Save Your ADR: Re-Record Live in Audition

Save Your ADR: Re-Record Live in Audition

  • Post author:
  • Post category:Edit / Mix

Recently, we were tasked with a difficult problem on a film we were working on. The editing staff had done a splendid job of inserting layers of sound effects, ambiences, music and dialog, but because it was shot in German, the English dub was SUPER CLEAN. So clean, in fact, it was taking audience members out of the story. Adding simple EQ or distortion wasn’t going to solve this problem, and we had to resort to some strange methods. In this article, we’ll explain how we saved this uber clean ADR. You Re-Record Live.

The video below explains how the process went and gives you examples of the various iterations of the process and their results (including the original ADR)

The Film:

Sky Shelter is the story of freedom fighters in a futuristic post apocalyptic earth who fight for their lives while bounty hunters track them down. There’s all manner of futuristic sound effects, orchestral music, forest ambiences and unabashed complete sci-fi audiophonic worlds. The mix had been expertly layered with sound effects. Even the weapons sounds were 8 tracks deep. We really appreciated the detail the employed to make it sound great. However, we did run into many issues with translating their Premiere Pro files into Audition. Look to this article to know how best to prepare your files for export to a post house.

You can see Sky Shelter Here.

The Problem:

As you can hear from the examples in the video: the ADR had been recorded in a perfectly quiet isolation room with a large diaphragm condenser microphone. The tracks were nearly perfectly recorded. But for ADR, we NEVER want to do this. As I describe in the ADR section of the Cinema Sound Education, we want to use the same kind of microphone we used while on set. We also want to position that mic in the same way it would have been done on set.

But since this was not done, and the audio was SO clean, we had limited solutions. Of course, we could always employ a plugin like SpeakerPhone to dirty up the sound – as well as employing all manner of EQ plugins and distortion to try to make it seem more real…but we knew it was never going to bring the kind of realism that the actors running around in a forest, laved, boomed etc. would bring.

Fortunately, Mark and his wife live in the woods outside of Nashville, TN, and they happened to be very near the kinds of trees seen in the movie. Mark got a wild idea:

To recreate the realism of shooting in a forest, what if we took a shotgun mic and a good two-way speaker and re-recorded the dialog from the speaker into the mic at 75 dB. The reflections from the forest would get picked up by the shotgun mic as they would be on set. But then also use a Mid Side mic to pick up any additional reflections from the forest as well as automatic forest room-tone. THEN, move the speaker around as if it were the head of the actor to create phase variances between the mic and the two-way speaker – as well as picking up those variances on the M/S mic. Use BOTH MICS in the final mix, but pan the M/S hard left and right (without transcoding it into stereo) and leave the shotgun dead mono.

For any of you who have been through Cinema Sound you’re already cringing. We NEVER – EVER have more than one microphone played in the mix – even if we record with more than one. And we certainly NEVER – EVER mix a stereo mic with a mono. WAY too many phase problems. And in any other situation, we’d agree with you.

But what we’re trying to do is take perfectly recorded audio and make it sound like it was not-so-perfect. We want to Ben-Burtize the audio to give it realism (like he did with the lightsaber sounds).The Film

The NTG-3 with Dead Wombat
The Zoom F8 with the M/S mic attached

The Gear:

The gear we used were the following:

  • Zoom F8 Audio Interface
  • Adobe Premiere Pro DAW
  • Røde NTG 3 with Blimp and Dead Wombat
    Zoom M/S Mic (attached to the Zoom)
  • JBL 305 MKII Speaker
There wasn’t much gear needed to do this. Mark simply played back Audition with the clean ADR, piped that sound through the 305s, and recorded them simultaneously through the Zoom F8 and back into Audition to stay in perfect sync.
Most folks don’t realize that on the back of the Zoom recorder is an interface for the H-series detachable microphones. We put the M/S mic directly on the back of the F8 and angled it accordingly (toward the forest). The NTG-3 we pointed directly at the speaker.
As the tracks were played, Mark moved the speaker around simulating the angle of an actor’s head to a shotgun microphone.
Sort of fun, really.
It was important to be fully surrounded by trees to simulate the forest ambience
Mark used the speaker's box to simulate a hooded actor

The Post Process

In post, after some strong denoising from the slight wind and leaves (it was December when we recorded and leaves had fallen everywhere), we were able to get a clean re-recorded set of files.

We did no special magic in combining the mono and stereo recordings from the NTG-3 and M/S mic. We just massaged the volume until it sounded like the M/S mic was ambience and the shotgun was the foreground sound source.

The results were beyond our wildest dreams.

Because of the phase variance between the NTG-3 and the M/S mic we had an undeniable realism as Mark move the two-way speaker in a performance. The phase variance alone between the woofer and tweeter on the shotgun mic were enough to create realism, but added to the M/S off-axis stereo sound – it’s almost like you’re standing next to the actor.

This actually created some problems when we needed to do DTC (Distance To Camera) mixing, as the sounds were so real, making them sound far away required us to mute the M/S mic in many cases.

We were also able to get a beautiful reflection from the old oak trees and reflective leaves on the ground which both the NTG-3 with its extra wide super cardioid polar pattern and the M/S mic expertly recorded.

Naturally, Mark doesn’t go anywhere without also recording impulse responses. He took his IRs and placed them in the Waves IR-L convolution reverb plugin and nearly perfectly re-recreated the refections of the forest for added ambience and realism when actors got loud or needed to sound far away.

To be fair, the success of this process was due, in no small part, to the fact that the forest outside of Mark’s home was nearly identical to the forest in the picture. We don’t recommend doing this for actors on sand dunes. In fact, a trip to some deserts would be the only course there. But even if we’d have done this process in a large room, the results in realism, de-perfecting the ADR, and giving space to the dialog would still be a powerful asset to the production.

Mark Used Sweatpants to simulate a masked actor

Join us on InstaGram, and stop missing out on all the fun! We do giveaways and all manner of things daily.


Join us!

Adobe Audition 1-to-3.This image has an empty alt attribute; its file name is Screen-Shot-2019-03-02-at-11.52.23-AM.jpgLearn Adobe Audition today!

Learn Adobe Audition the easy way, without all the difficulty, expense and time burn. Cinema Sound Educator, Mark Edward Lewis, takes you from the most basic understanding of Audition, to intermediate steps, and Hollywood-level sound mixing. The power of Adobe Audition can be yours in under 3.5 hours.

Share This