AVFilterBufferRef *avfilter_get_audio_buffer_ref_from_frame(const AVFrame *frame,
int perms)
{
AVFilterBufferRef *samplesref;
int channels = av_frame_get_channels(frame);
int64_t layout = av_frame_get_channel_layout(frame);
if(av_frame_get_channels(frame) > 8) // libavfilter does not suport more than 8 channels FIXME, remove once libavfilter is fixed
return NULL;
if (layout && av_get_channel_layout_nb_channels(layout) != av_frame_get_channels(frame)) {
av_log(0, AV_LOG_ERROR, "Layout indicates a different number of channels than actually present\n");
return NULL;
}
samplesref = avfilter_get_audio_buffer_ref_from_arrays_channels(
(uint8_t **)frame->data, frame->linesize[0], perms,
frame->nb_samples, frame->format, channels, layout);
if (!samplesref)
return NULL;
if (avfilter_copy_frame_props(samplesref, frame) < 0) {
samplesref->buf->data[0] = NULL;
avfilter_unref_bufferp(&samplesref);
}
return samplesref;
}
static bool audio_frame(struct ff_frame *frame, void *opaque)
{
struct ffmpeg_source *s = opaque;
struct obs_source_audio audio_data = {0};
uint64_t pts;
// Media ended
if (frame == NULL)
return true;
pts = (uint64_t)(frame->pts * 1000000000.0L);
int channels = av_frame_get_channels(frame->frame);
for(int i = 0; i < channels; i++)
audio_data.data[i] = frame->frame->data[i];
audio_data.samples_per_sec = frame->frame->sample_rate;
audio_data.frames = frame->frame->nb_samples;
audio_data.timestamp = pts;
audio_data.format =
convert_ffmpeg_sample_format(frame->frame->format);
audio_data.speakers = channels;
obs_source_output_audio(s->source, &audio_data);
return true;
}
int ff_filter_frame(AVFilterLink *link, AVFrame *frame)
{
FF_TPRINTF_START(NULL, filter_frame); ff_tlog_link(NULL, link, 1); ff_tlog(NULL, " "); ff_tlog_ref(NULL, frame, 1);
/* Consistency checks */
if (link->type == AVMEDIA_TYPE_VIDEO) {
if (strcmp(link->dst->filter->name, "scale") &&
strcmp(link->dst->filter->name, "idet")) {
av_assert1(frame->format == link->format);
av_assert1(frame->width == link->w);
av_assert1(frame->height == link->h);
}
} else {
av_assert1(frame->format == link->format);
av_assert1(av_frame_get_channels(frame) == link->channels);
av_assert1(frame->channel_layout == link->channel_layout);
av_assert1(frame->sample_rate == link->sample_rate);
}
/* Go directly to actual filtering if possible */
if (link->type == AVMEDIA_TYPE_AUDIO &&
link->min_samples &&
(link->partial_buf ||
frame->nb_samples < link->min_samples ||
frame->nb_samples > link->max_samples)) {
return ff_filter_frame_needs_framing(link, frame);
} else {
return ff_filter_frame_framed(link, frame);
}
}
string AudioSamples::channelsLayoutString() const
{
if (!m_raw)
return "";
char buf[128] = {0};
av_get_channel_layout_string(buf,
sizeof(buf),
av_frame_get_channels(m_raw),
av_frame_get_channel_layout(m_raw));
return string(buf);
}
const CGEAudioFrameBufferData* CGEVideoDecodeHandler::getCurrentAudioFrame()
{
m_cachedAudioFrame.timestamp = av_frame_get_best_effort_timestamp(m_context->pAudioFrame);
m_cachedAudioFrame.data = m_context->pAudioFrame->data[0];
m_cachedAudioFrame.nbSamples = m_context->pAudioFrame->nb_samples;
m_cachedAudioFrame.bytesPerSample = av_get_bytes_per_sample((AVSampleFormat)m_context->pAudioFrame->format);
m_cachedAudioFrame.channels = av_frame_get_channels(m_context->pAudioFrame);
m_cachedAudioFrame.linesize = m_context->pAudioFrame->linesize[0];
m_cachedAudioFrame.format = (CGESampleFormat)m_context->pAudioFrame->format;
return &m_cachedAudioFrame;
}
int ff_filter_frame(AVFilterLink *link, AVFrame *frame)
{
int ret;
FF_TPRINTF_START(NULL, filter_frame); ff_tlog_link(NULL, link, 1); ff_tlog(NULL, " "); ff_tlog_ref(NULL, frame, 1);
/* Consistency checks */
if (link->type == AVMEDIA_TYPE_VIDEO) {
if (strcmp(link->dst->filter->name, "buffersink") &&
strcmp(link->dst->filter->name, "format") &&
strcmp(link->dst->filter->name, "idet") &&
strcmp(link->dst->filter->name, "null") &&
strcmp(link->dst->filter->name, "scale")) {
av_assert1(frame->format == link->format);
av_assert1(frame->width == link->w);
