Could be useful to allow users with ffmpeg installed to use that as a sub-processed backend instead of opencv's cv2.VideoWriter (as at https://stackoverflow.com/a/42602576, (and https://stackoverflow.com/a/13298538)). If doing, make a similar interface to existing VideoWriter class for consistency and ease of use.
cv2.VideoWriter works but is quirky, and generally ends up with large file sizes even when just duplicating existing videos (likely due to decompression on read that's not recompressed on write).
Could be useful to allow users with ffmpeg installed to use that as a sub-processed backend instead of opencv's cv2.VideoWriter (as at https://stackoverflow.com/a/42602576, (and https://stackoverflow.com/a/13298538)). If doing, make a similar interface to existing VideoWriter class for consistency and ease of use.
cv2.VideoWriter works but is quirky, and generally ends up with large file sizes even when just duplicating existing videos (likely due to decompression on read that's not recompressed on write).