mirror of
https://github.com/CCExtractor/ccextractor.git
synced 2026-02-04 05:44:53 +00:00
[QUESTION] Support for extracting arib subs from bin_data pid #648
Reference in New Issue
Block a user
Delete Branch "%!s()"
Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?
Originally created by @jakubvojacek on GitHub (Jul 20, 2021).
CCExtractor version: 0.90
In raising this issue, I confirm the following:
Necessary information
ccextractor output.ts -datapid 0x102Video links
ffmpeg -y -i nsc.ts -map 0:0 -vcodec copy -map 0:1 -acodec copy -map 0:3 -scodec copy -copy_unknown -f mpegts output.tsccextractor output.ts -datapid 0x102found no subs-datastreamtypeand-streamtypebut no luckI am trying to extract arib captions from stream that is transcoded via ffmpeg (abr, resize, ...) and I need to copy the subs as well so that the timestamp of the video would be same as of subtitles. To simplify things, I am testing with simple ffmpeg copy, command above.
I know that ffmpeg does not support copying arib (isdbt) subtitles yet, it can only decode, not encode. But anyway, it tries and copies the data somehow.
ffprobesays that the subs areStream #0:2[0x102]: Data: bin_data ([6][0][0][0] / 0x0006),mediainfocommand correctly recognizesARIB STD B24/B37.I checked the subtitle PID of both the input and output files via
tsdump(part oftsduck) and the packets are pretty much 90% the same. Only the header seems to differ. I suspect that ffmpeg might have managed to copy the data correctly but the headers that ccextractor is using to choose which decoder to use, are missing.I tried forcing ccextractor to use isdb codec (i changed the else to be isdb decoder here https://github.com/CCExtractor/ccextractor/blob/master/src/lib_ccx/general_loop.c#L727) but still no luck extracting the subtitles - my C knowledge is rather bad
I would highly appreciate if anyone could take a look at the output.ts file and tried to extract the subs. Perhaps the data are not there but I think that they might and it could be quite easy to persuade ccextractor to extract them.
Thanks
Jakub