorder which uses High 8 video tape.

How do you feel about working with a computer based system versus something like an analog tape machine?

WM: Well, for me it’s fine. There are some people who claim to be able to tell the difference between professional digital equipment and analog equipment. I can’t. The advantages operationally of using digital are so great, I focus on that and not on what I guess might be the “digital” sound of it. “Touch of Evil” was a film that was done in 1958, so there wasn’t a wide range soundtrack to begin with.

Were you taking the existing sound track and mixing it with some sounds that were recorded now, or…?

Well no, we had separate dialog, and music and sound effects from the original magnetic masters, so we loaded those via DA88s into the Avid, onto the Avid’s hard disk, and I was editing them, supressing the music in some cases, lifting the level of the effects in some cases. All of this was following Orson Welles notes. Where we had to make changes, we simply stole (sound) from various places within the film. The goal was to make something that still sounded like it was all done in 1958 with a minimum of disruption of that particular kind of sound.

Did at any point in the process you use any kind of noise removal software, like (Sonic Solutions) No-Noise?

I have a device here which we also used down in Los Angeles when we did the final mixdown, which is a single ended noise reduction and hiss reduction device. The Behringer DHR 2000 – its a relatively inexpensive device – around $300, but its a very effective suppressor of noise, from wherever the noise comes from – it doesn’t matter if its tape hiss, or hiss from pre-amps or hiss from a tea kettle in the background of a scene, it will do its best to get rid of it. I’ve used it on all the films I’ve done since…1990. (!)

(At this point Walter’s wife Aggie who does radio programming for station KPFA in Berkeley came in, and he went off to help her remove some low end from an interview on her personal Pro-Tools system. “We’re a Mac kind of family” said his daughter Beatrice. When he returned he had an oddly shaped black box which he plunked down in front of me)

(incredulous)….its a Radio Shack…decible meter…

Yeah, you can buy them for about $30 and when I am lining up films in the theaters, I take that in and use it to make sure that all the speakers are lined up correctly, that the level is being played back at the same level as when we mixed it. Some times the theaters play the sound too loud or too soft.

So you do that when you are having a preview for a theater full of…critics?

(laughs) Or a preview for an audience. Any time you are going into a situation with a film that is not yet finished, there are a lot of variables, so you try to eliminate as many of them as you can. Once the film is done though, there is really nothing you can do – you put the sound tracks together as best you can, and then its just up to the theaters to do the right job.

I’ve heard that many theaters have equipment that is in poor states of repair.

I think that was something that George Lucas tried to rectify and is rectifying with the whole THX program. When Dolby went into theaters, what they sold the theater was the processing equipment that took the soundtrack track off the film, sent the signal to an amplifier, a Dolby amplifier that then made a line-out signal to the power amps in the theater. The problem was that the power amps and the speakers and the wiring was up to the theater. Dolby’s responsibility merely was to get the sound read off the projector via a special optical reader, into their processor, and from there produce a line out and they could ‘tune’ it to the theater, but the power amps might be widely deficient to the size of the room, or in extreme cases, sometimes the cones of the right hand speaker might be blown… and so there would be all kinds of problems and Dolby would just simply leave a memo saying “it would be good if you fix this”, but economics being what they are, many times the owners would just say “bye, thank you” and just kept on as usual.

That’s happenning less and less because theater owners are realizing more and more the importance of sound to help sell tickets. But they were bothered by a lack of standards, so what George did was to take responsibility for the power amplifiers and the lines for the speakers, and even manufactured special crossovers that help get the best out of certain speakers. George simply did something that should have been done all along but he did it in a way that standardized everything and gave people an idea of how high the bar was set, in order to be able to jump over it.

One movie you did that I saw a while ago was First Knight…Do you recall what equipment you used on that film?

That was released in SDDS which is the Sony system to get sound in the theater. There are really three delivery systems operating simultaneously now – there’s Dolby, which has the most units out there in the world, and then there’s DTS which I think is second, (Universal’s Disk Based System). The sound track is separate from the film and it’s running on a special CD which is in sync with the image. And then there is the Sony system, SDDS, which is digital sound on the film – its different from the other systems in that it gives you 5 channels of sound coming from behind the screen, All the systems are similar to each other – they all give a low frequency channel, they all have stereo surrounds, what the Sony system has in addition is two extra speakers behind the screen.

That film was pretty conventionally prepared in terms of the sound for the period – which we did in 1995, so with a few exceptions, it was done on magnetic tracks. It was prepared the way sound has been done in Hollywood for the last 30 or 40 years.

When It comes to lining up sounds on tape, I would imagine it would be much harder to do than on a computer based system where you could just ‘nudge’ it a few frames forward of backwards.

