It is common practice to record audio from movie test screenings and align it to the same version in Media Composer to see what is working or not in a particular version of a movie. Some films can have 10 or more test screenings before they are released. Importing the audience reaction allows the editor and director to pinpoint directly how scenes are playing out, be it for horror, comedy or other emotion they are trying to achieve in the story.
One thing to keep in mind when bringing the audio recorded from a test screening, is that DCP projection which is what is used for these screenings, is 24.000fps. Most digital productions working in NTSC based countries shoot and record at 23.976. When a proper DCP is made from a Media Composer 23.976 sequence, the film gets “pulled up” to 24.000 to meet the DCP standard. See related blog. Now when importing the audio back into the original project, the audio file is in sync with 24.000fps playback, and not 23.976. This results in the audio from the test screening to drift ahead of picture; roughly +1 frame every 00:00:42:16 and will be 01% out of sync by the end.
There are two ways to deal with this:
#1. Similar to 24.000 productions posting at 23.976, the recording in the theater could be done at 48.048kHz. When importing this file into Media Composer, it will be pulled down to 48.000kHz via a sample for sample import/playback slowing it up .1% and will be in sync.
#2. It is most likely that the recorders being used in these situations do not have the same control as professional audio field recorders, so the pulldown trick needs to be done in post before importing back into the 23.976 fps project. If it is an actual 48.000kHz file, then you need to change its sample rate via Sound Devices Wave Agent. Unlike field production, since the file is already 48.000kHz, you need to change it to 47.952kHz and not 48.048kHz.
Importing the file with the same pullup/pulldown settings in Media Composer will result in a file that is slowed down, now matching sync with the original timeline. If the file was recorded as an MP3 or other sample rate, I suggest first importing it into a 24.000 project in Media Composer with settings to not pull up or down, then exporting as a WAV at 48kHz, then doing what is described in step #2. For other frame rate/sample rate/pitch calculations, refer this blog that provides a handy spreadsheet to do the math for you. If in PAL based countries and post is being done as 25.000 and the DCP was made as 24.000 (via a slow down), the same steps will apply but the user can manually enter 50,000 in Wave Agent. Be sure to click “save” in the bottom left corner when using Wave Agent to ensure that the sample rate has been applied to the file(s). And of course, do these changes on a copy and not the original.
And as a final tip, if you are not getting the expected shorter/longer clip duration with the setting as seen above, turn off the second check box indicated by the arrow and try again. Despite the wording indicating that resulting files will be longer or shorter in duration after the import, I have found turning this off will actually change the duration in some scenarios.