[Bug 793705] msdk: Implement Video Post Processing element
GStreamer (GNOME Bugzilla)
bugzilla at gnome.org
Fri Mar 30 22:12:06 UTC 2018
https://bugzilla.gnome.org/show_bug.cgi?id=793705
--- Comment #22 from sreerenj <bsreerenj at gmail.com> ---
Review of attachment 370068:
--> (https://bugzilla.gnome.org/review?bug=793705&attachment=370068)
::: sys/msdk/gstmsdkvpp.c
@@ +58,3 @@
+ "width = (int) [ 1, MAX ], height = (int) [ 1, MAX ],"
+ "interlace-mode = (string) progressive")
+ );
Done.
@@ +69,3 @@
+ "width = (int) [ 1, MAX ], height = (int) [ 1, MAX ],"
+ "interlace-mode = (string) progressive")
+ );
Done!
@@ +83,3 @@
+
+#define gst_msdkvpp_parent_class parent_class
+G_DEFINE_TYPE (GstMsdkVPP, gst_msdkvpp, GST_TYPE_BASE_TRANSFORM);
MSDK has specific modes to double the frame rate in deinterlacing.
There are some misunderstandings in the way we interpret fps in deinterlacing.
MSDK requires the field-rate as input and frame-rate as output.
Suppose h264parse negotiate "framerate=25": here msdk expect input rate 50 and
output rate 25 (weave) or 50(bob).
I will add some comments and fixes to the deinterlacing later on (not included
in this patch series).
I have to double check the parser code, if parser negotiate filed-rate as
framerate, then we are not supposed to double it again :)
@@ +239,3 @@
+ goto error_pool_config;
+
+ *aligned_info = info;
Done.
@@ +315,3 @@
+ pool =
+ gst_msdkvpp_create_buffer_pool (thiz, GST_PAD_SRC, caps, min_buffers);
+ thiz->srcpad_buffer_pool = pool;
I don't understand, reuse downstream pool should work as it is (update_pool =
true)
Do you mean to update the size/min+max_buffersize? If so, yes that part can be
added. If there is a downstream msdk element (then only the pool will be
reused)
we should already have the correct size but min_max_count could be updated.
@@ +339,3 @@
+ if (caps)
+ gst_caps_unref (caps);
+
Done
@@ +795,3 @@
+static GstCaps *
+gst_msdkvpp_fixate_caps (GstBaseTransform * trans,
+ GstPadDirection direction, GstCaps * caps, GstCaps * othercaps)
Done
--
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