gst/: Segment update fix.

Original commit message from CVS:
* gst-libs/gst/audio/gstbaseaudiosink.c:
(gst_base_audio_sink_class_init), (gst_base_audio_sink_init),
(gst_base_audio_sink_provide_clock),
(gst_base_audio_sink_set_clock), (gst_base_audio_sink_render),
(gst_base_audio_sink_change_state):
* gst/audioresample/gstaudioresample.c:
Segment update fix.
This commit is contained in:
Wim Taymans 2005-11-21 17:14:02 +00:00
parent f405e12b4a
commit 0f2336cff6
3 changed files with 67 additions and 20 deletions

View file

@ -1,3 +1,13 @@
2005-11-21 Wim Taymans <wim@fluendo.com>
* gst-libs/gst/audio/gstbaseaudiosink.c:
(gst_base_audio_sink_class_init), (gst_base_audio_sink_init),
(gst_base_audio_sink_provide_clock),
(gst_base_audio_sink_set_clock), (gst_base_audio_sink_render),
(gst_base_audio_sink_change_state):
* gst/audioresample/gstaudioresample.c:
Segment update fix.
2005-11-21 Andy Wingo <wingo@pobox.com>
* *.h:

View file

@ -65,6 +65,7 @@ static GstStateChangeReturn gst_base_audio_sink_change_state (GstElement *
element, GstStateChange transition);
static GstClock *gst_base_audio_sink_provide_clock (GstElement * elem);
static void gst_base_audio_sink_set_clock (GstElement * elem, GstClock * clock);
static GstClockTime gst_base_audio_sink_get_time (GstClock * clock,
GstBaseAudioSink * sink);
static void gst_base_audio_sink_callback (GstRingBuffer * rbuf, guint8 * data,
@ -118,6 +119,8 @@ gst_base_audio_sink_class_init (GstBaseAudioSinkClass * klass)
GST_DEBUG_FUNCPTR (gst_base_audio_sink_change_state);
gstelement_class->provide_clock =
GST_DEBUG_FUNCPTR (gst_base_audio_sink_provide_clock);
gstelement_class->set_clock =
GST_DEBUG_FUNCPTR (gst_base_audio_sink_set_clock);
gstbasesink_class->event = GST_DEBUG_FUNCPTR (gst_base_audio_sink_event);
gstbasesink_class->preroll = GST_DEBUG_FUNCPTR (gst_base_audio_sink_preroll);
@ -134,10 +137,8 @@ gst_base_audio_sink_init (GstBaseAudioSink * baseaudiosink,
baseaudiosink->buffer_time = DEFAULT_BUFFER_TIME;
baseaudiosink->latency_time = DEFAULT_LATENCY_TIME;
baseaudiosink->clock = gst_audio_clock_new ("clock",
(GstAudioClockGetTimeFunc) gst_base_audio_sink_get_time, baseaudiosink);
}
static void
@ -166,15 +167,27 @@ gst_base_audio_sink_provide_clock (GstElement * elem)
sink = GST_BASE_AUDIO_SINK (elem);
#if 1
clock = GST_CLOCK_CAST (gst_object_ref (sink->clock));
#else
clock = gst_system_clock_obtain ();
#endif
return clock;
}
static void
gst_base_audio_sink_set_clock (GstElement * elem, GstClock * clock)
{
GstBaseAudioSink *sink;
sink = GST_BASE_AUDIO_SINK (elem);
GST_OBJECT_LOCK (sink);
if (clock != sink->clock) {
gst_clock_set_master (sink->clock, clock);
} else {
gst_clock_set_master (sink->clock, NULL);
}
GST_OBJECT_UNLOCK (sink);
}
static GstClockTime
gst_base_audio_sink_get_time (GstClock * clock, GstBaseAudioSink * sink)
{
@ -385,6 +398,8 @@ gst_base_audio_sink_render (GstBaseSink * bsink, GstBuffer * buf)
guint size;
guint samples;
gint bps;
gdouble crate;
GstClockTime cinternal, cexternal;
sink = GST_BASE_AUDIO_SINK (bsink);
@ -408,7 +423,7 @@ gst_base_audio_sink_render (GstBaseSink * bsink, GstBuffer * buf)
data = GST_BUFFER_DATA (buf);
GST_DEBUG ("time %" GST_TIME_FORMAT ", offset %llu, start %" GST_TIME_FORMAT,
GST_TIME_ARGS (time), in_offset, GST_TIME_ARGS (bsink->segment_start));
GST_TIME_ARGS (time), in_offset, GST_TIME_ARGS (bsink->segment.start));
/* if not valid timestamp or we don't need to sync, try to play
* sample ASAP */
@ -417,23 +432,30 @@ gst_base_audio_sink_render (GstBaseSink * bsink, GstBuffer * buf)
goto no_sync;
}
render_diff = time - bsink->segment_start;
render_diff = time - bsink->segment.start;
/* samples should be rendered based on their timestamp. All samples
* arriving before the segment_start are to be thrown away */