Well if you are used to cutting on film, it becomes second nature. But yeah, if you want to shift the sync of something in film, you have to shift it physically in its relation to the image, which means cutting out frames, before or after, or if you want to change the relationship of the sound within the soundtrack itself, you have to cut into the sound, physically – you are doing physically what you do electronically or virtually when you are working on a workstation.

a lot of splicing and dicing…

Yeah, you are physically cutting and then taping it back together, you are making something – a physical object which is a reel of film which will run in sync with the image. Whereas you do that only in a virtual sense when you are working on a workstation of some kind.

What dB level do you typically monitor at when you are mixing?

The standard now is 85 decibels per channel at 0 dB vu. Once you’ve set that level the question is what systems will give you the maximum headroom above that before you reach distortion. In conventional old mono films, up until say the mid 1970’s, the maximum decibels you could get were 88, so you could get just 3 decibels over 0db vu without distorting.

Some of the new hardware is supposed to get you virtually unlimited headroom…

The limit you will meet eventually will be own ear capable of hearing this without damaging you ear.

The old mono peaked out at 88 dB maximum. When Dolby A came in the peak was raised to 91. When Dolby SR came in the peak was raised to 97 – with digital systems it could easily be 110 -115 dB, which was also true of 70mm releases. For instance, when we did Apocalypse Now, we were mastering to magnetic masters, and magnetic soundtracks can take up to 110 -112 dB without distortion. When they do distort its a progressive distortion, whereas when you distort digitally it’s instant, and you really hear it, it’s very similar to optical, where there is an actually almost physical wall that you hit where you cannot get any louder without distorting. I prefer not to go much over 100 dB So if average dialogue is reading 78, that gives you 22 dB of headroom, which is plenty.

When you are finished with a film, is that the last step, or is there a mastering step after that? Are you the last person to touch that film?

There is one more step: transfer to the digital optical format. That’s supposed to be a completely neutral transfer from one medium to another without any tweaking or interpretation. What create for a Dolby digital master is a magneto-optical disk That’s sent to a place in LA where they make an optical negative – changing the M.O. to beams of light they expose that on photographic film, and then that becomes the negative master of the soundtrack which is sent to the lab and then printed on the film. But its very rare that you would do anything at the transfer stage that you haven’t already done – it has to be a hands off, do-it-by-the-numbers transfer.

So, this is different from mastering a record where there is still a lot of final interpretation. In film that stage is where we take all the theoretically balanced final tracks and make the last combination down to say, 6 tracks, which is called a printmaster. Two tracks in the case of Dolby SR, that’s like an encoded 4-track. Sort of like FM stereo. The basis of the Dolby system since 1976 has been that they use a phase correlation system to take the three front channels and compress them down into left and right. If anything is in phase on those two tracks, its extracted and sent to the center speaker and that in-phase information is suppressed in the left and right channels by 6 dB. So there is a shoulder. The same signal is coming out of all the speakers, but noticeably louder in the center than on the sides. Then there’s an out-of phase signal which is the surround information, very much like an FM signal. That’s all technology as of 1994. Still exists as a backup, but when you go digital, every track is discreet, whereas with Dolby SR, no track is exclusively itself – its sort of like a soft stereo with mono mixed in.

Is that the similar to 5.1 mixing?

No, see the tracks in 5.1 are discreet – there in no matrixing. 5.1 is 3 speakers behind the screen, stereo left and right in the back, another speaker that’s just low frequency, placed anywhere in the theater since low frequencies are non-directional.

Do you put one of those systems together here, or at a big studio.

Recently I’ve been working over at Berkeley over at Fantasy Records. Every studio that they have has a 5.1 setup.

What just happened in the last three years is that Tom Hohman, who created the THX system for theaters, has gone off on his own and created something called micro theaters – 5.1 systems you can put in your home or your studio – they are very carefully balanced and weighed speakers, such that you put yourself in the right place what you will hear will be identical to what you hear in a big theater.

The English Patient. What process did you follow for mixing that?

I produced a ‘guide’ track on the Avid, and then that was taken and transferred at a higher quality, onto a Sonic Solutions system at Fantasy Records, and then coming out of the Sonic solutions, after it had been cut, we would make transfers either onto 6 track film, or DA88’s

What we just did on “Touch of Evil” because I was working on the finished soundtrack right from the beginning was to take my Avid sessions and re-create it, opening it up through an OMF (Open Media Framework) file and convert it into ProTools, which is another sound editing situation (Digidesign and Protools are both owned by Avid). That was a real timesaver, in the sense that all of my decisions cutting and fading in and fades out and level setting were maintained when the sound track was opened up in the ProTools environment.

So all they had to do was to tweak what I had done and refine it, because the tools that they have in ProTools are much more precise than what I have on the Avid.

On The English Patient all they really had to rebuild everything that I had done from scratch which was a time consuming process.

What audio systems are most common in the movie industry?

