[Cin] Fuzzy "Align Timecodes" checked into GIT

Phyllis Smith phylsmith2017 at gmail.com
Mon Jul 20 05:55:55 CEST 2020

I ran out of time to document concerning BT 448 description, demo, and
documentation.  But here is a quick demo at:
The most IMPORTANT thing to know right now is that we had no "real" data to
test this but only "manufactured" data.  It would take video from
preferably 3 cameras filming the same scene from different angles but not
started at the same time AND cameras that are smart enough to supply "real"
timecodes in the data.

Here is the unfinished documentation:

*New Feature – Align Timecode*

New feature includes the following relevant parts:

- Align Timecode under the Window pulldown

- Clapperboard icon on the main timeline to set a timecode

- Timecode option added to timebar choices (where you have hh:mm:ss, etc)

- Time Code Start in the Resources window on the bottom of Info for the
media (when set)

- MMB on a track allows for Timecode option to display the timecode for
that particular track

Here are the comments from the code ; )


* 1) if the format context has a timecode

* return fmt_ctx->timecode - 0

* 2) if the layer/channel has a timecode

* return st->timecode - (start_time-nudge)

* 3) find the 1st program with stream, find 1st program video stream,

* if video stream has a timecode, return st->timecode - (start_time-nudge)

* 4) find timecode in any stream, return st->timecode

* 5) read 100 packets, save ofs=pkt.pts*st->time_base - st->nudge:

* decode frame for video stream of 1st program

* if frame->timecode has a timecode, return frame->timecode - ofs

* if side_data has gop timecode, return gop->timecode - ofs

* if side_data has smpte timecode, return smpte->timecode - ofs

* 6) if the filename/url scans *date_time.ext, return date_time

* 7) if stat works on the filename/url, return mtime

* 8) return -1 failure

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cinelerra-gg.org/pipermail/cin/attachments/20200719/c8f05f3c/attachment.htm>

More information about the Cin mailing list