[gst-devel] Audio Sinks Clocking Problem?

[Itay K. <ik...@gm...>]

Hi,

I'm trying to do audio streaming using gstreamer, and I can't seem to get it
synchronized properly.
Is there a real problem with using alsasink (or any audio sink for that
matter) with sync set to true? Or am I just not doing something right?

Even in a simple pipeline:
gst-launch audiotestsrc is-live=true ! audioconvert ! "audio/x-raw-int,
width=(int)16, depth=(int)16, signed=(boolean)true, endianness=(int)1234,
channels=(int)1, rate=(int)8000" ! alsasink sync=true

The sound is choppy.

In a more complex pipeline, that does audio RTP streaming, the sound starts
out alright for the first few samples, and then the clock_offset variable:

    gst_clock_get_calibration (sink->provided_clock, &cinternal, &cexternal,
&crate_num, &crate_denom);
    clock_offset = (gst_element_get_base_time (GST_ELEMENT_CAST (bsink)) -
cexternal) + cinternal;

suddenly goes from 0 to a very large value, throwing the whole thing out of
sync.

Any hints to what i'm doing wrong? Or how it make it work ok?

Thanks,
Itay.