After an evaluation, GNOME has moved from Bugzilla to GitLab. Learn more about GitLab.
No new issues can be reported in GNOME Bugzilla anymore.
To report an issue in a GNOME project, go to GNOME GitLab.
Do not go to GNOME Gitlab for: Bluefish, Doxygen, GnuCash, GStreamer, java-gnome, LDTP, NetworkManager, Tomboy.
Bug 116556 - mpeg2subt/subtitle plugins
mpeg2subt/subtitle plugins
Status: RESOLVED NOTABUG
Product: GStreamer
Classification: Platform
Component: gst-plugins
git master
Other Linux
: Normal enhancement
: NONE
Assigned To: GStreamer Maintainers
GStreamer Maintainers
Depends on:
Blocks:
 
 
Reported: 2003-07-02 18:02 UTC by Ronald Bultje
Modified: 2004-12-22 21:47 UTC
See Also:
GNOME target: ---
GNOME version: ---



Description Ronald Bultje 2003-07-02 18:02:20 UTC
We've currently got three text-on-video overlay plugins: mpeg2subt
(overlays mpeg2 subtitles on a mpeg2 image), textoverlay (in the sandbox)
and timeoverlay (in HEAD). These need merging in some way.

My opinion:
* we should merge the text-on-video function of text-/timeoverlay into one
that can overlay text on both YUV and RGB frames.
* we should move the time-to-text-stream function in timeoverlay into a
separate plugin if we want to.
* text-on-mpeg2 and mpeg2-subtitle-decoding in mpeg2subt should be
separated in two plugins. We might want to move the text-on-mpeg2 into the
"general" textoverlay plugin, but I don't think that's really needed.

Oh, and, of course, the subtitle plugins need to be moved into HEAD, too, then.
Comment 1 David Schleef 2003-07-02 22:16:57 UTC
mpeg2subt doesn't overlay text, it overlays run-length encoded
subtitle streams.

IMO, textoverlay should be moved to ext/pango, and the common bits of
textoverlay and timeoverlay be merged into a base class.  I'm guessing
that the needs for overlaying text will be greater than what can be
accomplished in one element.

The textoverlay base class should derive from Videofilter.  I still
need to figure out how to stuff extra pads onto a subclass of
videofilter, but I think its possible.
Comment 2 Ronald Bultje 2003-07-03 05:58:00 UTC
It's for accessibility and portability. We're currently thinking of
text on a video as the only thing, but how about text output to a
braille reading device as an a11y feature for people that cannot view
movies? Or how about text output to festival as a way of implementing
artificial karaoke? Or on systems that don't have/support pango (KDE!)?

A base class sounds fine, and I'm not saying we should make every
single line of code a separate element - let's just try to keep things
into our general principle of "everything is a pipeline". :).
Comment 3 David Schleef 2003-07-03 07:39:39 UTC
sorry, I wasn't questioning the wisdom of having text streams.  I was
more worried about the (perceived) suggestion that timeoverlay should
be replaced with:

            /- ! timetotext ! -\
  ... ! tee                     textoverlay
            \------- ! --------/

or something similar.

This may be just my personal preference, but I'd like to avoid designs
where one element splits a stream into two streams, just to be merged
in the next element.