* arriving before the segment.start are to be thrown away */
/* FIXME, for now we drop the sample completely, we should
* in fact clip the sample. Same for the segment_stop, actually. */
* in fact clip the sample. Same for the segment.stop, actually. */
if (render_diff < 0)
goto out_of_segment;
gst_clock_get_calibration (sink->clock, &cinternal, &cexternal, &crate);
GST_DEBUG_OBJECT (sink,
"internal %" G_GUINT64_FORMAT ", %" G_GUINT64_FORMAT ", rate %g",
cinternal, cexternal, crate);
/* bring buffer timestamp to stream time */
render_time = render_diff;
/* adjust for rate */
render_time /= ABS (bsink->segment_rate);
render_time /= ABS (bsink->segment.rate);
/* adjust for accumulated segments */
render_time += bsink->segment_accum;
render_time += bsink->segment.accum;
/* add base time to get absolute clock time */
render_time += gst_element_get_base_time (GST_ELEMENT_CAST (bsink));
render_time +=
(gst_element_get_base_time (GST_ELEMENT_CAST (bsink)) - cexternal) +
cinternal;
/* and bring the time to the offset in the buffer */
render_offset = render_time * ringbuf->spec.rate / GST_SECOND;
@ -461,14 +483,14 @@ gst_base_audio_sink_render (GstBaseSink * bsink, GstBuffer * buf)
no_sync:
/* clip length based on rate */
samples = MIN (samples, samples / ABS (bsink->segment_rate));
samples = MIN (samples, samples / (crate * ABS (bsink->segment.rate)));
/* the next sample should be current sample and its length */
sink->next_sample = render_offset + samples;
gst_ring_buffer_commit (ringbuf, render_offset, data, samples);
if (GST_CLOCK_TIME_IS_VALID (time) && time + duration >= bsink->segment_stop) {
if (GST_CLOCK_TIME_IS_VALID (time) && time + duration >= bsink->segment.stop) {
GST_DEBUG ("start playback because we are at the end of segment");
gst_ring_buffer_start (ringbuf);
}
@ -479,7 +501,7 @@ out_of_segment:
{
GST_DEBUG ("dropping sample out of segment time %" GST_TIME_FORMAT
", start %" GST_TIME_FORMAT,
GST_TIME_ARGS (time), GST_TIME_ARGS (bsink->segment_start));
GST_TIME_ARGS (time), GST_TIME_ARGS (bsink->segment.start));
return GST_FLOW_OK;
}
wrong_state:
@ -544,6 +566,23 @@ gst_base_audio_sink_change_state (GstElement * element,
gst_ring_buffer_set_flushing (sink->ringbuffer, FALSE);
break;
case GST_STATE_CHANGE_PAUSED_TO_PLAYING:
{
GstClockTime time;
gdouble rate;
time = gst_clock_get_internal_time (sink->clock);
GST_DEBUG_OBJECT (sink, "time: %" GST_TIME_FORMAT, GST_TIME_ARGS (time));
gst_clock_get_calibration (sink->clock, NULL, NULL, &rate);
/* Does not work yet.
gst_clock_set_calibration (sink->clock,
time, element->base_time, rate);
*/
break;
}
case GST_STATE_CHANGE_PAUSED_TO_READY:
gst_ring_buffer_set_flushing (sink->ringbuffer, TRUE);
break;
default:
break;
@ -556,8 +595,6 @@ gst_base_audio_sink_change_state (GstElement * element,
gst_ring_buffer_pause (sink->ringbuffer);
break;
case GST_STATE_CHANGE_PAUSED_TO_READY:
gst_ring_buffer_set_flushing (sink->ringbuffer, TRUE);
gst_ring_buffer_stop (sink->ringbuffer);
gst_ring_buffer_release (sink->ringbuffer);
gst_pad_set_caps (GST_BASE_SINK_PAD (sink), NULL);
break;

View file

@ -399,7 +399,7 @@ static GstFlowReturn
outsize, outsamples);
GST_BUFFER_OFFSET (outbuf) = audioresample->offset;
GST_BUFFER_TIMESTAMP (outbuf) = base->segment_start +
GST_BUFFER_TIMESTAMP (outbuf) = base->segment.start +
audioresample->offset * GST_SECOND / audioresample->o_rate;
audioresample->offset += outsamples;
@ -408,7 +408,7 @@ static GstFlowReturn
/* we calculate DURATION as the difference between "next" timestamp
* and current timestamp so we ensure a contiguous stream, instead of
* having rounding errors. */
GST_BUFFER_DURATION (outbuf) = base->segment_start +
GST_BUFFER_DURATION (outbuf) = base->segment.start +
audioresample->offset * GST_SECOND / audioresample->o_rate -
GST_BUFFER_TIMESTAMP (outbuf);