av_assert1(frame->height == link->h);
}
} else {
if (frame->format != link->format) {
av_log(link->dst, AV_LOG_ERROR, "Format change is not supported\n");
goto error;
}
if (av_frame_get_channels(frame) != link->channels) {
av_log(link->dst, AV_LOG_ERROR, "Channel count change is not supported\n");
goto error;
}
if (frame->channel_layout != link->channel_layout) {
av_log(link->dst, AV_LOG_ERROR, "Channel layout change is not supported\n");
goto error;
}
if (frame->sample_rate != link->sample_rate) {
av_log(link->dst, AV_LOG_ERROR, "Sample rate change is not supported\n");
goto error;
}
}
link->frame_blocked_in = link->frame_wanted_out = 0;
link->frame_count_in++;
filter_unblock(link->dst);
ret = ff_framequeue_add(&link->fifo, frame);
if (ret < 0) {
av_frame_free(&frame);
return ret;
}
ff_filter_set_ready(link->dst, 300);
return 0;
error:
av_frame_free(&frame);
return AVERROR_PATCHWELCOME;
}
void ffsox_convert_setup(convert_t *convert, frame_t *fr, frame_t *fw,
double q, intercept_t *intercept)
{
int nb_samples1,nb_samples2;
convert->fr=fr;
convert->fw=fw;
convert->q=q;
convert->intercept=intercept;
convert->channels=av_frame_get_channels(fr->frame);
nb_samples1=fr->frame->nb_samples-fr->nb_samples.frame;
nb_samples2=fw->frame->nb_samples-fw->nb_samples.frame;
convert->nb_samples=PBU_MIN(nb_samples1,nb_samples2);
}
int ff_filter_frame(AVFilterLink *link, AVFrame *frame)
{
FF_TPRINTF_START(NULL, filter_frame); ff_tlog_link(NULL, link, 1); ff_tlog(NULL, " "); ff_tlog_ref(NULL, frame, 1);
/* Consistency checks */
if (link->type == AVMEDIA_TYPE_VIDEO) {
if (strcmp(link->dst->filter->name, "buffersink") &&
strcmp(link->dst->filter->name, "format") &&
strcmp(link->dst->filter->name, "idet") &&
strcmp(link->dst->filter->name, "null") &&
strcmp(link->dst->filter->name, "scale")) {
av_assert1(frame->format == link->format);
av_assert1(frame->width == link->w);
av_assert1(frame->height == link->h);
}
} else {
if (frame->format != link->format) {
av_log(link->dst, AV_LOG_ERROR, "Format change is not supported\n");
goto error;
}
if (av_frame_get_channels(frame) != link->channels) {
av_log(link->dst, AV_LOG_ERROR, "Channel count change is not supported\n");
goto error;
}
if (frame->channel_layout != link->channel_layout) {
av_log(link->dst, AV_LOG_ERROR, "Channel layout change is not supported\n");
goto error;
}
if (frame->sample_rate != link->sample_rate) {
av_log(link->dst, AV_LOG_ERROR, "Sample rate change is not supported\n");
goto error;
}
}
/* Go directly to actual filtering if possible */
if (link->type == AVMEDIA_TYPE_AUDIO &&
link->min_samples &&
(link->partial_buf ||
frame->nb_samples < link->min_samples ||
frame->nb_samples > link->max_samples)) {
return ff_filter_frame_needs_framing(link, frame);
} else {
return ff_filter_frame_framed(link, frame);
}
error:
av_frame_free(&frame);
return AVERROR_PATCHWELCOME;
}
static int filter_frame(AVFilterLink *inlink, AVFrame *buf)
{
AVFilterContext *ctx = inlink->dst;
AShowInfoContext *s = ctx->priv;
char chlayout_str[128];
uint32_t checksum = 0;
int channels = av_get_channel_layout_nb_channels(buf->channel_layout);
int planar = av_sample_fmt_is_planar(buf->format);
int block_align = av_get_bytes_per_sample(buf->format) * (planar ? 1 : channels);
int data_size = buf->nb_samples * block_align;
int planes = planar ? channels : 1;
int i;
void *tmp_ptr = av_realloc(s->plane_checksums, channels * sizeof(*s->plane_checksums));
if (!tmp_ptr)
return AVERROR(ENOMEM);
s->plane_checksums = tmp_ptr;
for (i = 0; i < planes; i++) {
uint8_t *data = buf->extended_data[i];
s->plane_checksums[i] = av_adler32_update(0, data, data_size);
checksum = i ? av_adler32_update(checksum, data, data_size) :
s->plane_checksums[0];
}
av_get_channel_layout_string(chlayout_str, sizeof(chlayout_str), -1,
buf->channel_layout);
av_log(ctx, AV_LOG_INFO,
"n:%"PRId64" pts:%s pts_time:%s pos:%"PRId64" "
"fmt:%s channels:%d chlayout:%s rate:%d nb_samples:%d "
"checksum:%08X ",
inlink->frame_count,
av_ts2str(buf->pts), av_ts2timestr(buf->pts, &inlink->time_base),
