[FFmpeg-cvslog] avcodec: add code for a frame queue for use by audio encoders with delay

Justin Ruggles git at videolan.org
Thu Mar 22 01:33:55 CET 2012


ffmpeg | branch: master | Justin Ruggles <justin.ruggles at gmail.com> | Wed Feb 22 21:52:34 2012 -0500| [4bf64961a99f36b72b69e66310fa828525564166] | committer: Justin Ruggles

avcodec: add code for a frame queue for use by audio encoders with delay

This simplifies matching of timestamps between input frames and output
packets.

> http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commit;h=4bf64961a99f36b72b69e66310fa828525564166
---

 libavcodec/audio_frame_queue.c |  162 ++++++++++++++++++++++++++++++++++++++++
 libavcodec/audio_frame_queue.h |   90 ++++++++++++++++++++++
 2 files changed, 252 insertions(+), 0 deletions(-)

diff --git a/libavcodec/audio_frame_queue.c b/libavcodec/audio_frame_queue.c
new file mode 100644
index 0000000..156c3a1
--- /dev/null
+++ b/libavcodec/audio_frame_queue.c
@@ -0,0 +1,162 @@
+/*
+ * Audio Frame Queue
+ * Copyright (c) 2012 Justin Ruggles
+ *
+ * This file is part of Libav.
+ *
+ * Libav is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU Lesser General Public
+ * License as published by the Free Software Foundation; either
+ * version 2.1 of the License, or (at your option) any later version.
+ *
+ * Libav 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
+ * Lesser General Public License for more details.
+ *
+ * You should have received a copy of the GNU Lesser General Public
+ * License along with Libav; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
+ */
+
+#include "libavutil/mathematics.h"
+#include "internal.h"
+#include "audio_frame_queue.h"
+
+void ff_af_queue_init(AVCodecContext *avctx, AudioFrameQueue *afq)
+{
+    afq->avctx             = avctx;
+    afq->next_pts          = AV_NOPTS_VALUE;
+    afq->remaining_delay   = avctx->delay;
+    afq->remaining_samples = avctx->delay;
+    afq->frame_queue       = NULL;
+}
+
+static void delete_next_frame(AudioFrameQueue *afq)
+{
+    AudioFrame *f = afq->frame_queue;
+    if (f) {
+        afq->frame_queue = f->next;
+        f->next = NULL;
+        av_freep(&f);
+    }
+}
+
+void ff_af_queue_close(AudioFrameQueue *afq)
+{
+    /* remove/free any remaining frames */
+    while (afq->frame_queue)
+        delete_next_frame(afq);
+    memset(afq, 0, sizeof(*afq));
+}
+
+int ff_af_queue_add(AudioFrameQueue *afq, const AVFrame *f)
+{
+    AudioFrame *new_frame;
+    AudioFrame *queue_end = afq->frame_queue;
+
+    /* find the end of the queue */
+    while (queue_end && queue_end->next)
+        queue_end = queue_end->next;
+
+    /* allocate new frame queue entry */
+    if (!(new_frame = av_malloc(sizeof(*new_frame))))
+        return AVERROR(ENOMEM);
+
+    /* get frame parameters */
+    new_frame->next = NULL;
+    new_frame->duration = f->nb_samples;
+    if (f->pts != AV_NOPTS_VALUE) {
+        new_frame->pts = av_rescale_q(f->pts,
+                                      afq->avctx->time_base,
+                                      (AVRational){ 1, afq->avctx->sample_rate });
+        afq->next_pts = new_frame->pts + new_frame->duration;
+    } else {
+        new_frame->pts = AV_NOPTS_VALUE;
+        afq->next_pts  = AV_NOPTS_VALUE;
+    }
+
+    /* add new frame to the end of the queue */
+    if (!queue_end)
+        afq->frame_queue = new_frame;
+    else
+        queue_end->next = new_frame;
+
+    /* add frame sample count */
+    afq->remaining_samples += f->nb_samples;
+
+#ifdef DEBUG
+    ff_af_queue_log_state(afq);
+#endif
+
+    return 0;
+}
+
+void ff_af_queue_remove(AudioFrameQueue *afq, int nb_samples, int64_t *pts,
+                        int *duration)
+{
+    int64_t out_pts = AV_NOPTS_VALUE;
+    int removed_samples = 0;
+
+#ifdef DEBUG
+    ff_af_queue_log_state(afq);
+#endif
+
+    /* get output pts from the next frame or generated pts */
+    if (afq->frame_queue) {
+        if (afq->frame_queue->pts != AV_NOPTS_VALUE)
+            out_pts = afq->frame_queue->pts - afq->remaining_delay;
+    } else {
+        if (afq->next_pts != AV_NOPTS_VALUE)
+            out_pts = afq->next_pts - afq->remaining_delay;
+    }
+    if (pts) {
+        if (out_pts != AV_NOPTS_VALUE)
+            *pts = ff_samples_to_time_base(afq->avctx, out_pts);
+        else
+            *pts = AV_NOPTS_VALUE;
+    }
+
+    /* if the delay is larger than the packet duration, we use up delay samples
