[gstreamer-bugs] [Bug 640610] New: basesink: QoS events are wrong in live pipelines
GStreamer (bugzilla.gnome.org)
bugzilla at gnome.org
Tue Jan 25 18:16:00 PST 2011
https://bugzilla.gnome.org/show_bug.cgi?id=640610
GStreamer | gstreamer (core) | git
Summary: basesink: QoS events are wrong in live pipelines
Classification: Desktop
Product: GStreamer
Version: git
OS/Version: Linux
Status: UNCONFIRMED
Severity: normal
Priority: Normal
Component: gstreamer (core)
AssignedTo: gstreamer-bugs at lists.sourceforge.net
ReportedBy: olivier.crete at ocrete.ca
QAContact: gstreamer-bugs at lists.sourceforge.net
CC: felipe.contreras at nokia.com
GNOME target: ---
GNOME version: ---
We have two problem cases:
1. "videotestsrc is-live=1 ! xvimagesink"
If one looks at the QoS events this generates, "diff" is always > 0 .. So it
thinks all frames are late. Also, "proportion" is always just slightly about
1.. But in fact, my computer is very fast and could probably do much more
processing. So "proportion" should be close to 0.
2. "videotestsrc is-live=1 ! element-that-drops-framerate-and-duration !
xvimagesink"
That magical element could be a network using RTP. Since the sink doesn't know
the framerate or duration, the "stop" time is always NONE. So the only thing
that prevents frames from being dropped is the "max-lateness", which is often
not enough if there is enough processing hapenning. Also, in the QoS event,
the "proportion" computation is completely off. Especially if it knows the
duration of some frames but not others.
I think the cause of these two problems is that we ignore the compute latency.
I think there should be a global variable somewhere that is added to the
latency of every pipeline if it is live instead of ignoring the problem like we
currently do.
--
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