Audio NEVER sync’s right

[Jeremy Garchow]

The

[Brandon Lanski] “Camera-A and Camera-B fall out of sync with these audio tracks differently. “

But at that point in time, if you listen to camera A and camera A’s audio, is the audio in sync with the video? Or is the audio out of sync on the camera recording (not the console recording).

[Joe Marler]

[Jeremy Garchow] “But at that point in time, if you listen to camera A and camera A’s audio, is the audio in sync with the video? Or is the audio out of sync on the camera recording (not the console recording).”

This is a good question. If (by chance) Cam A video and Cam A audio are ever out of sync, this implies something unusual.

However I think his problem is the one discussed in this frame.io article. Note Brandon is trying to sync 24 fps and 23.98 fps material with external sound recorded at an unknown frame rate.

“When using an external sound recorder, it too must have 23.98fps as an available choice, or there will be drift between the sound and picture…Whatever you do, be consistent. Don’t mix footage with different frame rates..”

https://blog.frame.io/2017/07/17/timecode-and-frame-rates/

Of course audio doesn’t have a real frame rate — it’s recorded in samples. But it does have metadata which may contain a frame rate, and which various software can interpret differently, especially when conforming differing frame rates. This is the point Tangier was making in his earlier post.

Brandon might be able to laboriously work around the problem, but the underlying issue should ideally be investigated and resolved. A better workaround might be to use WaveAgent to modify the file metadata then re-import. To know what file to change, first use WaveAgent to examine all the files and write down which is the oddball group. If they are all 24 fps except for the one 23.98 fps camera, maybe try to change that one’s metadata to 24 fps. But I really don’t know — this could require trial and error. If that doesn’t work, some files might require resampling or even re-encoding.

Besides this issue, Jeremy’s statement is correct that long term sync (over hours) is not assured unless you have common timecode. The quartz oscillators in regular cameras are like a cheap wristwatch. They are pretty good but can drift over several hours. So Brandon may have two different issues — a more rapid drift problem caused by the 23.98/24 conform, and a slower drift problem caused by a camera timebase.

[Brandon Lanski]

The camera’s always seem to have their audio in sync. (A with A B with B but A not always syncing up with B)

After trying to sync smaller chunks at a time, I’ve realized that maybe fcpx gets confused and has a tendency to place things in the wrong places. I don’t know how the algorithm works but I guess if two songs have transients which look similar, a clip could be placed at the wrong place.

Still waiting to get wave agent on my system. But I will report back with what it tells me