I’m working on a film that needs stabilization. And I’ve to admit that after some years of staying far from Shotcut my favorite video editor wasn’t doing a great job in this particular case.
I’ve 2 videos sources (2 cams) and I’m doing a video that uses those 2 sources
- wide angle shot, with tripod, super stable but too far sometimes
- close shot, filmed with zoom by hand by someone that forgot that tripod has been invented
I’m trying to find the less “catastrophic” moments on the second cam clip so I can then remove trash moments (cut the clip) and adjust to match music.
So to do this I really need to stabilize before cutting, as I need to see the stabilized result to be able to decide at what points I can cut my stab video.
As you know, for now (or I really missed something here) when a stabilized clip is cutted, it looses all stabilization information (as explained here).
Would it be hard to code the cut of stabilization data ? As I’m not an expert on it, I’ve the feeling that it wouldn’t be too hard as I imagine stab data being a big array of transformation matrix (that defines translation, rotation and zoom), and cutting this array at a certain timecode shouldn’t be too complex.
Of course I’m aware that if a Shotcut dev reads this he probably would wants to insult me and the task is way more complex !