Parameters for each pass in two-pass MPEG2 encoding
I use ffmpeg to convert videos to DVD (MPEG2) format, using two-pass
encoding for the best results. My question is whether the settings used
in both passes need to be the same, or if some optimizations can be left
out to make a faster first pass.
- Must the two passes be run with the same bitrate?
- I generally use -sws_flags lanczos. Video scaling is almost always
necessary. I have a feeling this should be used in both passes, is that
- What about settings such as -mbd 2, -flags trell, -cmp 2, -subcmp and
similar optimization flags? Must they be present in both passes, or the
information gathered in the first pass is independent on their use?
When one burns one's bridges, what a very nice fire it makes.
-- Dylan Thomas