[Bug 607342] patch that Add 2 news plugins jp462bayer and bayer2rgb2

GStreamer (bugzilla.gnome.org) bugzilla at gnome.org
Tue May 21 06:16:46 PDT 2013


https://bugzilla.gnome.org/show_bug.cgi?id=607342
  GStreamer | gst-plugins-bad | git

--- Comment #33 from Sebastian Dröge <slomo at circular-chaos.org> 2013-05-21 13:16:38 UTC ---
(In reply to comment #32)
> (In reply to comment #31)
> > Still the question about why having this as a separate element instead of
> > merging it into the current element and just adding a method property to select
> > the algorithm.
> 
> I suggest you take a closer look at the patch; the confusion is probably a
> result of bad file naming. gstdc1394bayer.c just contains a slightly modified
> version of bayer.c from the libdc1394 project, it does not contain another
> element. Perhaps the gst prefix should be removed to avoid this confusion,
> especially since it doesn't depend on any GStreamer library.

Ah sorry, nevermind then :)

> > As the stride with bayer seems to be undefined in general, make a good choice
> > for a default that is used often... and then add a GstMeta similar to
> > GstVideoMeta to define additional information about the format... like the
> > stride.
> 
> In my experience with framegrabbers, they don't bother to align strides at all,
> however I'll actually most often capture the video as GRAY16 and then use
> capssetter, so perhaps we should stick with the 4 byte alignment (this is
> actually not a problem with 16-bit bayer, assuming there's never a
> half-pattern).

So most hardware is generating bayer with no aligned strides? In that case we
should do that by default too I guess, not useful to work against what most
hardware already does.

-- 
Configure bugmail: https://bugzilla.gnome.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the QA contact for the bug.
You are the assignee for the bug.


More information about the gstreamer-bugs mailing list