[Bug 728438] v4l2: Implement a v4l2 video encoder

[GStreamer (GNOME Bugzilla)]

https://bugzilla.gnome.org/show_bug.cgi?id=728438

Jean-Christophe Trotin <jean-christophe.trotin at st.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |jean-christophe.trotin at st.c
                   |                            |om

--- Comment #58 from Jean-Christophe Trotin <jean-christophe.trotin at st.com> ---
(In reply to Fnoop from comment #55)
> Thanks for everyone's work on this.  I've applied some patches (ayaka, I
> think?) to current master and 1.10.2 here (1.10.2-patched and master
> branches):
> https://github.com/fnoop/gst-plugins-good
> 
> They work perfectly for me on odroid xu4 to use the MFC for encoding.  It
> would be great to get this working in the main tree somehow.  Even if it's
> not perfect, it will at least do something rather than the current nothing,
> and give something to work on and improve in the future.

Thank you for your work on this.
I’ve applied your patches (1.10.2-patched branch of
https://github.com/fnoop/gst-plugins-good) on my environment (GStreamer 1.8.0)
to test them with STMicroelectronics HVA (Hardware Video Accelerator) V4L2
driver (H.264 encoding).
The patches applied easily, and the resulting V4L2 video encoder plugin works
correctly on HVA V4L2 driver, even if there are some troubles with some
configurations (e.g. management of the alignment between the decoder and the
encoder, last frame not encoded). Note that these troubles might be due to the
fact that I used an “old” version of GStreamer.
The control parameters for the encoder driver are correctly set through the
"extra-controls" property of the GStreamer encoder plugin.
Example: gst-launch-1.0 videotestsrc num-buffers=100 !
"video/x-raw(ANY),format=NV12" ! v4l2video1h264enc
extra-controls="ctrl,h264_profile=4,h264_entropy_mode=1" ! h264parse ! qtmux !
filesink location=./out.mp4

In short, from the tests that I ran on HVA, I agree with you that “it would be
great to get this working in the main tree somehow”. It would give the
opportunity to improve this common plugin.

-- 
You are receiving this mail because:
You are the QA Contact for the bug.