Audio recording & syncing with different frame rates
- Thread starter IKONIC
- Start date
Unlike video or film, audio does not operate at a frame rate. There are audio recorders that will record time code, but that's just a stream of reference data that has absolutely nothing to do with the audio itself.
Audio runs with a sample rate: 33kHz, 44.1kHz, 48kHz, 88.2kHz, 96kHz, etc. At that, :60 of audio is :60 of audio, no matter what the frame rate of the matching video. So you can sync the same audio file to two shots of the same subject from different cameras at different frame rates... if you really wanted to (not that shooting two cameras at separate frame rates is something you'd necessarily want to do).
However, without a link between audio and video recordings to make sure that they are running on the same clock, you may (read: will) experience drift over time. And unless your camera and recorder have sync connections (which the Zoom does not), you won't have any master clock to keep them running together. Your audio may run slightly faster, or slightly slower, than your video. To correct this, simply perform a slight time-compression or expansion of the audio as needed. I find that my Zoom H4 usually runs a little longer (slower) than my DVX, so adjusting the audio length to about 99.9% of the original is usually enough. This may differ based on your camera.
Audio runs with a sample rate: 33kHz, 44.1kHz, 48kHz, 88.2kHz, 96kHz, etc. At that, :60 of audio is :60 of audio, no matter what the frame rate of the matching video. So you can sync the same audio file to two shots of the same subject from different cameras at different frame rates... if you really wanted to (not that shooting two cameras at separate frame rates is something you'd necessarily want to do).
However, without a link between audio and video recordings to make sure that they are running on the same clock, you may (read: will) experience drift over time. And unless your camera and recorder have sync connections (which the Zoom does not), you won't have any master clock to keep them running together. Your audio may run slightly faster, or slightly slower, than your video. To correct this, simply perform a slight time-compression or expansion of the audio as needed. I find that my Zoom H4 usually runs a little longer (slower) than my DVX, so adjusting the audio length to about 99.9% of the original is usually enough. This may differ based on your camera.
Richard Crowley
Veteran
In addition to everything C2V said (with which I completely concur), I should also add that for short-form clips (shooting individual scenes in a scripted production), the drift between non-sync audio and video equipment is typically not enough to worry about. At least that has been my experience with pro-sumer gear.
OTOH, for longer stuff (over 10 minutes?) there may be noticeable drift. HOWEVER, using a 10-minute (or longer) clip of video is so deadly-dull that nobody does that unless it is absolutely necessary. We typically have reverse-angle or close-ups or inserts or "B-roll" or other video that is slotted into the timeline to keep it interesting. And every point where you switch video is an opportunity to "pull-up" the video sync to match the audio track.
My preference is to use the original, unmodified audio track as the "master" and pull up the video to match it. IME, it is more reliable and less noticeable to add or drop random frames of video to keep pace with the audio than the other way around. Converting between, for example, 44.1K sample rate and 48K sample rate is pretty easy to do without significant artifacts. However adjusting audio by small increments sometimes introduces noticeable artifacts. That has been my experience, but it is probably very dependent on exactly what software (and what vintage) you are using.
OTOH, for longer stuff (over 10 minutes?) there may be noticeable drift. HOWEVER, using a 10-minute (or longer) clip of video is so deadly-dull that nobody does that unless it is absolutely necessary. We typically have reverse-angle or close-ups or inserts or "B-roll" or other video that is slotted into the timeline to keep it interesting. And every point where you switch video is an opportunity to "pull-up" the video sync to match the audio track.
My preference is to use the original, unmodified audio track as the "master" and pull up the video to match it. IME, it is more reliable and less noticeable to add or drop random frames of video to keep pace with the audio than the other way around. Converting between, for example, 44.1K sample rate and 48K sample rate is pretty easy to do without significant artifacts. However adjusting audio by small increments sometimes introduces noticeable artifacts. That has been my experience, but it is probably very dependent on exactly what software (and what vintage) you are using.
