ď...plans are useless, but planning is indispensable.Ē Dwight D.
Planning for postproduction can be a mystifying process. There are so
many variables involved that it may seem pointless to plan anything at
all. While there are true-life tales of technical horror and ruined
finances, and Iíve been involved in a few, in the end itís usually not
so bad. Rather than preach a litany of correct behavior, Iíll take you
step by step through the audio post of an actual project with a modest
budget. There were enough bumps to make it interesting, so Iíll give it
to you straight.
The film is ďWell Fed & ComfortableĒ, written and directed by Lou Howe,
a Harvard undergraduate film student who has graciously consented to
air his process. All production personnel were undergraduates and
volunteers. Louís script was good enough that he was able to cast a
top-shelf professional actor, Tony Roberts, in the film.
The usual Harvard undergrad project is shot on 16mm and cut on a
flatbed. Lou had done one film this way and found the process
satisfying. However, for his thesis project, he wanted to carve out his
own path and work with the kinds of systems that are widely used
professionally, so he purchased a copy of Final Cut Pro for his laptop.
He set out to do as much of the work as possible on his own system, but
knew that he would do a video online and a professional sound edit and
mix. His goal was to continue working on the film using his own gear
after the close of the semester, and to be able to learn as much about
current post practices as possible.
He shot his film on 16mm with sound recorded to a Tascam non-time code
DAT machine. Slates were done the old fashioned way, with a clapper.
Lou shot his film in New York City over several weekends; that and the
distance from Cambridge resulted in different people crewing on each
day of shooting.
Things were going well until the imposition of a true New York moment.
At the end of a shooting day, some one stole the sound bag with the DAT
machine from Louís car. The machine had the tape with the dayís sound
in it, and the audio was lost. Of course, this was the day with Tonyís
scenes. Lou decided to defer the problem to post and ADR the scenes.
The rest of the shoot went as planned, with the usual mix of good and
not so good recording quality that tends to accompany student
Lou transferred the film dailies to Digibeta masters MOS and dubbed to
DVCam for editing in his Final Cut Pro system. He checked out another
DAT machine from the Harvard production office and plugged it into his
Mac via a thirty-five dollar iMic interface, through which he
transferred the audio for editing. He synchronized his dailies in FCP
and then began cutting his film. He wasnít particularly happy with the
overall sound quality, but assumed that it was inherent to the
production recordings and went ahead with the edit.
At this point Lou called on me to engage in some complimentary audio
post supervision. We discussed the missing sound and the cost and
effort involved in replacing the dialogue. I explained that looping a
scene with no production sound as a reference would probably result in
poor sync, and asked if he might be able to re-shoot the scene. He
discussed it with Tony, who did not have the time to re-shoot, but felt
he could do the ADR in an hour or two. We also planned the rest of the
audio post at that time allowing 2-3 days in total for ADR, sound
editing, and Foley as well as a day for the mix.
We started with the ADR. Because he had edited the two problem scenes
with no sound whatever, Lou wanted to bring the ADR selects back into
his FCP system to sync and be sure the cut still worked before
proceeding with the rest of the post. One of the big problems with
looping an MOS scene is determining the actual lines the actors spoke
on the set. Lou reviewed the shooting script and spent an evening
scrutinizing each mouth movement to determine the exact lines spoken.
He got ninety percent of it figured out, and would consult with the
actors for the rest during the session.
As the ADR date approached, I discussed the project with DuArt sound
designer and mixer Kevin Wilson. An experienced dubbing engineer, he
knew whatever was looped would also require the addition of detailed
ambience and effects, including Foley of all movements. A couple of
days before dubbing, Lou provided Kevin with final picture on tape for
the two scenes we had to loop. We digitized the picture into Pro Tools
and we were ready to begin.