av_frame_get_pkt_pos(buf),
av_get_sample_fmt_name(buf->format), av_frame_get_channels(buf), chlayout_str,
buf->sample_rate, buf->nb_samples,
checksum);
av_log(ctx, AV_LOG_INFO, "plane_checksums: [ ");
for (i = 0; i < planes; i++)
av_log(ctx, AV_LOG_INFO, "%08X ", s->plane_checksums[i]);
av_log(ctx, AV_LOG_INFO, "]\n");
return ff_filter_frame(inlink->dst->outputs[0], buf);
}
static void audio_frame_cksum(AVBPrint *bp, AVFrame *frame)
{
int nb_planes, nb_samples, p;
const char *name;
nb_planes = av_frame_get_channels(frame);
nb_samples = frame->nb_samples;
if (!av_sample_fmt_is_planar(frame->format)) {
nb_samples *= nb_planes;
nb_planes = 1;
}
name = av_get_sample_fmt_name(frame->format);
av_bprintf(bp, ", %d samples", frame->nb_samples);
av_bprintf(bp, ", %s", name ? name : "unknown");
for (p = 0; p < nb_planes; p++) {
uint32_t cksum = 0;
void *d = frame->extended_data[p];
switch (frame->format) {
case AV_SAMPLE_FMT_U8:
case AV_SAMPLE_FMT_U8P:
cksum_line_u8(&cksum, d, nb_samples);
break;
case AV_SAMPLE_FMT_S16:
case AV_SAMPLE_FMT_S16P:
cksum_line_s16(&cksum, d, nb_samples);
break;
case AV_SAMPLE_FMT_S32:
case AV_SAMPLE_FMT_S32P:
cksum_line_s32(&cksum, d, nb_samples);
break;
case AV_SAMPLE_FMT_FLT:
case AV_SAMPLE_FMT_FLTP:
cksum_line_flt(&cksum, d, nb_samples);
break;
case AV_SAMPLE_FMT_DBL:
case AV_SAMPLE_FMT_DBLP:
cksum_line_dbl(&cksum, d, nb_samples);
break;
default:
av_assert0(!"reached");
}
av_bprintf(bp, ", 0x%08x", cksum);
}
}
static int ff_filter_frame_needs_framing(AVFilterLink *link, AVFrame *frame)
{
int insamples = frame->nb_samples, inpos = 0, nb_samples;
AVFrame *pbuf = link->partial_buf;
int nb_channels = av_frame_get_channels(frame);
int ret = 0;
link->flags |= FF_LINK_FLAG_REQUEST_LOOP;
/* Handle framing (min_samples, max_samples) */
while (insamples) {
if (!pbuf) {
AVRational samples_tb = { 1, link->sample_rate };
pbuf = ff_get_audio_buffer(link, link->partial_buf_size);
if (!pbuf) {
av_log(link->dst, AV_LOG_WARNING,
"Samples dropped due to memory allocation failure.\n");
return 0;
}
av_frame_copy_props(pbuf, frame);
pbuf->pts = frame->pts;
if (pbuf->pts != AV_NOPTS_VALUE)
pbuf->pts += av_rescale_q(inpos, samples_tb, link->time_base);
pbuf->nb_samples = 0;
}
nb_samples = FFMIN(insamples,
link->partial_buf_size - pbuf->nb_samples);
av_samples_copy(pbuf->extended_data, frame->extended_data,
pbuf->nb_samples, inpos,
nb_samples, nb_channels, link->format);
inpos += nb_samples;
insamples -= nb_samples;
pbuf->nb_samples += nb_samples;
if (pbuf->nb_samples >= link->min_samples) {
ret = ff_filter_frame_framed(link, pbuf);
pbuf = NULL;
}
}
av_frame_free(&frame);
link->partial_buf = pbuf;
return ret;
}
static int filter_frame(AVFilterLink *inlink, AVFrame *inbuf)
{
AudioPhaserContext *p = inlink->dst->priv;
AVFilterLink *outlink = inlink->dst->outputs[0];
AVFrame *outbuf;
if (av_frame_is_writable(inbuf)) {
outbuf = inbuf;
} else {
outbuf = ff_get_audio_buffer(inlink, inbuf->nb_samples);
if (!outbuf)
return AVERROR(ENOMEM);
av_frame_copy_props(outbuf, inbuf);
}
p->phaser(p, inbuf->extended_data, outbuf->extended_data,
outbuf->nb_samples, av_frame_get_channels(outbuf));
if (inbuf != outbuf)
av_frame_free(&inbuf);
return ff_filter_frame(outlink, outbuf);
}
int attribute_align_arg av_buffersrc_add_frame_flags(AVFilterContext *ctx, AVFrame *frame, int flags)
{
AVFrame *copy = NULL;
int ret = 0;
if (frame && frame->channel_layout &&
av_get_channel_layout_nb_channels(frame->channel_layout) != av_frame_get_channels(frame)) {
av_log(ctx, AV_LOG_ERROR, "Layout indicates a different number of channels than actually present\n");
return AVERROR(EINVAL);
}
if (!(flags & AV_BUFFERSRC_FLAG_KEEP_REF) || !frame)
return av_buffersrc_add_frame_internal(ctx, frame, flags);
if (!(copy = av_frame_alloc()))
return AVERROR(ENOMEM);
ret = av_frame_ref(copy, frame);
if (ret >= 0)
ret = av_buffersrc_add_frame_internal(ctx, copy, flags);
av_frame_free(©);
return ret;
}
static int av_buffersrc_add_frame_internal(AVFilterContext *ctx,
AVFrame *frame, int flags)
