After an evaluation, GNOME has moved from Bugzilla to GitLab. Learn more about GitLab.
No new issues can be reported in GNOME Bugzilla anymore.
To report an issue in a GNOME project, go to GNOME GitLab.
Do not go to GNOME Gitlab for: Bluefish, Doxygen, GnuCash, GStreamer, java-gnome, LDTP, NetworkManager, Tomboy.
Bug 525095 - Non-constant buffer size in raw decoded video frames received from fakesink's handoff signal
Non-constant buffer size in raw decoded video frames received from fakesink's...
Status: RESOLVED INVALID
Product: GStreamer
Classification: Platform
Component: dont know
0.10.18
Other Windows
: Normal normal
: git master
Assigned To: GStreamer Maintainers
GStreamer Maintainers
Depends on:
Blocks:
 
 
Reported: 2008-03-30 10:28 UTC by jones.noamle
Modified: 2008-03-30 11:15 UTC
See Also:
GNOME target: ---
GNOME version: ---


Attachments
GStreamer log (with interleaved application logs, you can ignore) (230.66 KB, application/x-gzip)
2008-03-30 10:33 UTC, jones.noamle
Details

Description jones.noamle 2008-03-30 10:28:32 UTC
I have a callback on the "handoff" signal of fakesink, in a pipeline (see below) that decodes video. 
The problem is that GST_BUFFER_SIZE on the recieved buffer varies instead of having the same constant size as expected by width*height*bpp. The variation is: width*height*bpp + a few dozen bytes, with the "extra" number varying from 0 to a few dozen.

This is a version of gstreamer built using OABuild2 (the build was given to me by the maintainer of http://people.collabora.co.uk/~oleavr/OABuild/)

Using the following pipeline:

udpsrc port=%d caps=\"application/x-rtp, media=(string)video, clock-rate=(int)90000, encoding-name=(string)H263\" ! rtph263depay ! ffdec_h263 ! ffmpegcolorspace ! video/x-raw-rgb,depth=(int)24 !  fakesink name=sink
Comment 1 jones.noamle 2008-03-30 10:33:46 UTC
Created attachment 108263 [details]
GStreamer log (with interleaved application logs, you can ignore)
Comment 2 jones.noamle 2008-03-30 11:15:34 UTC
Added more debug info and now it is obvious that this is not the problem.
Sorry...