This Week: Memorial Day Sale | Shop Now »

John Purcell on Dialogue Editing for Motion Pictures

Nov 06, 2010

John Purcell is author of Dialogue Editing for Motion Pictures: A Guide to the Invisible Art, the definitive book on the subject. In this interview with Waves, John was kind enough to shed some light on this somewhat mysterious corner of the audio universe.

What's the main goal of a dialogue editor?

A dialogue editor is responsible for every sound that was recorded during the shoot. He takes the more or less finished film from the picture editor, makes sense of the edited sounds, organizes them, and finds out what works and what doesn't. The dialogue editor wades through the outtakes to find better articulations, quieter passages, sexier breaths, and less vulgar lip smacks. He replaces washy wide-shot sound with clean close-up takes, establishes depth in otherwise flat scenes, and edits tracks for maximum punch and clarity.

Dialogue editors also work to remove the filmmaking from the film. Dolly squeaks, camera noise, crew rustling, and light buzzes must go; otherwise, the magic of the movies is compromised. These editors help present the actors in their best light, quieting dentures, eliminating belly noises, and sobering slurred syllables. And when the production sound can't be saved, the dialogue editor is involved in the ADR process, that is, the re-recording of voices in the studio, to replace problem field recordings or to beef up performances.

Dialogue editing is all of these things and more. Dialogue is what makes most films work. The dialogue editor makes the dialogue work.

Could you describe your working environment?

Usually, I edit in a small, quiet cutting room, with good sound isolation and really quiet air conditioning. These days, I cut on a Pro Tools rig, although in the past I've used AMS Audiofile, Sonic Solutions, and mag for dialogue. I monitor with Tannoy loudspeakers. The idea is to have sound in the cutting room that is reasonably close to that of the Dolby mix room, thus avoiding too many embarrassing surprises during the mix.

What are some of the problems you encounter as a dialogue editor?

There are two basic kinds of problems: production problems and problems with the tracks. Production problems include not getting all the materials that you need to get started, receiving materials that aren't to spec, endless changes after the picture is locked, actors who can't perform ADR, inexperienced directors who don't have a clue, and producers who don't pay.

Track problems range from problematic recordings to noisy locations to clothing rustle with lavaliere mics to off-microphone shots, to scenes that just won't cut. The list goes on and on.

The greatest frustration of any editor is that there's rarely enough time or money to give the film the love it deserves.

Does the technology at your disposal affect how you approach solving these problems?

Of course technology has changed the way we work. Although editing is editing, you think differently when you're sitting in front of a Steenbeck rather than a DAW. The thinking vs. doing ratio has changed dramatically, sometimes for the better, sometimes not. Technology has not solved the production problems; after all, these are people problems, not sound. But technology has drastically altered the way we address track problems. The ability to quickly organize a scene, to remove transient noises, to clean and modify unseemly room tones, and to easily fix shot transitions are but a few of the benefits of the modern way of working.

Digital audio workstations allow endless opportunity for changes, for new versions. This is a dual-edged sword that can provide for painless creativity or doom you to follow the whims of a capricious director.

For me, perhaps the most serious weakness of digital, random access editing is that it's too efficient. There are no natural pauses in the process, as there were during the mag days. You can edit — quickly — until you drop. Today, there's neither time to ponder the movie or your dialogue strategy, nor to take a moment to rest your brain and your back.

How much time do you spend on a typical project?

On a typical dramatic feature, I'll spend between 250 – 300 hours editing the dialogue and ADR. This doesn't include the ADR recording, which can vary wildly between zero and 50 or more hours. On a normal film, we spend about a day per reel pre-mixing the dialogue.

How does one learn the skills to become a dialogue editor?

The classic way to learn the trade is to work under someone who knows more than you. Whether you're a union apprentice or an intern at a studio, nothing beats watching, asking questions, trying and failing and succeeding, and having a critical, interested boss to show you the ropes.

There are numerous film schools that teach film sound, but I've yet to find a school that offers serious training in dialogue editing. It's just too much of a niche craft to warrant its own full curriculum.

And, of course, there's my book, Dialogue Editing for Motion Pictures: A Guide to the Invisible Art.

How does the Waves WNS fit into your workflow?

I'm a dialogue editor, not a mixer, so I bring my edited tracks to a large mix room for the dialogue pre-mix. This means that I rarely process tracks in the box, but instead I leave this for the mix, where there's a Cat. 43, a Cedar DNS, a console, and other processing tools. However, there are times when I must do some cleaning in the editing room. If, for example, there's a scene with one shot that's full of noise, while the rest of the scene is rather quiet and cuts well, I'll clean up the noisy shot so that I stand a chance of making the entire scene sound like...well, a scene. Sometimes I'll noise suppress an entire scene, just to see if it can be "saved" in the mix, or, if I must, spot the scene for ADR recording.

As budgets drop and editing and dialogue pre-mix times are being slashed, I find myself doing more and more noise reduction in the cutting room. When I clean tracks in-the-box, I have to be very gentle, so not to "overcook" the tracks, and I must absolutely avoid artifacts. Now that I have the Waves WNS in my bag of tricks, noise reduction is easier and faster, and the resulting sound is much more cinematic. I can quickly use WNS as an AudioSuite processor to tame a noisy track without creating artifacts. I always err on the side of under-processing, so that the mixer has some "wiggle room" and there's no worry of hearing the awful sound of overcooked dialogue. Plus, I create an edited, muted, unprocessed version of the regions to be processed, in case the mixer decides he doesn't like my noise reduction. But that rarely happens.

What are the advantages of WNS over hardware noise suppressors?

The most obvious advantage is cost. But mobility is also very important. My WNS goes with me on my iLok, so I can work under the best or worst of conditions and still have this great cleaning tool.

What's your overall impression of WNS?

WNS is a valuable additional tool in my noise reduction arsenal. I use X-Click and X-Crackle for small clicks and radio microphone rustle. I like Renaissance EQ for gentle rounding, Q10 for harmonic notches, and Z-Noise for my broadband cleaning. Sometimes I'll do some light noise reduction with Izotope RX and then apply Waves WNS for the finish. This makes for a lovely sound. But I've not found a broadband noise reduction software solution that's as effective yet gentle on dialogue tracks as WNS.