ffmpeg/ffmpeg_filter.c
Michael Niedermayer 91af76099e Merge commit '23aae62c2cb4504a09ceb8cd0cabc1c8b260f521'
* commit '23aae62c2cb4504a09ceb8cd0cabc1c8b260f521':
  alsdec: Check k used for rice decoder.
  avfiltergraph: silence an uninitialized variable warning
  xsubenc: reindent
  lavc: replace AVCodecContext.encode with subtitle-specific callback
  lavc: add const to private codec class initialization.
  avconv: don't pass a bogus parameter to avfilter_graph_create_filter().
  id3v2: strdup the genre name explicitly.
  lavf/id3v2: do not export empty fields.
  buffersrc: add const to the AVFrame* argument of av_buffersrc_write_frame()
  lavfi: replace empty input/output lists with null pointers

Conflicts:
	ffmpeg_filter.c
	libavcodec/alsdec.c
	libavcodec/dvdsubenc.c
	libavcodec/utils.c
	libavcodec/v210dec.h
	libavfilter/af_channelsplit.c
	libavfilter/avfiltergraph.c
	libavfilter/buffersrc.c
	libavfilter/src_movie.c
	libavfilter/vf_ass.c

Merged-by: Michael Niedermayer <michaelni@gmx.at>
2012-09-18 14:45:44 +02:00

775 lines
31 KiB
C

/*
* ffmpeg filter configuration
*
* This file is part of FFmpeg.
*
* FFmpeg 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.
*
* FFmpeg 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 FFmpeg; if not, write to the Free Software
* Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
*/
#include "ffmpeg.h"
#include "libavfilter/avfilter.h"
#include "libavfilter/avfiltergraph.h"
#include "libavfilter/buffersink.h"
#include "libavutil/audioconvert.h"
#include "libavutil/avassert.h"
#include "libavutil/avstring.h"
#include "libavutil/bprint.h"
#include "libavutil/pixdesc.h"
#include "libavutil/pixfmt.h"
#include "libavutil/imgutils.h"
#include "libavutil/samplefmt.h"
enum PixelFormat choose_pixel_fmt(AVStream *st, AVCodec *codec, enum PixelFormat target)
{
if (codec && codec->pix_fmts) {
const enum PixelFormat *p = codec->pix_fmts;
int has_alpha= av_pix_fmt_descriptors[target].nb_components % 2 == 0;
enum PixelFormat best= PIX_FMT_NONE;
if (st->codec->strict_std_compliance <= FF_COMPLIANCE_UNOFFICIAL) {
if (st->codec->codec_id == AV_CODEC_ID_MJPEG) {
p = (const enum PixelFormat[]) { PIX_FMT_YUVJ420P, PIX_FMT_YUVJ422P, PIX_FMT_YUV420P, PIX_FMT_YUV422P, PIX_FMT_NONE };
} else if (st->codec->codec_id == AV_CODEC_ID_LJPEG) {
p = (const enum PixelFormat[]) { PIX_FMT_YUVJ420P, PIX_FMT_YUVJ422P, PIX_FMT_YUVJ444P, PIX_FMT_YUV420P,
PIX_FMT_YUV422P, PIX_FMT_YUV444P, PIX_FMT_BGRA, PIX_FMT_NONE };
}
}
for (; *p != PIX_FMT_NONE; p++) {
best= avcodec_find_best_pix_fmt_of_2(best, *p, target, has_alpha, NULL);
if (*p == target)
break;
}
if (*p == PIX_FMT_NONE) {
if (target != PIX_FMT_NONE)
av_log(NULL, AV_LOG_WARNING,
"Incompatible pixel format '%s' for codec '%s', auto-selecting format '%s'\n",
av_pix_fmt_descriptors[target].name,
codec->name,
av_pix_fmt_descriptors[best].name);
return best;
}
}
return target;
}
void choose_sample_fmt(AVStream *st, AVCodec *codec)
{
if (codec && codec->sample_fmts) {
const enum AVSampleFormat *p = codec->sample_fmts;
for (; *p != -1; p++) {
if (*p == st->codec->sample_fmt)
break;
}
if (*p == -1) {
if((codec->capabilities & CODEC_CAP_LOSSLESS) && av_get_sample_fmt_name(st->codec->sample_fmt) > av_get_sample_fmt_name(codec->sample_fmts[0]))