{
BufferSourceContext *s = ctx->priv;
AVFrame *copy;
int refcounted, ret;
s->nb_failed_requests = 0;
if (!frame) {
s->eof = 1;
return 0;
} else if (s->eof)
return AVERROR(EINVAL);
refcounted = !!frame->buf[0];
if (!(flags & AV_BUFFERSRC_FLAG_NO_CHECK_FORMAT)) {
switch (ctx->outputs[0]->type) {
case AVMEDIA_TYPE_VIDEO:
CHECK_VIDEO_PARAM_CHANGE(ctx, s, frame->width, frame->height,
frame->format);
break;
case AVMEDIA_TYPE_AUDIO:
/* For layouts unknown on input but known on link after negotiation. */
if (!frame->channel_layout)
frame->channel_layout = s->channel_layout;
CHECK_AUDIO_PARAM_CHANGE(ctx, s, frame->sample_rate, frame->channel_layout,
av_frame_get_channels(frame), frame->format);
break;
default:
return AVERROR(EINVAL);
}
}
if (!av_fifo_space(s->fifo) &&
(ret = av_fifo_realloc2(s->fifo, av_fifo_size(s->fifo) +
sizeof(copy))) < 0)
return ret;
if (!(copy = av_frame_alloc()))
return AVERROR(ENOMEM);
if (refcounted) {
av_frame_move_ref(copy, frame);
} else {
ret = av_frame_ref(copy, frame);
if (ret < 0) {
av_frame_free(©);
return ret;
}
}
if ((ret = av_fifo_generic_write(s->fifo, ©, sizeof(copy), NULL)) < 0) {
if (refcounted)
av_frame_move_ref(frame, copy);
av_frame_free(©);
return ret;
}
if ((flags & AV_BUFFERSRC_FLAG_PUSH))
if ((ret = ctx->output_pads[0].request_frame(ctx->outputs[0])) < 0)
return ret;
return 0;
}
int AudioDecoder::audio_thread(void *arg)
{
VideoState *is = (VideoState *) arg;
AVStreamsParser* ps = is->getAVStreamsParser();
AVFrame *frame = av_frame_alloc();
Frame *af;
#if CONFIG_AVFILTER
int last_serial = -1;
int64_t dec_channel_layout;
int reconfigure;
#endif
int got_frame = 0;
AVRational tb;
int ret = 0;
if (!frame)
return AVERROR(ENOMEM);
do {
if ((got_frame = is->auddec().decode_frame(frame)) < 0)
goto the_end;
if (got_frame) {
tb = (AVRational){1, frame->sample_rate};
#if CONFIG_AVFILTER
dec_channel_layout = get_valid_channel_layout(frame->channel_layout, av_frame_get_channels(frame));
reconfigure =
cmp_audio_fmts(is->audio_filter_src.fmt, is->audio_filter_src.channels,
(AVSampleFormat)frame->format, av_frame_get_channels(frame)) ||
is->audio_filter_src.channel_layout != dec_channel_layout ||
is->audio_filter_src.freq != frame->sample_rate ||
is->auddec().pkt_serial != last_serial;
if (reconfigure) {
char buf1[1024], buf2[1024];
av_get_channel_layout_string(buf1, sizeof(buf1), -1, is->audio_filter_src.channel_layout);
av_get_channel_layout_string(buf2, sizeof(buf2), -1, dec_channel_layout);
av_log(NULL, AV_LOG_DEBUG,
"Audio frame changed from rate:%d ch:%d fmt:%s layout:%s serial:%d to rate:%d ch:%d fmt:%s layout:%s serial:%d\n",
is->audio_filter_src.freq, is->audio_filter_src.channels, av_get_sample_fmt_name(is->audio_filter_src.fmt), buf1, last_serial,
frame->sample_rate, av_frame_get_channels(frame), av_get_sample_fmt_name((AVSampleFormat)frame->format), buf2, is->auddec().pkt_serial);
is->audio_filter_src.fmt = (AVSampleFormat)frame->format;
is->audio_filter_src.channels = av_frame_get_channels(frame);
is->audio_filter_src.channel_layout = dec_channel_layout;
is->audio_filter_src.freq = frame->sample_rate;
last_serial = is->auddec().pkt_serial;
if ((ret = configure_audio_filters(is,gOptions. afilters, 1)) < 0)
goto the_end;
}
if ((ret = av_buffersrc_add_frame(is->in_audio_filter, frame)) < 0)
goto the_end;
while ((ret = av_buffersink_get_frame_flags(is->out_audio_filter, frame, 0)) >= 0) {
tb = is->out_audio_filter->inputs[0]->time_base;
#endif
if (!(af = is->sampq().peek_writable()))
goto the_end;
af->pts = (frame->pts == AV_NOPTS_VALUE) ? NAN : frame->pts * av_q2d(tb);
af->pos = av_frame_get_pkt_pos(frame);
af->serial = is->auddec().pkt_serial;
af->duration = av_q2d((AVRational){frame->nb_samples, frame->sample_rate});
av_frame_move_ref(af->frame, frame);
is->sampq().push();
#if CONFIG_AVFILTER
if (ps->audioq.serial != is->auddec().pkt_serial)
break;
}
if (ret == AVERROR_EOF)
is->auddec().finished = is->auddec().pkt_serial;
#endif
}
} while (ret >= 0 || ret == AVERROR(EAGAIN) || ret == AVERROR_EOF);
the_end:
