I was talking with someone last night who does marionette workshops for kids. He said he had videos but he didn't want to put them on YouTube because there were people walking in front of the camera. I said to myself, there must be some open source libraries for detecting dark shapes moving across the screen and automatically clipping that from the video... There is certainly some code for that in one of the Eye Gaze projects. Since I too ran into some encoding problems when asking QuickTime to turn its screencasts into a format for YouTube, I decided to look into it.
So I did a search on Wikipedia for a comparison of video editing software and this is what I came up with.
Blender looks to be a pretty complex 3D graphics software, not exactly what I was looking for... but considering its scriptable, perhaps there are some ways to get it done. It also seems to be rather user friendly with a pretty strong community. Indeed, the infamous Big Buck Bunny was made in Blender.
I've always wanted to have 3D animations for our linguistics representations, either to give true meaning to Halle's wirebrush visualization of tiers in phonology, or to visualiize linearlization in spell-out from syntactic structures to linear strings of morphemes. So perhaps I'll hire someone to work with Blender in the future...
Avidemux seems to be a video editing software, maybe not completely user friendly but at least cross platform and open source. I tried it for transcoding .mov -> .avi and it seems to work.