Just to summarize for posterity, here is what works for me: I load the files from my 7D into Cinelerra patched to read SOWT audio. I edit in Cinelerra. I render audio to a wav file 16 bit linear audio. I render video to uncompressed RGB in a quicktime container. I user qtstreamize to move the moov atom to the front of the video file. I use avidemux to mux and encode the video and audio into an h.264 video / pcm or ac3 audio mkv file. I have to use the swap U and V filter to get the colors right. I use a 2 pass 25 megabit / sec h.264 encoding.
The quality is very good and the sound sync is fine. Rendering video and audio at the same time resulted in audio stopping after about 20 minutes. Rendering to an AVI resulted in crashing Cinelerra. h.264 encoding did not work. Avidemux cannot read the video file if it is longer than 4 Gig until the moov atom is moved to the beginning of the file. I investigated this and it looks like avidemux 2.5.3 ADM_mp4Leaf.cpp and ADM_mp4Analyzer.cpp use uint32_t various places, resulting in a 4 Gig limit for finding the moov atom. I tried patching to use uint64_t, but was unsuccessful in getting it to compile. Anything other than Uncompressed RGB in a quicktime container either resulted in a crash in Cinelerra or some video anomaly in avidemux decoding. I did not use a pipe from yuv4mpeg because it may be lower quality, only being yuv420, and it does not really improve the workflow as I have multiple steps after that. Mark _______________________________________________ Cinelerra mailing list [email protected] https://init.linpro.no/mailman/skolelinux.no/listinfo/cinelerra