#if CONFIG_AVFILTER
avfilter_graph_free(&is->agraph);
#endif
av_frame_free(&frame);
return ret;
}
static int decode_packet(int *got_frame, int cached)
{
int ret = 0;
if (pkt.stream_index == video_stream_idx) {
/* decode video frame */
ret = avcodec_decode_video2(video_dec_ctx, frame, got_frame, &pkt);
if (ret < 0) {
fprintf(stderr, "Error decoding video frame\n");
return ret;
}
if (*got_frame) {
printf("video_frame%s n:%d coded_n:%d pts:%s\n",
cached ? "(cached)" : "",
video_frame_count++, frame->coded_picture_number,
av_ts2timestr(frame->pts, &video_dec_ctx->time_base));
/* copy decoded frame to destination buffer:
* this is required since rawvideo expects non aligned data */
av_image_copy(video_dst_data, video_dst_linesize,
(const uint8_t **)(frame->data), frame->linesize,
video_dec_ctx->pix_fmt, video_dec_ctx->width, video_dec_ctx->height);
/* write to rawvideo file */
fwrite(video_dst_data[0], 1, video_dst_bufsize, video_dst_file);
}
} else if (pkt.stream_index == audio_stream_idx) {
/* decode audio frame */
ret = avcodec_decode_audio4(audio_dec_ctx, frame, got_frame, &pkt);
if (ret < 0) {
fprintf(stderr, "Error decoding audio frame\n");
return ret;
}
if (*got_frame) {
printf("audio_frame%s n:%d nb_samples:%d pts:%s\n",
cached ? "(cached)" : "",
audio_frame_count++, frame->nb_samples,
av_ts2timestr(frame->pts, &audio_dec_ctx->time_base));
ret = av_samples_alloc(audio_dst_data, &audio_dst_linesize, av_frame_get_channels(frame),
frame->nb_samples, frame->format, 1);
if (ret < 0) {
fprintf(stderr, "Could not allocate audio buffer\n");
return AVERROR(ENOMEM);
}
/* TODO: extend return code of the av_samples_* functions so that this call is not needed */
audio_dst_bufsize =
av_samples_get_buffer_size(NULL, av_frame_get_channels(frame),
frame->nb_samples, frame->format, 1);
/* copy audio data to destination buffer:
* this is required since rawaudio expects non aligned data */
av_samples_copy(audio_dst_data, frame->data, 0, 0,
frame->nb_samples, av_frame_get_channels(frame), frame->format);
/* write to rawaudio file */
fwrite(audio_dst_data[0], 1, audio_dst_bufsize, audio_dst_file);
av_freep(&audio_dst_data[0]);
}
}
return ret;
}
/**
* Decode a frame to a packet, run the result through SwrContext, if desired, encode it via an appropriate
* encoder, and write the results to the Java-side native buffer.
*
* @param aio FFAudio context
* @param cached true or false
* @return number of bytes placed into java buffer or a negative value, if something went wrong
*/
static int decode_packet(FFAudioIO *aio, int cached) {
int res = 0;
uint8_t **resample_buf = NULL;
jobject byte_buffer = NULL;
uint8_t *javaBuffer = NULL;
uint32_t out_buf_size = 0;
int out_buf_samples = 0;
int64_t out_channel_count;
int64_t out_sample_rate;
int flush = aio->got_frame;
enum AVSampleFormat out;
int bytesConsumed = 0;
init_ids(aio->env, aio->java_instance);
av_opt_get_int(aio->swr_context, "out_channel_count", 0, &out_channel_count);
av_opt_get_int(aio->swr_context, "out_sample_rate", 0, &out_sample_rate);
av_opt_get_sample_fmt(aio->swr_context, "out_sample_fmt", 0, &out);
resample_buf = av_mallocz(sizeof(uint8_t *) * 1); // one plane!
// make sure we really have an audio packet
if (aio->decode_packet.stream_index == aio->stream_index) {
// decode frame
// got_frame indicates whether we got a frame
bytesConsumed = avcodec_decode_audio4(aio->decode_context, aio->decode_frame, &aio->got_frame, &aio->decode_packet);
if (bytesConsumed < 0) {
throwUnsupportedAudioFileExceptionIfError(aio->env, bytesConsumed, "Failed to decode audio frame.");
return bytesConsumed;
}
if (aio->got_frame) {
aio->decoded_samples += aio->decode_frame->nb_samples;
out_buf_samples = aio->decode_frame->nb_samples;
#ifdef DEBUG
fprintf(stderr, "samples%s n:%" PRIu64 " nb_samples:%d pts:%s\n",
cached ? "(cached)" : "",
aio->decoded_samples, aio->decode_frame->nb_samples,
av_ts2timestr(aio->decode_frame->pts, &aio->decode_context->time_base));
#endif
// adjust out sample number for a different sample rate
// this is an estimate!!