+       for the output packet and leave all frames in the queue */
+    if (afq->remaining_delay >= nb_samples) {
+        removed_samples      += nb_samples;
+        afq->remaining_delay -= nb_samples;
+    }
+    /* remove frames from the queue until we have enough to cover the
+       requested number of samples or until the queue is empty */
+    while (removed_samples < nb_samples && afq->frame_queue) {
+        removed_samples += afq->frame_queue->duration;
+        delete_next_frame(afq);
+    }
+    afq->remaining_samples -= removed_samples;
+
+    /* if there are no frames left and we have room for more samples, use
+       any remaining delay samples */
+    if (removed_samples < nb_samples && afq->remaining_samples > 0) {
+        int add_samples = FFMIN(afq->remaining_samples,
+                                nb_samples - removed_samples);
+        removed_samples        += add_samples;
+        afq->remaining_samples -= add_samples;
+    }
+    if (removed_samples > nb_samples)
+        av_log(afq->avctx, AV_LOG_WARNING, "frame_size is too large\n");
+    if (duration)
+        *duration = ff_samples_to_time_base(afq->avctx, removed_samples);
+}
+
+void ff_af_queue_log_state(AudioFrameQueue *afq)
+{
+    AudioFrame *f;
+    av_log(afq->avctx, AV_LOG_DEBUG, "remaining delay   = %d\n",
+           afq->remaining_delay);
+    av_log(afq->avctx, AV_LOG_DEBUG, "remaining samples = %d\n",
+           afq->remaining_samples);
+    av_log(afq->avctx, AV_LOG_DEBUG, "frames:\n");
+    f = afq->frame_queue;
+    while (f) {
+        av_log(afq->avctx, AV_LOG_DEBUG, "  [ pts=%9"PRId64" duration=%d ]\n",
+               f->pts, f->duration);
+        f = f->next;
+    }
+}
diff --git a/libavcodec/audio_frame_queue.h b/libavcodec/audio_frame_queue.h
new file mode 100644
index 0000000..cfcc6a0
--- /dev/null
+++ b/libavcodec/audio_frame_queue.h
@@ -0,0 +1,90 @@
+/*
+ * Audio Frame Queue
+ * Copyright (c) 2012 Justin Ruggles
+ *
+ * This file is part of Libav.
+ *
+ * Libav is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU Lesser General Public
+ * License as published by the Free Software Foundation; either
+ * version 2.1 of the License, or (at your option) any later version.
+ *
+ * Libav 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
+ * Lesser General Public License for more details.
+ *
+ * You should have received a copy of the GNU Lesser General Public
+ * License along with Libav; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
+ */
+
+#ifndef AVCODEC_AUDIO_FRAME_QUEUE_H
+#define AVCODEC_AUDIO_FRAME_QUEUE_H
+
+#include "avcodec.h"
+
+typedef struct AudioFrame {
+    int64_t pts;
+    int duration;
+    struct AudioFrame *next;
+} AudioFrame;
+
+typedef struct AudioFrameQueue {
+    AVCodecContext *avctx;
+    int64_t next_pts;
+    int remaining_delay;
+    int remaining_samples;
+    AudioFrame *frame_queue;
+} AudioFrameQueue;
+
+/**
+ * Initialize AudioFrameQueue.
+ *
+ * @param avctx context to use for time_base and av_log
+ * @param afq   queue context
+ */
+void ff_af_queue_init(AVCodecContext *avctx, AudioFrameQueue *afq);
+
+/**
+ * Close AudioFrameQueue.
+ *
+ * Frees memory if needed.
+ *
+ * @param afq queue context
+ */
+void ff_af_queue_close(AudioFrameQueue *afq);
+
+/**
+ * Add a frame to the queue.
+ *
+ * @param afq queue context
+ * @param f   frame to add to the queue
+ */
+int ff_af_queue_add(AudioFrameQueue *afq, const AVFrame *f);
+
+/**
+ * Remove frame(s) from the queue.
+ *
+ * Retrieves the pts of the next available frame, or a generated pts based on
+ * the last frame duration if there are no frames left in the queue. The number
+ * of requested samples should be the full number of samples represented by the
+ * packet that will be output by the encoder. If fewer samples are available
+ * in the queue, a smaller value will be used for the output duration.
+ *
+ * @param afq           queue context
+ * @param nb_samples    number of samples to remove from the queue
+ * @param[out] pts      output packet pts
+ * @param[out] duration output packet duration
+ */
+void ff_af_queue_remove(AudioFrameQueue *afq, int nb_samples, int64_t *pts,
+                        int *duration);
+
+/**
+ * Log the current state of the queue.
+ *
+ * @param afq queue context
+ */
+void ff_af_queue_log_state(AudioFrameQueue *afq);
+
+#endif /* AVCODEC_AUDIO_FRAME_QUEUE_H */



More information about the ffmpeg-cvslog mailing list