Inside Out

Get an uncensored look at our products and the people who make and use them.

AGENT MX-Z3RO—Making an Independent Movie (Part 6)

“Sound is 50 percent of the movie going experience, and I’ve always believed audiences are moved and excited by what they hear in my movies at least as much as by what they see.” —George Lucas

 

“The eye sees better when the sound is great.” —Steven Spielberg

 

“Video without audio is just surveillance.” —David Scally, AES

 

 

It’s my strong belief that the number one dead-giveaway of spotting an “amateur” piece of filmmaking is hearing ambient, noisy, cheap-mic-on-camera, sync-challenged production dialog. Let’s face it—since the days of the late 90s low budget but huge box office smash, The Blair Witch Project, to the more recent Paranormal Activity, people will now tolerate quite a wide array of low-resolution picture quality and handheld “shaky cam footage.” But, I would argue, if the audio is subpar—especially if the dialog intelligibility is poor—your audience will not stay with you. This is why I’d like to stress the importance of getting good production sound.

 

Have a sound supervisor in the mix from the start

One of the key elements that we planned for with AGENT MX-Z3RO (but not common in Hollywood at the top levels of moviemaking) was to have the post audio sound supervisor consult with the production sound mixer and director before the first frame of film/video was rolled. Typically, the sound supervisor isn’t even hired until after post video editorial has begun (or even been completed), and the production sound mixer, by this time, is on to another job. It’s really hard to get what you want, let alone what you need, from a sound standpoint if the two teams pass like ships in the night.

 

I learned this from my friend and mentor, Larry Blake, who has been the sound supervisor and re-recording mixer for Steven Soderbergh for many years. Larry would conduct test reel shoots with his DP, production sound mixer, and himself ahead of any big Soderbergh movie, and then test the resulting footage in Media Composer and in Pro Tools to make sure the sync/workflow was solid. It kinda makes sense, right?

 

When we started writing the script for AGENT MX-Z3RO and planning what the movie would be, Scott Weber, our sound supervisor (and my close friend), was already consulting with myself and director/co-writer Brian Barnhart about what we wanted, as far as soundtrack elements go, and how we would go about getting it. He had some solid advice for us before we started shooting, which can be seen here in this video interview.

 

 

 

 

If you want higher quality audio in your project, here are three rules to follow during the shoot:

 

1. Always run sound, never shoot MOS (aka, without sound)—Even the worst production sound can be a good guide for lip sync and gauging an actor’s emotive performance/vocal tone, which will help during the post production dialogue edit process and potential replacement. Footsteps and production prop handling are good audio cues in the same way. In truth, we didn’t always abide by this rule, and the scenes in which our Foley crew re-created the sounds turned out great, but, for sure, it took them more time to do. I’ll break down and share this scene with you in an upcoming blog about Foley and ADR.

 

2. Take 30 seconds (or more!) to capture the room tone—By this, I mean the sound or “air” on set when no one’s talking. This is critical, as it gives your dialog editor and ADR editor the best chance of being able to preserve and smooth out the production dialogue. Plus, it’ll help make the ADR match better, since they can use the tone to fill the hole where the production dialogue is pulled out. When you’re running and gunning to make 30 setups a day, this is tough to do, but will really help you out on the back end. We took it a step further and tried (though not successfully) to also capture a loud clap after we had the room tone, hoping to capture some of the ambience/reflections/reverb of the room in case we wanted to replicate it with a possible IR (Impulse Response) convolution reverb, which we ultimately didn’t use. In the video clip, you’ll notice that we ran picture for all room tone captures, which was very useful for identifying all of the takes. We then strung them together on the Media Composer timeline and exported the AAF.

 

3. Have the actors do “wild takes” on set—This makes sense to me in so many ways. On set, your actor is fresh in the moment and remembers the emotion, pitch, and volume to deliver a line. This is hard to do four months later when he or she has moved on to other projects, but trying to get back in character at a point in a scene—in a sterile ADR studio. The mics on set and ambience of the set are exactly the same as the production take, which eliminates 75% (my guesstimate) of the difficult work of matching the sound of the ADR with the original audio. I must say that we relied heavily on Synchro Arts VocALign to align the ADR takes and some alternate takes—more on this in a later blog. Another great reason for doing this (which happened to us on this project!) is that one of our key actors had major dental work done between shooting and the ADR session months later. The actor’s vocal character and tone were so different, we felt it couldn’t be matched and didn’t use any of the ADR takes!

 

 

Before we started shooting, we set up a call with our production sound mixer, Jim Machowski; sound supervisor, Scott Weber; and our audio workflow consultant, Scott Wood, to talk through the plan and determine what would be best for our needs. Jim had a 4-channel Sound Devices 744T as his basic recorder, and we used that to capture two actors in a scene for all but one of the days. We used one channel for the boom mic, two channels for two lavalier mics, and the fourth as a combined mix track of Jim’s mix of those three mics, which would be used for our video editor. (We also sent a mix of the lavs to the right F35 camera channel and the boom to the left.)

 

For our big card game scene on day two of shooting, which featured up to six actors on camera at once, we planned to use a Zaxcom Deva IV 8-channel recorder. This was provided by Coffey Sound in Los Angeles, who was interested in helping us on our quest to generate demo/training material, so we could show off the proper workflow from a field recorder, such as Sound Devices or Deva, to Media Composer and then to Pro Tools. Robert Kennedy from Coffey Sound was a great asset to help us on that day, making sure everything worked and flowed smoothly.

 

Putting the modern field recorder workflow into action

It’s very difficult for Avid (or any other company) to get full access to a major motion picture or a TV show’s raw dailies and production dialogue for obvious reasons. So why are we trying, you ask? To show you how easily you can integrate field-recorded audio into your workflow. Truth be told, we are a bit spoiled now—with today’s technology, we don’t have to deal with timecode DATs or analog Nagras and EDLs, or hope that the deck will control and ultimately conform our tapes properly.

 

But even in the more modern era, it hasn’t always been so easy. During post production for the 2005 movie Syriana, I watched my friend (and sound supervisor on the film) Larry and his assistant editor, Mick Gormaley, painstakingly spend great amounts of time pulling down a region of dialog from the OMF, opening the handles all the way to the top to identify the take, searching the bin for the alternate mic, spotting that in and trimming it up, and then putting back the piece they pulled down and opened up. We talked about how there had to be a better way, and now there is. Many people have used a great app called Titan from Synchro Arts to accomplish this work, and now much of that functionality is included in Pro Tools.

 

In fact, Pro Tools 9.0.3 improves the performance of the match criteria and such. As you’ll see in the video below, Scott Weber and I found that the default match settings seemed to work the best for our project. One thing you’ll want to do when importing an AAF is to ignore rendered effects from Media Composer, as it will return false results on the match criteria, since the file no longer has the true metadata from the recorder.

 

Check out this video to see the workflow in action—a moving picture is worth a thousand words…

 

 

 

 

Until next time!

Leave a Comment

login or create an account to post a comment.

About Thomas Graham

Avid marketing manager for Post Audio and Pro Mixing, I am also a veteran engineer/recordist/editor. I've worked on music scores for dozens of feature films, including Ice Age, Collateral, and The Spirit.

© Copyright 2011 Avid Technology, Inc.  Terms of Use |  Privacy Policy |  Site Map |  Find a Reseller