/* GStreamer * Copyright (C) <1999> Erik Walthinsen * * This library is free software; you can redistribute it and/or * modify it under the terms of the GNU Library General Public * License as published by the Free Software Foundation; either * version 2 of the License, or (at your option) any later version. * * This library is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU * Library General Public License for more details. * * You should have received a copy of the GNU Library General Public * License along with this library; if not, write to the * Free Software Foundation, Inc., 59 Temple Place - Suite 330, * Boston, MA 02111-1307, USA. */ #include #ifdef HAVE_CONFIG_H #include "config.h" #endif #include #include #include #define GST_CAT_DEFAULT audio_rate_debug GST_DEBUG_CATEGORY_STATIC (audio_rate_debug); #define GST_TYPE_AUDIO_RATE \ (gst_audio_rate_get_type()) #define GST_AUDIO_RATE(obj) \ (G_TYPE_CHECK_INSTANCE_CAST((obj),GST_TYPE_AUDIO_RATE,GstAudioRate)) #define GST_AUDIO_RATE_CLASS(klass) \ (G_TYPE_CHECK_CLASS_CAST((klass),GST_TYPE_AUDIO_RATE,GstAudioRate)) #define GST_IS_AUDIO_RATE(obj) \ (G_TYPE_CHECK_INSTANCE_TYPE((obj),GST_TYPE_AUDIO_RATE)) #define GST_IS_AUDIO_RATE_CLASS(klass) \ (G_TYPE_CHECK_CLASS_TYPE((klass),GST_TYPE_AUDIO_RATE)) typedef struct _GstAudioRate GstAudioRate; typedef struct _GstAudioRateClass GstAudioRateClass; struct _GstAudioRate { GstElement element; GstPad *sinkpad, *srcpad; /* audio format */ gint bytes_per_sample; gint rate; /* stats */ guint64 in, out, add, drop; gboolean silent; /* audio state */ guint64 next_offset; guint64 next_ts; gboolean discont; gboolean new_segment; /* we accept all formats on the sink */ GstSegment sink_segment; /* we output TIME format on the src */ GstSegment src_segment; }; struct _GstAudioRateClass { GstElementClass parent_class; }; /* elementfactory information */ static const GstElementDetails audio_rate_details = GST_ELEMENT_DETAILS ("Audio rate adjuster", "Filter/Effect/Audio", "Drops/duplicates/adjusts timestamps on audio samples to make a perfect stream", "Wim Taymans "); /* GstAudioRate signals and args */ enum { /* FILL ME */ LAST_SIGNAL }; #define DEFAULT_SILENT TRUE enum { ARG_0, ARG_IN, ARG_OUT, ARG_ADD, ARG_DROP, ARG_SILENT, /* FILL ME */ }; static GstStaticPadTemplate gst_audio_rate_src_template = GST_STATIC_PAD_TEMPLATE ("src", GST_PAD_SRC, GST_PAD_ALWAYS, GST_STATIC_CAPS (GST_AUDIO_INT_PAD_TEMPLATE_CAPS ";" GST_AUDIO_FLOAT_PAD_TEMPLATE_CAPS) ); static GstStaticPadTemplate gst_audio_rate_sink_template = GST_STATIC_PAD_TEMPLATE ("sink", GST_PAD_SINK, GST_PAD_ALWAYS, GST_STATIC_CAPS (GST_AUDIO_INT_PAD_TEMPLATE_CAPS ";" GST_AUDIO_FLOAT_PAD_TEMPLATE_CAPS) ); static void gst_audio_rate_base_init (gpointer g_class); static void gst_audio_rate_class_init (GstAudioRateClass * klass); static void gst_audio_rate_init (GstAudioRate * audiorate); static gboolean gst_audio_rate_sink_event (GstPad * pad, GstEvent * event); static gboolean gst_audio_rate_src_event (GstPad * pad, GstEvent * event); static GstFlowReturn gst_audio_rate_chain (GstPad * pad, GstBuffer * buf); static void gst_audio_rate_set_property (GObject * object, guint prop_id, const GValue * value, GParamSpec * pspec); static void gst_audio_rate_get_property (GObject * object, guint prop_id, GValue * value, GParamSpec * pspec); static GstStateChangeReturn gst_audio_rate_change_state (GstElement * element, GstStateChange transition); static GstElementClass *parent_class = NULL; /*static guint gst_audio_rate_signals[LAST_SIGNAL] = { 0 }; */ static GType gst_audio_rate_get_type (void) { static GType audio_rate_type = 0; if (!audio_rate_type) { static const GTypeInfo audio_rate_info = { sizeof (GstAudioRateClass), gst_audio_rate_base_init, NULL, (GClassInitFunc) gst_audio_rate_class_init, NULL, NULL, sizeof (GstAudioRate), 0, (GInstanceInitFunc) gst_audio_rate_init, }; audio_rate_type = g_type_register_static (GST_TYPE_ELEMENT, "GstAudioRate", &audio_rate_info, 0); } return audio_rate_type; } static void gst_audio_rate_base_init (gpointer g_class) { GstElementClass *element_class = GST_ELEMENT_CLASS (g_class); gst_element_class_set_details (element_class, &audio_rate_details); gst_element_class_add_pad_template (element_class, gst_static_pad_template_get (&gst_audio_rate_sink_template)); gst_element_class_add_pad_template (element_class, gst_static_pad_template_get (&gst_audio_rate_src_template)); } static void gst_audio_rate_class_init (GstAudioRateClass * klass) { GObjectClass *object_class = G_OBJECT_CLASS (klass); GstElementClass *element_class = GST_ELEMENT_CLASS (klass); parent_class = g_type_class_peek_parent (klass); object_class->set_property = gst_audio_rate_set_property; object_class->get_property = gst_audio_rate_get_property; g_object_class_install_property (object_class, ARG_IN, g_param_spec_uint64 ("in", "In", "Number of input samples", 0, G_MAXUINT64, 0, G_PARAM_READABLE)); g_object_class_install_property (object_class, ARG_OUT, g_param_spec_uint64 ("out", "Out", "Number of output samples", 0, G_MAXUINT64, 0, G_PARAM_READABLE)); g_object_class_install_property (object_class, ARG_ADD, g_param_spec_uint64 ("add", "Add", "Number of added samples", 0, G_MAXUINT64, 0, G_PARAM_READABLE)); g_object_class_install_property (object_class, ARG_DROP, g_param_spec_uint64 ("drop", "Drop", "Number of dropped samples", 0, G_MAXUINT64, 0, G_PARAM_READABLE)); g_object_class_install_property (object_class, ARG_SILENT, g_param_spec_boolean ("silent", "silent", "Don't emit notify for dropped and duplicated frames", DEFAULT_SILENT, G_PARAM_READWRITE)); element_class->change_state = gst_audio_rate_change_state; } static void gst_audio_rate_reset (GstAudioRate * audiorate) { audiorate->next_offset = -1; audiorate->next_ts = -1; audiorate->discont = TRUE; gst_segment_init (&audiorate->sink_segment, GST_FORMAT_UNDEFINED); gst_segment_init (&audiorate->src_segment, GST_FORMAT_TIME); GST_DEBUG_OBJECT (audiorate, "handle reset"); } static gboolean gst_audio_rate_setcaps (GstPad * pad, GstCaps * caps) { GstAudioRate *audiorate; GstStructure *structure; GstPad *otherpad; gboolean ret = FALSE; gint channels, width, rate; audiorate = GST_AUDIO_RATE (gst_pad_get_parent (pad)); structure = gst_caps_get_structure (caps, 0); if (!gst_structure_get_int (structure, "channels", &channels)) goto wrong_caps; if (!gst_structure_get_int (structure, "width", &width)) goto wrong_caps; if (!gst_structure_get_int (structure, "rate", &rate)) goto wrong_caps; audiorate->bytes_per_sample = channels * (width / 8); if (audiorate->bytes_per_sample == 0) goto wrong_format; audiorate->rate = rate; /* the format is correct, configure caps on other pad */ otherpad = (pad == audiorate->srcpad) ? audiorate->sinkpad : audiorate->srcpad; ret = gst_pad_set_caps (otherpad, caps); done: gst_object_unref (audiorate); return ret; /* ERRORS */ wrong_caps: { GST_DEBUG_OBJECT (audiorate, "could not get channels/width from caps"); goto done; } wrong_format: { GST_DEBUG_OBJECT (audiorate, "bytes_per_samples gave 0"); goto done; } } static void gst_audio_rate_init (GstAudioRate * audiorate) { audiorate->sinkpad = gst_pad_new_from_static_template (&gst_audio_rate_sink_template, "sink"); gst_pad_set_event_function (audiorate->sinkpad, gst_audio_rate_sink_event); gst_pad_set_chain_function (audiorate->sinkpad, gst_audio_rate_chain); gst_pad_set_setcaps_function (audiorate->sinkpad, gst_audio_rate_setcaps); gst_pad_set_getcaps_function (audiorate->sinkpad, gst_pad_proxy_getcaps); gst_element_add_pad (GST_ELEMENT (audiorate), audiorate->sinkpad); audiorate->srcpad = gst_pad_new_from_static_template (&gst_audio_rate_src_template, "src"); gst_pad_set_event_function (audiorate->srcpad, gst_audio_rate_src_event); gst_pad_set_setcaps_function (audiorate->srcpad, gst_audio_rate_setcaps); gst_pad_set_getcaps_function (audiorate->srcpad, gst_pad_proxy_getcaps); gst_element_add_pad (GST_ELEMENT (audiorate), audiorate->srcpad); audiorate->in = 0; audiorate->out = 0; audiorate->drop = 0; audiorate->add = 0; audiorate->silent = DEFAULT_SILENT; } static gboolean gst_audio_rate_sink_event (GstPad * pad, GstEvent * event) { gboolean res; GstAudioRate *audiorate; audiorate = GST_AUDIO_RATE (gst_pad_get_parent (pad)); switch (GST_EVENT_TYPE (event)) { case GST_EVENT_FLUSH_STOP: GST_DEBUG_OBJECT (audiorate, "handling FLUSH_STOP"); gst_audio_rate_reset (audiorate); res = gst_pad_push_event (audiorate->srcpad, event); break; case GST_EVENT_NEWSEGMENT: { GstFormat format; gdouble rate, arate; gint64 start, stop, time; gboolean update; gst_event_parse_new_segment_full (event, &update, &rate, &arate, &format, &start, &stop, &time); GST_DEBUG_OBJECT (audiorate, "handle NEWSEGMENT"); /* FIXME: * - sparse stream support. For this, the update flag is TRUE and the * start/time positions are updated, meaning that time progressed by * time - old_time amount and we need to fill that gap with empty * samples. * - fill the current segment if it has a valid stop position. This * happens when the update flag is FALSE. With the segment helper we can * calculate the accumulated time and compare this to the next_offset. */ if (!update) { /* a new segment starts. We need to figure out what will be the next * sample offset. We mark the offsets as invalid so that the _chain * function will perform this calculation. */ audiorate->next_offset = -1; audiorate->next_ts = -1; } /* we accept all formats */ gst_segment_set_newsegment_full (&audiorate->sink_segment, update, rate, arate, format, start, stop, time); GST_DEBUG_OBJECT (audiorate, "updated segment: %" GST_SEGMENT_FORMAT, &audiorate->sink_segment); if (format == GST_FORMAT_TIME) { /* TIME formats can be copied to src and forwarded */ res = gst_pad_push_event (audiorate->srcpad, event); memcpy (&audiorate->src_segment, &audiorate->sink_segment, sizeof (GstSegment)); } else { /* other formats will be handled in the _chain function */ gst_event_unref (event); res = TRUE; } break; } case GST_EVENT_EOS: /* FIXME, fill last segment */ res = gst_pad_push_event (audiorate->srcpad, event); break; default: res = gst_pad_push_event (audiorate->srcpad, event); break; } gst_object_unref (audiorate); return res; } static gboolean gst_audio_rate_src_event (GstPad * pad, GstEvent * event) { gboolean res; GstAudioRate *audiorate; audiorate = GST_AUDIO_RATE (gst_pad_get_parent (pad)); switch (GST_EVENT_TYPE (event)) { default: res = gst_pad_push_event (audiorate->sinkpad, event); break; } gst_object_unref (audiorate); return res; } static gboolean gst_audio_rate_convert (GstAudioRate * audiorate, GstFormat src_fmt, gint64 src_val, GstFormat dest_fmt, gint64 * dest_val) { if (src_fmt == dest_fmt) { *dest_val = src_val; return TRUE; } switch (src_fmt) { case GST_FORMAT_DEFAULT: switch (dest_fmt) { case GST_FORMAT_BYTES: *dest_val = src_val * audiorate->bytes_per_sample; break; case GST_FORMAT_TIME: *dest_val = gst_util_uint64_scale_int (src_val, GST_SECOND, audiorate->rate); break; default: return FALSE;; } break; case GST_FORMAT_BYTES: switch (dest_fmt) { case GST_FORMAT_DEFAULT: *dest_val = src_val / audiorate->bytes_per_sample; break; case GST_FORMAT_TIME: *dest_val = gst_util_uint64_scale_int (src_val, GST_SECOND, audiorate->rate * audiorate->bytes_per_sample); break; default: return FALSE;; } break; case GST_FORMAT_TIME: switch (dest_fmt) { case GST_FORMAT_BYTES: *dest_val = gst_util_uint64_scale_int (src_val, audiorate->rate * audiorate->bytes_per_sample, GST_SECOND); break; case GST_FORMAT_DEFAULT: *dest_val = gst_util_uint64_scale_int (src_val, audiorate->rate, GST_SECOND); break; default: return FALSE;; } break; default: return FALSE; } return TRUE; } static gboolean gst_audio_rate_convert_segments (GstAudioRate * audiorate) { GstFormat src_fmt, dst_fmt; src_fmt = audiorate->sink_segment.format; dst_fmt = audiorate->src_segment.format; #define CONVERT_VAL(field) gst_audio_rate_convert (audiorate, \ src_fmt, audiorate->sink_segment.field, \ dst_fmt, &audiorate->src_segment.field); audiorate->sink_segment.rate = audiorate->src_segment.rate; audiorate->sink_segment.abs_rate = audiorate->src_segment.abs_rate; audiorate->sink_segment.flags = audiorate->src_segment.flags; audiorate->sink_segment.applied_rate = audiorate->src_segment.applied_rate; CONVERT_VAL (start); CONVERT_VAL (stop); CONVERT_VAL (time); CONVERT_VAL (accum); CONVERT_VAL (last_stop); #undef CONVERT_VAL return TRUE; } static GstFlowReturn gst_audio_rate_chain (GstPad * pad, GstBuffer * buf) { GstAudioRate *audiorate; GstClockTime in_time, in_duration, in_stop, run_time; guint64 in_offset, in_offset_end, in_samples; guint in_size; GstFlowReturn ret = GST_FLOW_OK; audiorate = GST_AUDIO_RATE (gst_pad_get_parent (pad)); /* need to be negotiated now */ if (audiorate->bytes_per_sample == 0) goto not_negotiated; /* we have a new pending segment */ if (audiorate->next_offset == -1) { gint64 pos; /* update the TIME segment */ gst_audio_rate_convert_segments (audiorate); /* first buffer, we are negotiated and we have a segment, calculate the * current expected offsets based on the segment.start, which is the first * media time of the segment and should match the media time of the first * buffer in that segment, which is the offset expressed in DEFAULT units. */ /* convert first timestamp of segment to sample position */ pos = gst_util_uint64_scale_int (audiorate->src_segment.start, audiorate->rate, GST_SECOND); GST_DEBUG_OBJECT (audiorate, "resync to offset %" G_GINT64_FORMAT, pos); audiorate->next_offset = pos; audiorate->next_ts = gst_util_uint64_scale_int (audiorate->next_offset, GST_SECOND, audiorate->rate); } audiorate->in++; in_time = GST_BUFFER_TIMESTAMP (buf); in_size = GST_BUFFER_SIZE (buf); in_samples = in_size / audiorate->bytes_per_sample; /* get duration from the size because we can and it's more accurate */ in_duration = gst_util_uint64_scale_int (in_samples, GST_SECOND, audiorate->rate); in_stop = in_time + in_duration; /* Figure out the total accumulated segment time. */ run_time = in_time + audiorate->src_segment.accum; /* calculate the buffer offset */ in_offset = gst_util_uint64_scale_int (run_time, audiorate->rate, GST_SECOND); in_offset_end = in_offset + in_samples; GST_LOG_OBJECT (audiorate, "in_time:%" GST_TIME_FORMAT ", run_time:%" GST_TIME_FORMAT ", in_duration:%" GST_TIME_FORMAT ", in_size:%u, in_offset:%lld, in_offset_end:%lld" ", ->next_offset:%lld", GST_TIME_ARGS (in_time), GST_TIME_ARGS (run_time), GST_TIME_ARGS (in_duration), in_size, in_offset, in_offset_end, audiorate->next_offset); /* do we need to insert samples */ if (in_offset > audiorate->next_offset) { GstBuffer *fill; gint fillsize; guint64 fillsamples; fillsamples = in_offset - audiorate->next_offset; fillsize = fillsamples * audiorate->bytes_per_sample; fill = gst_buffer_new_and_alloc (fillsize); /* FIXME, 0 might not be the silence byte for the negotiated format. */ memset (GST_BUFFER_DATA (fill), 0, fillsize); GST_DEBUG_OBJECT (audiorate, "inserting %lld samples", fillsamples); GST_BUFFER_OFFSET (fill) = audiorate->next_offset; audiorate->next_offset += fillsamples; GST_BUFFER_OFFSET_END (fill) = audiorate->next_offset; /* Use next timestamp, then calculate following timestamp based on in_offset * to get duration. Neccesary complexity to get 'perfect' streams */ GST_BUFFER_TIMESTAMP (fill) = audiorate->next_ts; audiorate->next_ts = gst_util_uint64_scale_int (in_offset, GST_SECOND, audiorate->rate); GST_BUFFER_DURATION (fill) = audiorate->next_ts - GST_BUFFER_TIMESTAMP (fill); /* we created this buffer to fill a gap */ GST_BUFFER_FLAG_SET (fill, GST_BUFFER_FLAG_GAP); /* set discont if it's pending, this is mostly done for the first buffer and * after a flushing seek */ if (audiorate->discont) { GST_BUFFER_FLAG_SET (fill, GST_BUFFER_FLAG_DISCONT); audiorate->discont = FALSE; } gst_buffer_set_caps (fill, GST_PAD_CAPS (audiorate->srcpad)); ret = gst_pad_push (audiorate->srcpad, fill); if (ret != GST_FLOW_OK) goto beach; audiorate->out++; audiorate->add += fillsamples; if (!audiorate->silent) g_object_notify (G_OBJECT (audiorate), "add"); } else if (in_offset < audiorate->next_offset) { /* need to remove samples */ if (in_offset_end <= audiorate->next_offset) { guint64 drop = in_size / audiorate->bytes_per_sample; audiorate->drop += drop; GST_DEBUG_OBJECT (audiorate, "dropping %lld samples", drop); /* we can drop the buffer completely */ gst_buffer_unref (buf); if (!audiorate->silent) g_object_notify (G_OBJECT (audiorate), "drop"); goto beach; } else { guint64 truncsamples; guint truncsize, leftsize; GstBuffer *trunc; /* truncate buffer */ truncsamples = audiorate->next_offset - in_offset; truncsize = truncsamples * audiorate->bytes_per_sample; leftsize = in_size - truncsize; trunc = gst_buffer_create_sub (buf, truncsize, leftsize); gst_buffer_unref (buf); buf = trunc; gst_buffer_set_caps (buf, GST_PAD_CAPS (audiorate->srcpad)); audiorate->drop += truncsamples; } } /* Now