av_log(NULL, AV_LOG_ERROR, "Conversion will not be lossless.\n");
if(av_get_sample_fmt_name(st->codec->sample_fmt))
av_log(NULL, AV_LOG_WARNING,
"Incompatible sample format '%s' for codec '%s', auto-selecting format '%s'\n",
av_get_sample_fmt_name(st->codec->sample_fmt),
codec->name,
av_get_sample_fmt_name(codec->sample_fmts[0]));
st->codec->sample_fmt = codec->sample_fmts[0];
}
}
}
static char *choose_pix_fmts(OutputStream *ost)
{
if (ost->keep_pix_fmt) {
if (ost->filter)
avfilter_graph_set_auto_convert(ost->filter->graph->graph,
AVFILTER_AUTO_CONVERT_NONE);
if (ost->st->codec->pix_fmt == PIX_FMT_NONE)
return NULL;
return av_strdup(av_get_pix_fmt_name(ost->st->codec->pix_fmt));
}
if (ost->st->codec->pix_fmt != PIX_FMT_NONE) {
return av_strdup(av_get_pix_fmt_name(choose_pixel_fmt(ost->st, ost->enc, ost->st->codec->pix_fmt)));
} else if (ost->enc && ost->enc->pix_fmts) {
const enum PixelFormat *p;
AVIOContext *s = NULL;
uint8_t *ret;
int len;
if (avio_open_dyn_buf(&s) < 0)
exit_program(1);
p = ost->enc->pix_fmts;
if (ost->st->codec->strict_std_compliance <= FF_COMPLIANCE_UNOFFICIAL) {
if (ost->st->codec->codec_id == AV_CODEC_ID_MJPEG) {
p = (const enum PixelFormat[]) { PIX_FMT_YUVJ420P, PIX_FMT_YUVJ422P, PIX_FMT_YUV420P, PIX_FMT_YUV422P, PIX_FMT_NONE };
} else if (ost->st->codec->codec_id == AV_CODEC_ID_LJPEG) {
p = (const enum PixelFormat[]) { PIX_FMT_YUVJ420P, PIX_FMT_YUVJ422P, PIX_FMT_YUVJ444P, PIX_FMT_YUV420P,
PIX_FMT_YUV422P, PIX_FMT_YUV444P, PIX_FMT_BGRA, PIX_FMT_NONE };
}
}
for (; *p != PIX_FMT_NONE; p++) {
const char *name = av_get_pix_fmt_name(*p);
avio_printf(s, "%s:", name);
}
len = avio_close_dyn_buf(s, &ret);
ret[len - 1] = 0;
return ret;
} else
return NULL;
}
/**
* Define a function for building a string containing a list of
* allowed formats,
*/
#define DEF_CHOOSE_FORMAT(type, var, supported_list, none, get_name, separator)\
static char *choose_ ## var ## s(OutputStream *ost) \
{ \
if (ost->st->codec->var != none) { \
get_name(ost->st->codec->var); \
return av_strdup(name); \
} else if (ost->enc->supported_list) { \
const type *p; \
AVIOContext *s = NULL; \
uint8_t *ret; \
int len; \
\
if (avio_open_dyn_buf(&s) < 0) \
exit_program(1); \
\
for (p = ost->enc->supported_list; *p != none; p++) { \
get_name(*p); \
avio_printf(s, "%s" separator, name); \
} \
len = avio_close_dyn_buf(s, &ret); \
ret[len - 1] = 0; \
return ret; \
} else \
return NULL; \
}
// DEF_CHOOSE_FORMAT(enum PixelFormat, pix_fmt, pix_fmts, PIX_FMT_NONE,
// GET_PIX_FMT_NAME, ":")
DEF_CHOOSE_FORMAT(enum AVSampleFormat, sample_fmt, sample_fmts,
AV_SAMPLE_FMT_NONE, GET_SAMPLE_FMT_NAME, ",")
DEF_CHOOSE_FORMAT(int, sample_rate, supported_samplerates, 0,
GET_SAMPLE_RATE_NAME, ",")
DEF_CHOOSE_FORMAT(uint64_t, channel_layout, channel_layouts, 0,
GET_CH_LAYOUT_NAME, ",")
FilterGraph *init_simple_filtergraph(InputStream *ist, OutputStream *ost)
{
FilterGraph *fg = av_mallocz(sizeof(*fg));
if (!fg)
exit_program(1);
fg->index = nb_filtergraphs;
fg->outputs = grow_array(fg->outputs, sizeof(*fg->outputs), &fg->nb_outputs,
fg->nb_outputs + 1);
if (!(fg->outputs[0] = av_mallocz(sizeof(*fg->outputs[0]))))
exit_program(1);
fg->outputs[0]->ost = ost;
fg->outputs[0]->graph = fg;
ost->filter = fg->outputs[0];
fg->inputs = grow_array(fg->inputs, sizeof(*fg->inputs), &fg->nb_inputs,
