gst/extend/discoverer.py: Make the queue buffer up 1s of data before outputting to the sinks. This should give time f...

Original commit message from CVS:
* gst/extend/discoverer.py: Make the queue buffer up 1s of data before
outputting to the sinks. This should give time for some demuxers like
mpegdemux or fluasfdemux to discover a bit more about the muxed
stream and add the correct pads.
Fixes #371969
This commit is contained in:
Edward Hervey 2006-11-07 12:24:13 +00:00
parent 0650a27089
commit f32834da5f
2 changed files with 16 additions and 0 deletions

View file

@ -1,3 +1,11 @@
2006-11-07 Edward Hervey <edward@fluendo.com>
* gst/extend/discoverer.py: Make the queue buffer up 1s of data before
outputting to the sinks. This should give time for some demuxers like
mpegdemux or fluasfdemux to discover a bit more about the muxed
stream and add the correct pads.
Fixes #371969
2006-11-07 Andy Wingo <wingo@pobox.com>
* examples/Makefile.am (examples_DATA):

View file

@ -292,6 +292,14 @@ class Discoverer(gst.Pipeline):
pad.info("adding queue->fakesink")
fakesink = gst.element_factory_make("fakesink")
queue = gst.element_factory_make("queue")
# we want the queue to buffer up to 2 seconds of data before outputting
# This enables us to cope with formats that don't create their source
# pads straight away, but instead wait for the first buffer of that
# stream.
queue.props.min_threshold_time = 1 * gst.SECOND
queue.props.max_size_time = 2 * gst.SECOND
queue.props.max_size_buffers = 0
queue.props.max_size_bytes = 0
self.add(fakesink, queue)
queue.link(fakesink)
sinkpad = fakesink.get_pad("sink")