I've come across this several times in the past but never gave much thought to it. Until now, when there's no way around it. Let's just start with what needs fixing first:
I wrote some music for a half hour dance recital. This was transfered onto a PC (from a USB key) in glorious 44.1 quality, so the light designer could program the cues. He would also be playing the music from his program so I really didn't have much to do once I set up the sound system.
He would also be recording the show onto a camcorder for archival purposes. This is where the fun starts. He recorded the audio in magnificent 48k and tried to sync it up with my original audio and of course it didn't sync up. I doubt the difference in sample rate is the culprit. Here's why:
A while back I wanted to record a live show I was doing. I had an output straight from the board into Audition, recording at 16/44. I also set up a Zoom H1 to record the room and audience also at 16/44. It didn't sync up.
On another occasion I did the same thing, except instead of the PC I recorded the board onto another Zoom H1. It synced up.
This tells me it might be a word clock thing. Since there's no way of syncing word clocks with these portable recorders, is there a way to fix this in post?
Now with this video, what can I do? Maybe if I had the original audio, I could "stretch" the original track to make it fit, but that doesn't seem right, and I'm not even sure it would work...
Ideas? Should I be brushing up on jitter? :biggrin:
Comments
This is to do with lack of synchronization between clocks and ha
This is to do with lack of synchronization between clocks and has nothing to do with jitter, at least to first order.
The 48K/44.1K is an obvious difference, and you can do sample-rate conversion (SRC) to get the 48K down to 44.1K, and then you would be at a similar point as having the 44.1K recordings with no sync beween them.
On most of the occasions when I have had to rescue video shoots with unsynchronized audio, I at least had the luxury of multiple cameras, so I could make sure that any one camera shot was lined up with the audio at that point, and then cut to another camera before the audio sync drift became evident. The one time where I knew we would be working with only one camera, I made sure the camera operator had captured lots of audience shots, band rear views and other non-specific footage during tune-ups and announcements that we could cut to during post to allow us the chance to regain visual sync.
I don't think it's worth trying fancy audio stretching or other audio processing until you have examined all the possibilities of cross-cutting the video. It will look a bit odd to have only one or two cross-cuts per song, so you may have to invent a few more to soften up the viewer and disguise the real purpose of re-syncing the full shots with the audio.