somebody help me fix lame ;)

Original commit message from CVS:
somebody help me fix lame ;)

I commented out the state change function because it is called before lame has the right caps.
Is the state change function still necessary ?
in any case, at least now lame actually listens to osssrc re: rate and channels
This commit is contained in:
Thomas Vander Stichele 2002-02-19 12:55:16 +00:00
parent ec94be0220
commit 9ee4ef8061

View file

@ -320,7 +320,7 @@ gst_lame_class_init (GstLameClass *klass)
gobject_class->set_property = gst_lame_set_property;
gobject_class->get_property = gst_lame_get_property;
gstelement_class->change_state = gst_lame_change_state;
//gstelement_class->change_state = gst_lame_change_state;
}
static GstPadConnectReturn
@ -340,8 +340,9 @@ gst_lame_sinkconnect (GstPad *pad, GstCaps *caps)
gst_event_new_info ("channels", GST_PROPS_INT (lame->num_channels), NULL));
gst_element_send_event (GST_ELEMENT (lame),
gst_event_new_info ("rate", GST_PROPS_INT (lame->samplerate), NULL));
GST_DEBUG (0, "rate=%d, channels=%d\n", lame->samplerate, lame->num_channels);
g_print ("rate=%d, channels=%d\n", lame->samplerate, lame->num_channels);
if (gst_lame_setup (lame)) {
lame->initialized = TRUE;
@ -710,8 +711,11 @@ gst_lame_setup (GstLame *lame)
{
GST_DEBUG_ENTER ("(\"%s\")", gst_element_get_name (GST_ELEMENT (lame)));
g_assert (!lame->initialized);
/* copy the parameters over */
lame_set_in_samplerate (lame->lgf, lame->samplerate);
/* force mono encoding if we only have one channel */
if (lame->num_channels == 1)
lame->mode = 3;