[digiKam-users] Reading/writing Video metadata to Caption/Title

[Gilles Caulier]

Le lun. 11 mars 2019 à 14:16, Andrey Goreev <aegoreev at gmail.com> a écrit :

> Problem with ffmpeg is that it always produces an output file. So, for
> example, from file.mp4 digiKam will have to create file_out.mp4, then
> remove file.mp4 then rename file_out.mp4 to file.mp4. I think Gilles tested
> it and it slowed down the application a lot (too much).
>


For reading, we use the 'information' mode from ffmpeg to extra metadata
only. Something similar to "ffmpeg -i" but working with ffmpeg shared
library.

For writing, metadata can be only update on few video format (typically
mostly MP4), and for a few metadata entry.
I use it at work to tag MP4 files written from a IR video recorder, but
it's really a limited way. ffmpeg infrastructure have not been designed for
this kind of job. And yes, ffmpeg generate an output file by this way.


Maybe one day someone will write a plugin for that. In that case only the
> users who want write to video files will install and activate the plugin. I
> am one of them too. So I might chip in into the development if I understand
> how the new plugin interface works.
>

The plugin interface is not yet able to work with low level metadata
component.

Best

Gilles Caulier
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.kde.org/pipermail/digikam-users/attachments/20190311/b005a0cd/attachment.html>