I have a sequence of frames, which I want to render to video. The first frame is fully non-transparent, and the next ones are border-rendered - they are partially transparent. And these transparent areas I want to have filled based on the previous frame - simply "copy" these parts from the previous frame (the non-transparent areas are the only ones, where movement or other changes occur). Is it possible to do so with Blender?
Thanks in advance for any help :)