by Chimay on 07 Dec 2009 15:01
I've shot (and edited) several music videos without timecoded playback sources, since the early 90s. As long as your playback source doesn't drift (practically any digital device these days), all you have to do is also capture the playback audio with your camera. Then it becomes very easy to sync in post by using the reference audio, not the video.
This is made even easier with modern nonlinear edit systems because you can see the audio waveforms and you can make sure that the peaks are aligned (you first lay down the audio track and lock it, then add the video cuts together with the audio). There's even software that will do this automatically for you.
An even easier way to do it is by using the multicam functions offered by Final Cut Pro, Avid, or any other software that supports it. What you do is you shoot the whole song without stopping the camera, each and every time. Then you just sync the takes on the timeline and cut between them, as if you were doing a live TV show. You can do the same thing even without multicam, although it's not as fast. Just sync each take and lay it on a separate layer. You'll end up with lots of layers, one on top of each other. By opening a space on the top layer for a take, you reveal that take on the layer underneath. So, if you desire to use the take on the third layer, just open the respective spaces on the two layers above it. It's actually a very easy and quick process.
Of course there will be times when the performers screw something up. Just have them keep going, no matter what, so that you can still use the good bits. Or, as an alternative, just restart from scratch and ignore the bad take.