calculate parameters for whichever buffer (either the original * or truncated one) we're pushing. */ GST_BUFFER_OFFSET (buf) = audiorate->next_offset; GST_BUFFER_OFFSET_END (buf) = in_offset_end; GST_BUFFER_TIMESTAMP (buf) = audiorate->next_ts; audiorate->next_ts = gst_util_uint64_scale_int (in_offset_end, GST_SECOND, audiorate->rate); GST_BUFFER_DURATION (buf) = audiorate->next_ts - GST_BUFFER_TIMESTAMP (buf); if (audiorate->discont) { /* we need to output a discont buffer, do so now */ GST_DEBUG_OBJECT (audiorate, "marking DISCONT on output buffer"); buf = gst_buffer_make_metadata_writable (buf); GST_BUFFER_FLAG_SET (buf, GST_BUFFER_FLAG_DISCONT); audiorate->discont = FALSE; } else if (GST_BUFFER_IS_DISCONT (buf)) { /* else we make everything continuous so we can safely remove the DISCONT * flag from the buffer if there was one */ GST_DEBUG_OBJECT (audiorate, "removing DISCONT from buffer"); buf = gst_buffer_make_metadata_writable (buf); GST_BUFFER_FLAG_UNSET (buf, GST_BUFFER_FLAG_DISCONT); } /* set last_stop on segment */ gst_segment_set_last_stop (&audiorate->src_segment, GST_FORMAT_TIME, GST_BUFFER_TIMESTAMP (buf) + GST_BUFFER_DURATION (buf)); ret = gst_pad_push (audiorate->srcpad, buf); audiorate->out++; audiorate->next_offset = in_offset_end; beach: gst_object_unref (audiorate); return ret; /* ERRORS */ not_negotiated: { GST_ELEMENT_ERROR (audiorate, STREAM, FORMAT, (NULL), ("pipeline error, format was not negotiated")); return GST_FLOW_NOT_NEGOTIATED; } } static void gst_audio_rate_set_property (GObject * object, guint prop_id, const GValue * value, GParamSpec * pspec) { GstAudioRate *audiorate = GST_AUDIO_RATE (object); switch (prop_id) { case ARG_SILENT: audiorate->silent = g_value_get_boolean (value); break; default: G_OBJECT_WARN_INVALID_PROPERTY_ID (object, prop_id, pspec); break; } } static void gst_audio_rate_get_property (GObject * object, guint prop_id, GValue * value, GParamSpec * pspec) { GstAudioRate *audiorate = GST_AUDIO_RATE (object); switch (prop_id) { case ARG_IN: g_value_set_uint64 (value, audiorate->in); break; case ARG_OUT: g_value_set_uint64 (value, audiorate->out); break; case ARG_ADD: g_value_set_uint64 (value, audiorate->add); break; case ARG_DROP: g_value_set_uint64 (value, audiorate->drop); break; case ARG_SILENT: g_value_set_boolean (value, audiorate->silent); break; default: G_OBJECT_WARN_INVALID_PROPERTY_ID (object, prop_id, pspec); break; } } static GstStateChangeReturn gst_audio_rate_change_state (GstElement * element, GstStateChange transition) { GstAudioRate *audiorate = GST_AUDIO_RATE (element); switch (transition) { case GST_STATE_CHANGE_PAUSED_TO_READY: break; case GST_STATE_CHANGE_READY_TO_PAUSED: audiorate->in = 0; audiorate->out = 0; audiorate->drop = 0; audiorate->bytes_per_sample = 0; audiorate->add = 0; gst_audio_rate_reset (audiorate); break; default: break; } if (parent_class->change_state) return parent_class->change_state (element, transition); return GST_STATE_CHANGE_SUCCESS; } static gboolean plugin_init (GstPlugin * plugin) { GST_DEBUG_CATEGORY_INIT (audio_rate_debug, "audiorate", 0, "AudioRate stream fixer"); return gst_element_register (plugin, "audiorate", GST_RANK_NONE, GST_TYPE_AUDIO_RATE); } GST_PLUGIN_DEFINE (GST_VERSION_MAJOR, GST_VERSION_MINOR, "audiorate", "Adjusts audio frames", plugin_init, VERSION, GST_LICENSE, GST_PACKAGE_NAME, GST_PACKAGE_ORIGIN)