We did separate looping sessions with Tony Roberts and then Mike
Hoagland, the other actor in the two scenes. It took about 4 hours to
do the thirty or so lines. Both actors were well prepared and supplied
all of the energy necessary to make the scenes work. With the addition
of Foley and ambience, we all felt that that the ADR would be
Around that time, I met with Lou in Central Park to do some field
recording on the Fostex FR2. We got some great stereo ambiences that
perfectly fit the looped scenes, as well as wild sounds of joggers and
close micíed footsteps. This location-specific sound would be essential
to help the ADR play transparently. The flash card format of the FR2
meant that I was able to log all takes in the field and transfer them
by USB as wav files into our Pro Tools system. Zippity zip!
Then we discovered another problem. Kevin noticed pervasive noise in
the OMF that Lou gave us from his Final Cut system, so we set up an A/B
test. When we listened to the original production DAT against the sound
from Final Cut, we were distressed. The iMic hardware that Lou had used
to load in his audio was not of high enough quality for the post
process. Though the audio was at full resolution 48 Khz, the
quantization noise from the cheap converters was clearly audible to all
of us when compared to the original DATís. Rather than try to de-noise
it, which would have added artifacts, we chose to re-import the DATís
as sound files in Pro Tools, and then give the files to Lou to re-sync
After re-transferring the DATís and handing them off to Lou, he called
me to tell he had found the sound from one of the missing scenes! There
was no sound report from the production mixer, a fellow student, and
there were no logs for the missing tapes. When Lou listened to the
sound files from the newly transferred DATís, he heard the elusive
voices of Tony and Mike. The embarrassment of poor logging aside, we
were very happy that the original takes had been located. Kevin chose
to use production sound on the newfound scene instead of the ADR, which
meant we now had only one looped scene. Since we had not yet done the
Foleys, it saved us a bit of recording and editing time.
We did the video online and the sound work at the same time. We figured
that Kevin could get through the editing in two days, and the mix would
take less than a day. We knew that the mix might go a bit long due to
the ADR, which requires a more detailed approach.
The big day finally arrived, and Kevin and Lou started mixing on a Pro
Tools HD3 system with a Control24 mixing surface. Kevin used a favorite
tool, the Cedar DNS1000, to de-noise the sound from a couple of
troublesome scenes with high levels of traffic noise. In one scene, a
few of the lines recorded off mic didnít blend in well. He programmed a
reverb from the Waves Renaissance plug-in to help smooth it out. In an
unusual scene that takes place inside a tent in an apartment during a
party, he cheated the music down to emphasize the emotional space in
Kevin said, ďLou had a clear vision of what he wanted, which made the
mix easy because he could make decisions quickly.Ē Kevin also noted
that the editing and particularly the A/B test, where he pinpointed the
badly transferred audio, was a real lifesaver. ďIf we had been trying
to figure that out and clean it up during the mix, it would have taken
a lot longer and certainly wouldnít have sounded as good in the end.Ē
Lou felt good about the mix. He particularly enjoyed the collaboration,
and felt comfortable speaking up to get what he needed. We mixed the
film in stereo and finished it on Digibeta for theatrical screening at
Soho House in New York, where the invited crowd had a visibly good
time. Lou is in the submission zone now sending the piece off for
festival consideration. I believe it will get some good play.
I asked Lou what heíll take away from the project. He said that he made
a lot of mistakes and learned from all of them. I would add that
despite the technical problems, he ended up with a great little film.
The writing and acting are excellent; the film plays well sonically and
visually. He may have ultimately figured out all of his problems on his
own, but I think that his decision to use a post facility for finishing
work was a good one as we were able to help him resolve some
accumulated technical issues. Although there were a number of wrinkles
to iron out, the budget didnít get out of hand. He ended up spending a
bit more for the re-transfer of the DATís, but the rest came out in the
Ok. So, things didnít go exactly as Lou originally planned. All told, I
would say that this project was on the difficult side of average in
terms of technical issues. But with the dizzying selection of workflows
out there, itís unusual to have a project that doesnít have some kind
of quirk. The important thing is that we got it done right, and now the
film is out there finding its audience.
Carmen Borgia is the head of audio services for DuArt Film & Video
in New York City. He oversees a post production sound department that
provides mixing, sound design, restoration, transfer and printmastering.
His department caters to independent projects in all formats from mono
optical up to digital 5.1.