After some more testing suggested by compn: - Checking the timestamps of the packets (both dts and pts) -> the patch doesn't change them - Checking mp4->mkv conversion: before/after produces the same binary file, except for some padding bytes in the header - Checking a MOV with many streams: tested with 48 streams. Behaviour is ok, memory usage is almost the same (see previous mails), perf is same or better (including images).
Tested for both per-stream and per-media. I'll submit a v2 with James' suggestions. On Mon, Nov 25, 2024 at 12:53 AM compn <f...@hawaiiantel.net> wrote: > > On Sun, 24 Nov 2024 18:14:03 +0100 > Arnaud Masserann <arnaud1...@gmail.com> wrote: > > > Good question. I just benchmarked all the .mov files in the samples repo. > > On average, there is a 1.3x speedup. A few files are slower, but 75% > > of the files have at least a 1.11x speedup, and 25% of the files have > > at least 1.49x. > > > > Outliers include openquicktime/aletrek-tga-8bit.mov (0.8x, so slower), > > and wmv3/Bluem.mov (10.4x). > > Full data: > > https://etconlyview-my.sharepoint.com/:x:/g/personal/a_masserann_etc-onlyview_com/EXomK72G3LdJgoBHCgnYcBUBnxd0FAe90nrIkL4EFZKEuw?e=vj3TRl > > > > And FATE passes with valgrind, btw. > > thanks for testing. if the change doesnt break any old files, i have no > issue with it slowing the speeds of demuxing on old weird files. > > more ancient mov samples if you feel like doing more testing. > https://samples.ffmpeg.org/A-codecs/MACE/MAC6-mov/ > https://samples.ffmpeg.org/A-codecs/QDM2/ > https://samples.ffmpeg.org/A-codecs/QDMC/ > https://samples.ffmpeg.org/A-codecs/qclp/ > https://samples.ffmpeg.org/A-codecs/qtNONE/ > https://samples.ffmpeg.org/A-codecs/qtfl32/ > https://samples.ffmpeg.org/A-codecs/qtraw/ > https://samples.ffmpeg.org/A-codecs/suite/3G2/ > https://samples.ffmpeg.org/A-codecs/suite/3GP/ > https://samples.ffmpeg.org/A-codecs/uLaw/ > https://samples.ffmpeg.org/mobileVideo_3gp/ > https://samples.ffmpeg.org/archive/container/mov/ > https://samples.ffmpeg.org/V-codecs/SVQ1/ > https://samples.ffmpeg.org/V-codecs/SVQ3/ > https://samples.ffmpeg.org/V-codecs/QT-qdrw/Airplane.mov > https://samples.ffmpeg.org/V-codecs/QTRLE/ > https://samples.ffmpeg.org/V-codecs/ZyGo/ > https://samples.ffmpeg.org/V-codecs/icod/ > https://samples.ffmpeg.org/V-codecs/pxlt-ApplePixlet/ > https://samples.ffmpeg.org/MPEG-4/ > https://samples.ffmpeg.org/ffmpeg-bugs/trac/ticket4471/spot-day.mj2 > > probably a good idea to test the rest of the mov formats like > psp,m4b,ism,ismv,isma,f4v,avif,heic,heif which i didnt link. > > especially popular camera photo format heic/heif could do with some > tests. > > full list of samples https://samples.ffmpeg.org/allsamples.txt > > note that mp4 files are the 2nd most popular format that i see on a > daily basis , just behind mkv. with support by default in > windows, mac/ios, android and web browsers, mov/mp4 is probably the #1 > popular format in the world currently. testing and speeding up our > mov/mp4 demuxer is a good idea. > > 392092 mov.c > ugh. > > thanks, > -compn > _______________________________________________ > ffmpeg-devel mailing list > ffmpeg-devel@ffmpeg.org > https://ffmpeg.org/mailman/listinfo/ffmpeg-devel > > To unsubscribe, visit link above, or email > ffmpeg-devel-requ...@ffmpeg.org with subject "unsubscribe". _______________________________________________ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org https://ffmpeg.org/mailman/listinfo/ffmpeg-devel To unsubscribe, visit link above, or email ffmpeg-devel-requ...@ffmpeg.org with subject "unsubscribe".
