Closed bturkus closed 9 months ago
Hi Ben, the -f
flag says in the documentation
forces dvpackager to ignore changes in significant technical characteristics of the dv stream when splitting the output
so it's essentially ignoring all the prep work dvpackager usually does and mimicing an ffmpeg -i hey.dv -c copy -map 0 out.mkv
, so if the dv changes specs in the middle then the output will present it all mangled.
I've been considering removing this option, as unless the dv is perfectly consistent, frame to frame, then the output would be mangled. I had this option to help with debugging in dvpackager development, but don't need it for that anymore.
Heya @bturkus, I removed the -f
option in https://github.com/mipops/dvrescue/pull/788/commits/fda477874408e2af4de04db487a7ae342eabf76f
Hi,
Using the most recent proper release of dvrescue, and also the latest macOS daily build that I could find, I've been having some issues packaging DV files with unusual audio configurations (these files in particular switch from 2 CH/48 to 4 CH/32 and then back to 2CH/48; essentially opening and closing title/credit sequences that surround program content).
What I'm finding is that the files package properly when using dvpackager without the -f flag (the resulting files represent all three sections combined), but when invoking -f, the only resulting files are the 2CH/48 opening credit sequences.
Attaching the dvrescue xml for reference, and including the verbose terminal output, but I was curious to know:
I also tried using the various Audio Resample options, but that didn't seem to help.
Thanks!
Ben
myd_723175_v01_pm.mkv.dvrescue.xml.zip