The three systems that are most commonly used in the industry are ProTools, Sonic Solutions, and WaveFrame. It really has to do with which system you are comfortable with. ProTools is a good system, my wife is using it upstairs for the radio programs that she does.

In a recent interview you expressed the desire to be able to capture the acoustic properties of an environment. That ability is available now with acoustic modeling software. Do you think that will change the way you work?

What I used to do was to actually go to the space that I wanted to capture, have two tape recorders and play the sound from one tape recorder into that space, and record the result on another tape recorder, and in the mix, have both tracks running simultaneously. That was just because I was dissatisfied with the echo chambers at the time which simply gave you a certain kind of twangy sound. But I wanted “gymnasium”, and “auditorium” and “bathroom” and other sounds. Back in the 70’s the only way was actually to go to the space and record it there.

You’re a Mac user – do you think you would ever use a PC based system?

No. (laughs)

What are the deficiencies that you see in the technology at the moment.

We’re in a transition phase, which means a lot of redundancy 100 years ago if you had a chandelier, it was likely to be BOTH electric and gas – electric was unreliable, and gas was dangerous. Film itself, the Celluloid part of film, is threatened and on the way out. But we have to live with one foot in the Digital world and one foot in the conventional Analog world and find a way to blend those two things, at least for the moment..

Do you think that Digital Video will replace Film?

I think it will. It won’t be in the form we know it now, but I just think the economy of the whole process… I mean Godzilla made 7000 prints at… I don’t know what they cost these days, but at least $5000 – That’s 35 million dollars. If they were able to beam the film from a central station by satellite into the theaters, they would probably save many many millions of dollars. Kodak now has a digital projector that is reportedly the equal of film. It uses up a tremendous amount of memory, and its an expensive item because it isn’t built on an assembly line. It just takes something to start a process going. 6 years ago probably 30% of films were done on Avid and 70% done on film. Now it’s probably 10% done on film and 90% done on Avid.

Once those figures reach a certain level it becomes almost impossible to edit on film because all the support is headed in the digital direction, so it would become more expensive to edit on film. I think eventually there will be digital projectors in movie theaters, then it will become cost inefficient to do film.

What was Zoetrope and how did it start?

George Lucas and Francis Ford Coppola and myself were all film students who had gone to film school in the mid-sixties. When we got out we found out that Hollywood was in a very depressed state. I think 1967-1968 was the absolute bottom of people going to movie theaters ever, so there wasn’t the kind of money that there is today, or was back in the 40’s. In addition the studios were still owned and operated by the people who had owned and operated them in the 40s – Jack Warner, Adolph Zukor, Walt Disney. Anyway. But they were all reaching retirement age, or about to die, and the idea of completely re-inventing Hollywood or did not occur to them or appeal to them. So we found ourselves frustrated not being able to get interesting employment. As a result we moved up to San Francisco, thinking that if we just changed where we were it might loosen things up a bit. We decided to make all low budget films and make them on location, and try to do as much as possible to lower the cost, and also provide a stylistic unity to the process.

As far as sound goes, it was really possible at that time because of the recent invention of the transistor – audio equipment was beginning to be miniaturized and so the cost was lower and the quality was nearing the professional level, and we had an intuition that this would open up the whole process in a way that had been impossible in the 40’s and 50’s.

Ok, One last question: When you go out in the field and record sound, what are your favorite microphones to use?

You know, it’s been so long since I’ve done it… I’ve now become a film editor and mixer. That middle part of going out a recording them I actually don’t do. When I was doing it, I liked the Shoeps Microphones.

I started out with Sennheisers, throughout the 60’s and early 70’s. There was something about them that sounded very good when you heard them on immediate playback, but when you had gone through several generations of transfers, the sound became a little peaky in the sibilant regions. Shoeps at the time had a little bit of a softer sound, but it aged well through different generations. That isn’t really a problem any more with digital.

With analog, if things were calibrated right, the noise reduction systems are very good. On Apocalypse Now we had a 24 track with DBX noise reduction. We daisy chained a sound across all 24 tracks – copied the sound from track 1 to track 2, track 2 to track 3 track 3 to track 4 etc, and there was really no noticeable generation loss, even by the 24th track.

Well that should blow some illusions about what you can and cannot do with analog tape.

The Murch Resume

The Rainpeople (1969)
THX-1138 (1971)
The Godfather (1972)
American Graffiti (1973)
The Godfather, Part II (1974)
The Conversation (1974)
Julia (1977)
The Black Stallion (1979)
Apocalyspe Now (1979)
The Right Stuff (1983)
Return to Oz (1985)
The Unbearable Lightness of Being (1988)
Ghost (1990)
The Godfather, Part III (1990)
The Godfather Trilogy (1991)
House of Cards (1992)
Romeo is Bleeding (1994)
Crumb (1994)
First Knight (1995)
The English Patient (1996)
Touch of Evil (reconstruction) (1998)