out_buf_samples = av_rescale_rnd(
swr_get_delay(aio->swr_context, aio->stream->codecpar->sample_rate) + aio->decode_frame->nb_samples,
out_sample_rate,
aio->stream->codecpar->sample_rate,
AV_ROUND_UP
);
// allocate new aio->audio_data buffers
res = av_samples_alloc(aio->audio_data, NULL, av_frame_get_channels(aio->decode_frame),
aio->decode_frame->nb_samples, aio->decode_frame->format, 1);
if (res < 0) {
throwIOExceptionIfError(aio->env, res, "Could not allocate audio buffer.");
return AVERROR(ENOMEM);
}
// copy audio data to aio->audio_data
av_samples_copy(aio->audio_data, aio->decode_frame->data, 0, 0,
aio->decode_frame->nb_samples, av_frame_get_channels(aio->decode_frame), aio->decode_frame->format);
res = resample(aio, resample_buf, out_buf_samples, (const uint8_t **)aio->audio_data, aio->decode_frame->nb_samples);
if (res < 0) goto bail;
else out_buf_samples = res;
} else if (flush && swr_get_delay(aio->swr_context, aio->stream->codecpar->sample_rate)) {
res = resample(aio, resample_buf, swr_get_delay(aio->swr_context, aio->stream->codecpar->sample_rate), NULL, 0);
if (res < 0) goto bail;
else out_buf_samples = res;
} else {
#ifdef DEBUG
fprintf(stderr, "Got no frame.\n");
#endif
}
if (out_buf_samples > 0) {
res = av_samples_get_buffer_size(NULL, (int)out_channel_count, out_buf_samples, out, 1);
if (res < 0) goto bail;
else out_buf_size = res;
// ensure native buffer capacity
if (aio->java_buffer_capacity < out_buf_size) {
aio->java_buffer_capacity = (*aio->env)->CallIntMethod(aio->env, aio->java_instance, setNativeBufferCapacity_MID, (jint)out_buf_size);
}
// get java-managed byte buffer reference
byte_buffer = (*aio->env)->GetObjectField(aio->env, aio->java_instance, nativeBuffer_FID);
if (!byte_buffer) {
res = -1;
throwIOExceptionIfError(aio->env, 1, "Failed to get native buffer.");
goto bail;
}
// we have some samples, let's copy them to the java buffer, using the desired encoding
javaBuffer = (uint8_t *)(*aio->env)->GetDirectBufferAddress(aio->env, byte_buffer);
if (!javaBuffer) {
throwIOExceptionIfError(aio->env, 1, "Failed to get address for native buffer.");
goto bail;
}
if (aio->encode_context) {
aio->encode_frame->nb_samples = out_buf_samples;
res = encode_buffer(aio, resample_buf[0], out_buf_size, javaBuffer);
if (res < 0) {
out_buf_size = 0;
goto bail;
}
out_buf_size = res;
} else {
memcpy(javaBuffer, resample_buf[0], out_buf_size);
}
// we already wrote to the buffer, now we still need to
// set new bytebuffer limit and position to 0.
(*aio->env)->CallObjectMethod(aio->env, byte_buffer, rewind_MID);
(*aio->env)->CallObjectMethod(aio->env, byte_buffer, limit_MID, out_buf_size);
}
}
aio->resampled_samples += out_buf_size;
bail:
if (resample_buf) {
if (resample_buf[0]) av_freep(&resample_buf[0]);
av_free(resample_buf);
}
if (aio->audio_data[0]) av_freep(&aio->audio_data[0]);
return res;
}
static gboolean
gst_ffmpegauddec_negotiate (GstFFMpegAudDec * ffmpegdec,
AVCodecContext * context, AVFrame * frame, gboolean force)
{
GstFFMpegAudDecClass *oclass;
GstAudioFormat format;
gint channels;
GstAudioChannelPosition pos[64] = { 0, };
oclass = (GstFFMpegAudDecClass *) (G_OBJECT_GET_CLASS (ffmpegdec));
format = gst_ffmpeg_smpfmt_to_audioformat (frame->format);
if (format == GST_AUDIO_FORMAT_UNKNOWN)
goto no_caps;
channels =
av_get_channel_layout_nb_channels (av_frame_get_channel_layout (frame));
if (channels == 0)
channels = av_frame_get_channels (frame);
if (channels == 0)
goto no_caps;
if (!force && !settings_changed (ffmpegdec, frame))
return TRUE;
GST_DEBUG_OBJECT (ffmpegdec,
"Renegotiating audio from %dHz@%dchannels (%d) to %dHz@%dchannels (%d)",
ffmpegdec->info.rate, ffmpegdec->info.channels,
ffmpegdec->info.finfo->format, av_frame_get_sample_rate (frame), channels,
format);
gst_ffmpeg_channel_layout_to_gst (av_frame_get_channel_layout (frame),
channels, pos);
memcpy (ffmpegdec->ffmpeg_layout, pos,
sizeof (GstAudioChannelPosition) * channels);
/* Get GStreamer channel layout */
gst_audio_channel_positions_to_valid_order (pos, channels);
ffmpegdec->needs_reorder =
memcmp (pos, ffmpegdec->ffmpeg_layout, sizeof (pos[0]) * channels) != 0;
gst_audio_info_set_format (&ffmpegdec->info, format,
av_frame_get_sample_rate (frame), channels, pos);
if (!gst_audio_decoder_set_output_format (GST_AUDIO_DECODER (ffmpegdec),
&ffmpegdec->info))
goto caps_failed;
return TRUE;
/* ERRORS */
no_caps:
{
#ifdef HAVE_LIBAV_UNINSTALLED
/* using internal ffmpeg snapshot */
