[Bug 731791] videometa: add GstVideoAffineTransformationMeta

GStreamer (bugzilla.gnome.org) bugzilla at gnome.org
Tue Jun 17 10:21:29 PDT 2014


https://bugzilla.gnome.org/show_bug.cgi?id=731791
  GStreamer | gst-plugins-base | git

--- Comment #3 from Matthieu Bouron <matthieu.bouron at collabora.com> 2014-06-17 17:21:24 UTC ---
(In reply to comment #2)
> Review of attachment 278612 [details]:
> 
> How is it negotiated? Are users of this API expected to handle all possible
> transformations? What's the expected behaviour for transformations that convert
> a rectangular video to a non-rectangular one (or non multiple of 90°
> rotations)?

I haven't though about it yet.

> 
> ::: gst-libs/gst/video/gstvideoaffinetransformationmeta.h
> @@ +46,3 @@
> + * gst_video_affine_transformation_meta_get_matrix.
> + * The user can provide its own get_matrix function, which can be used to
> + * retrieve a transformation matrix from the buffer.
> 
> Define if this is matrix is stored by rows or by columns, and maybe use a
> [4][4] array instead?

The matrix is stored by rows.

> 
> @@ +57,3 @@
> +
> +  gfloat matrix[16];
> +  GstVideoAffineTransformationGetMatrix get_matrix;
> 
> Why the get_matrix function pointer? What is this supposed to do other than
> returning the matrix?

As I said in the bug description, it is used to provide a way for the user to
fetch the matrix from the buffer / buffer memory. I'm thinking about the
android buffers where the matrix has to be retrieved from the surface texture,
after the texture is updated with the updateTexImage. But i'm sure if this
design is right since it will imply that the upload function is called before
the [...]_get_matrix function is. 

The default implementation just returns the stored matrix.

-- 
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