At the Blender conference 2012 At Mind organized an open discussion about Blender compositor. During this discussion the current issues were discussed, some proposals were made and we discussed what our next step should be. Here are the notes.
- Access compositor from video sequencer: Make the compositor accessible from the video sequencer.
- Border compositing: Let the user select an area and only calculate this part of the image.
- Buffering: Do not recalculate the compositor on every change, but only that what has been changed. Make it possible to record and playback a number of frames.
- Canvas compositing: Make it possible to drag/rotate/scale images inside the Node editor (backdrop) and Image editor in a friendly manner.
- Deep compositing: Deep compositing is a technique where multiple colors per pixels are stored and can solve alpha-blending and occlusion issues.
- Plugin API: Having an API to handle custom nodes.
The second part of the discussion was about the top priority. The top-three are Buffering followed by Canvas compositing and Border compositing.
The last part of the discussion was about what to do next. The community thinks that these projects can be crowd-funded, but needs to be marketed well and validated by the committers.