GST_ELEMENT_ERROR (ffmpegdec, CORE, NEGOTIATION,
("Could not find GStreamer caps mapping for libav codec '%s'.",
oclass->in_plugin->name), (NULL));
#else
/* using external ffmpeg */
GST_ELEMENT_ERROR (ffmpegdec, CORE, NEGOTIATION,
("Could not find GStreamer caps mapping for libav codec '%s', and "
"you are using an external libavcodec. This is most likely due to "
"a packaging problem and/or libavcodec having been upgraded to a "
"version that is not compatible with this version of "
"gstreamer-libav. Make sure your gstreamer-libav and libavcodec "
"packages come from the same source/repository.",
oclass->in_plugin->name), (NULL));
#endif
return FALSE;
}
caps_failed:
{
GST_ELEMENT_ERROR (ffmpegdec, CORE, NEGOTIATION, (NULL),
("Could not set caps for libav decoder (%s), not fixed?",
oclass->in_plugin->name));
memset (&ffmpegdec->info, 0, sizeof (ffmpegdec->info));
return FALSE;
}
}
static int ff_filter_frame_framed(AVFilterLink *link, AVFrame *frame)
{
int (*filter_frame)(AVFilterLink *, AVFrame *);
AVFilterContext *dstctx = link->dst;
AVFilterPad *dst = link->dstpad;
AVFrame *out = NULL;
int ret;
AVFilterCommand *cmd= link->dst->command_queue;
int64_t pts;
if (!(filter_frame = dst->filter_frame))
filter_frame = default_filter_frame;
/* copy the frame if needed */
if (dst->needs_writable && !av_frame_is_writable(frame)) {
av_log(link->dst, AV_LOG_DEBUG, "Copying data in avfilter.\n");
switch (link->type) {
case AVMEDIA_TYPE_VIDEO:
out = ff_get_video_buffer(link, link->w, link->h);
break;
case AVMEDIA_TYPE_AUDIO:
out = ff_get_audio_buffer(link, frame->nb_samples);
break;
default:
ret = AVERROR(EINVAL);
goto fail;
}
if (!out) {
ret = AVERROR(ENOMEM);
goto fail;
}
ret = av_frame_copy_props(out, frame);
if (ret < 0)
goto fail;
switch (link->type) {
case AVMEDIA_TYPE_VIDEO:
av_image_copy(out->data, out->linesize, (const uint8_t **)frame->data, frame->linesize,
frame->format, frame->width, frame->height);
break;
case AVMEDIA_TYPE_AUDIO:
av_samples_copy(out->extended_data, frame->extended_data,
0, 0, frame->nb_samples,
av_frame_get_channels(frame),
frame->format);
break;
default:
ret = AVERROR(EINVAL);
goto fail;
}
av_frame_free(&frame);
} else
out = frame;
while(cmd && cmd->time <= out->pts * av_q2d(link->time_base)){
av_log(link->dst, AV_LOG_DEBUG,
"Processing command time:%f command:%s arg:%s\n",
cmd->time, cmd->command, cmd->arg);
avfilter_process_command(link->dst, cmd->command, cmd->arg, 0, 0, cmd->flags);
ff_command_queue_pop(link->dst);
cmd= link->dst->command_queue;
}
pts = out->pts;
if (dstctx->enable_str) {
int64_t pos = av_frame_get_pkt_pos(out);
dstctx->var_values[VAR_N] = link->frame_count_out;
dstctx->var_values[VAR_T] = pts == AV_NOPTS_VALUE ? NAN : pts * av_q2d(link->time_base);
dstctx->var_values[VAR_W] = link->w;
dstctx->var_values[VAR_H] = link->h;
dstctx->var_values[VAR_POS] = pos == -1 ? NAN : pos;
dstctx->is_disabled = fabs(av_expr_eval(dstctx->enable, dstctx->var_values, NULL)) < 0.5;
if (dstctx->is_disabled &&
(dstctx->filter->flags & AVFILTER_FLAG_SUPPORT_TIMELINE_GENERIC))
filter_frame = default_filter_frame;
}
ret = filter_frame(link, out);
link->frame_count_out++;
ff_update_link_current_pts(link, pts);
return ret;
fail:
av_frame_free(&out);
av_frame_free(&frame);
return ret;
}
static int run_test(AVCodec *enc, AVCodec *dec, AVCodecContext *enc_ctx,
AVCodecContext *dec_ctx)
{
AVPacket enc_pkt;
AVFrame *in_frame, *out_frame;
uint8_t *raw_in = NULL, *raw_out = NULL;
int in_offset = 0, out_offset = 0;
int result = 0;
int got_output = 0;
int i = 0;
int in_frame_bytes, out_frame_bytes;
in_frame = av_frame_alloc();
if (!in_frame) {
av_log(NULL, AV_LOG_ERROR, "Can't allocate input frame\n");
return AVERROR(ENOMEM);
}
in_frame->nb_samples = enc_ctx->frame_size;
in_frame->format = enc_ctx->sample_fmt;
in_frame->channel_layout = enc_ctx->channel_layout;
if (av_frame_get_buffer(in_frame, 32) != 0) {
av_log(NULL, AV_LOG_ERROR, "Can't allocate a buffer for input frame\n");
return AVERROR(ENOMEM);
}
out_frame = av_frame_alloc();
if (!out_frame) {
av_log(NULL, AV_LOG_ERROR, "Can't allocate output frame\n");
return AVERROR(ENOMEM);
}
raw_in = av_malloc(in_frame->linesize[0] * NUMBER_OF_FRAMES);
if (!raw_in) {
av_log(NULL, AV_LOG_ERROR, "Can't allocate memory for raw_in\n");
return AVERROR(ENOMEM);
}
raw_out = av_malloc(in_frame->linesize[0] * NUMBER_OF_FRAMES);
if (!raw_out) {
av_log(NULL, AV_LOG_ERROR, "Can't allocate memory for raw_out\n");
return AVERROR(ENOMEM);
}
for (i = 0; i < NUMBER_OF_FRAMES; i++) {
av_init_packet(&enc_pkt);
enc_pkt.data = NULL;
enc_pkt.size = 0;
generate_raw_frame((uint16_t*)(in_frame->data[0]), i, enc_ctx->sample_rate,
enc_ctx->channels, enc_ctx->frame_size);
in_frame_bytes = in_frame->nb_samples * av_frame_get_channels(in_frame) * sizeof(uint16_t);
if (in_frame_bytes > in_frame->linesize[0]) {
av_log(NULL, AV_LOG_ERROR, "Incorrect value of input frame linesize\n");
return 1;
}
memcpy(raw_in + in_offset, in_frame->data[0], in_frame_bytes);
in_offset += in_frame_bytes;
result = avcodec_encode_audio2(enc_ctx, &enc_pkt, in_frame, &got_output);
if (result < 0) {
av_log(NULL, AV_LOG_ERROR, "Error encoding audio frame\n");
return result;
}
/* if we get an encoded packet, feed it straight to the decoder */
if (got_output) {
result = avcodec_decode_audio4(dec_ctx, out_frame, &got_output, &enc_pkt);
if (result < 0) {
av_log(NULL, AV_LOG_ERROR, "Error decoding audio packet\n");
return result;
}
if (got_output) {
if (result != enc_pkt.size) {
av_log(NULL, AV_LOG_INFO, "Decoder consumed only part of a packet, it is allowed to do so -- need to update this test\n");
return AVERROR_UNKNOWN;
}
if (in_frame->nb_samples != out_frame->nb_samples) {
av_log(NULL, AV_LOG_ERROR, "Error frames before and after decoding has different number of samples\n");
return AVERROR_UNKNOWN;
}
if (in_frame->channel_layout != out_frame->channel_layout) {
av_log(NULL, AV_LOG_ERROR, "Error frames before and after decoding has different channel layout\n");
return AVERROR_UNKNOWN;
}
if (in_frame->format != out_frame->format) {
av_log(NULL, AV_LOG_ERROR, "Error frames before and after decoding has different sample format\n");
return AVERROR_UNKNOWN;
}
out_frame_bytes = out_frame->nb_samples * av_frame_get_channels(out_frame) * sizeof(uint16_t);
if (out_frame_bytes > out_frame->linesize[0]) {
av_log(NULL, AV_LOG_ERROR, "Incorrect value of output frame linesize\n");
return 1;
}
memcpy(raw_out + out_offset, out_frame->data[0], out_frame_bytes);
out_offset += out_frame_bytes;
}
}
av_free_packet(&enc_pkt);
}
if (memcmp(raw_in, raw_out, out_frame_bytes * NUMBER_OF_FRAMES) != 0) {
av_log(NULL, AV_LOG_ERROR, "Output differs\n");
return 1;
}
av_log(NULL, AV_LOG_INFO, "OK\n");
av_freep(&raw_in);
av_freep(&raw_out);
av_frame_free(&in_frame);
av_frame_free(&out_frame);
return 0;
}
static int filter_frame(AVFilterLink *inlink, AVFrame *buf)
{
AVFilterContext *ctx = inlink->dst;
AShowInfoContext *s = ctx->priv;
char chlayout_str[128];
uint32_t checksum = 0;
int channels = inlink->channels;
int planar = av_sample_fmt_is_planar(buf->format);
int block_align = av_get_bytes_per_sample(buf->format) * (planar ? 1 : channels);
int data_size = buf->nb_samples * block_align;
int planes = planar ? channels : 1;
int i;
void *tmp_ptr = av_realloc_array(s->plane_checksums, channels, sizeof(*s->plane_checksums));
if (!tmp_ptr)
return AVERROR(ENOMEM);
s->plane_checksums = tmp_ptr;
for (i = 0; i < planes; i++) {
uint8_t *data = buf->extended_data[i];
s->plane_checksums[i] = av_adler32_update(0, data, data_size);
checksum = i ? av_adler32_update(checksum, data, data_size) :
s->plane_checksums[0];
}
av_get_channel_layout_string(chlayout_str, sizeof(chlayout_str), -1,
buf->channel_layout);
av_log(ctx, AV_LOG_INFO,
"n:%"PRId64" pts:%s pts_time:%s pos:%"PRId64" "
"fmt:%s channels:%d chlayout:%s rate:%d nb_samples:%d "
"checksum:%08"PRIX32" ",
inlink->frame_count_out,
av_ts2str(buf->pts), av_ts2timestr(buf->pts, &inlink->time_base),
av_frame_get_pkt_pos(buf),
av_get_sample_fmt_name(buf->format), av_frame_get_channels(buf), chlayout_str,
buf->sample_rate, buf->nb_samples,
checksum);
av_log(ctx, AV_LOG_INFO, "plane_checksums: [ ");
for (i = 0; i < planes; i++)
av_log(ctx, AV_LOG_INFO, "%08"PRIX32" ", s->plane_checksums[i]);
av_log(ctx, AV_LOG_INFO, "]\n");
for (i = 0; i < buf->nb_side_data; i++) {
AVFrameSideData *sd = buf->side_data[i];
av_log(ctx, AV_LOG_INFO, " side data - ");
switch (sd->type) {
case AV_FRAME_DATA_MATRIXENCODING: dump_matrixenc (ctx, sd); break;
case AV_FRAME_DATA_DOWNMIX_INFO: dump_downmix (ctx, sd); break;
case AV_FRAME_DATA_REPLAYGAIN: dump_replaygain(ctx, sd); break;
case AV_FRAME_DATA_AUDIO_SERVICE_TYPE: dump_audio_service_type(ctx, sd); break;
default: dump_unknown (ctx, sd); break;
}
av_log(ctx, AV_LOG_INFO, "\n");
}
return ff_filter_frame(inlink->dst->outputs[0], buf);
}
int AudioSamples::channelsCount() const
{
return m_raw ? av_frame_get_channels(m_raw) : 0;
}