fg->nb_inputs + 1);
if (!(fg->inputs[0] = av_mallocz(sizeof(*fg->inputs[0]))))
exit_program(1);
fg->inputs[0]->ist = ist;
fg->inputs[0]->graph = fg;
ist->filters = grow_array(ist->filters, sizeof(*ist->filters),
&ist->nb_filters, ist->nb_filters + 1);
ist->filters[ist->nb_filters - 1] = fg->inputs[0];
filtergraphs = grow_array(filtergraphs, sizeof(*filtergraphs),
&nb_filtergraphs, nb_filtergraphs + 1);
filtergraphs[nb_filtergraphs - 1] = fg;
return fg;
}
static void init_input_filter(FilterGraph *fg, AVFilterInOut *in)
{
InputStream *ist = NULL;
enum AVMediaType type = avfilter_pad_get_type(in->filter_ctx->input_pads, in->pad_idx);
int i;
// TODO: support other filter types
if (type != AVMEDIA_TYPE_VIDEO && type != AVMEDIA_TYPE_AUDIO) {
av_log(NULL, AV_LOG_FATAL, "Only video and audio filters supported "
"currently.\n");
exit_program(1);
}
if (in->name) {
AVFormatContext *s;
AVStream *st = NULL;
char *p;
int file_idx = strtol(in->name, &p, 0);
if (file_idx < 0 || file_idx >= nb_input_files) {
av_log(NULL, AV_LOG_FATAL, "Invalid file index %d in filtergraph description %s.\n",
file_idx, fg->graph_desc);
exit_program(1);
}
s = input_files[file_idx]->ctx;
for (i = 0; i < s->nb_streams; i++) {
enum AVMediaType stream_type = s->streams[i]->codec->codec_type;
if (stream_type != type &&
!(stream_type == AVMEDIA_TYPE_SUBTITLE &&
type == AVMEDIA_TYPE_VIDEO /* sub2video hack */))
continue;
if (check_stream_specifier(s, s->streams[i], *p == ':' ? p + 1 : p) == 1) {
st = s->streams[i];
break;
}
}
if (!st) {
av_log(NULL, AV_LOG_FATAL, "Stream specifier '%s' in filtergraph description %s "
"matches no streams.\n", p, fg->graph_desc);
exit_program(1);
}
ist = input_streams[input_files[file_idx]->ist_index + st->index];
} else {
/* find the first unused stream of corresponding type */
for (i = 0; i < nb_input_streams; i++) {
ist = input_streams[i];
if (ist->st->codec->codec_type == type && ist->discard)
break;
}
if (i == nb_input_streams) {
av_log(NULL, AV_LOG_FATAL, "Cannot find a matching stream for "
"unlabeled input pad %d on filter %s\n", in->pad_idx,
in->filter_ctx->name);
exit_program(1);
}
}
av_assert0(ist);
ist->discard = 0;
ist->decoding_needed++;
ist->st->discard = AVDISCARD_NONE;
fg->inputs = grow_array(fg->inputs, sizeof(*fg->inputs),
&fg->nb_inputs, fg->nb_inputs + 1);
if (!(fg->inputs[fg->nb_inputs - 1] = av_mallocz(sizeof(*fg->inputs[0]))))
exit_program(1);
fg->inputs[fg->nb_inputs - 1]->ist = ist;
fg->inputs[fg->nb_inputs - 1]->graph = fg;
ist->filters = grow_array(ist->filters, sizeof(*ist->filters),
&ist->nb_filters, ist->nb_filters + 1);
ist->filters[ist->nb_filters - 1] = fg->inputs[fg->nb_inputs - 1];
}
static int configure_output_video_filter(FilterGraph *fg, OutputFilter *ofilter, AVFilterInOut *out)
{
char *pix_fmts;
OutputStream *ost = ofilter->ost;
AVCodecContext *codec = ost->st->codec;
AVFilterContext *last_filter = out->filter_ctx;
int pad_idx = out->pad_idx;
int ret;
char name[255];
AVBufferSinkParams *buffersink_params = av_buffersink_params_alloc();
snprintf(name, sizeof(name), "output stream %d:%d", ost->file_index, ost->index);
ret = avfilter_graph_create_filter(&ofilter->filter,
avfilter_get_by_name("ffbuffersink"),
name, NULL, NULL, fg->graph);
av_freep(&buffersink_params);
if (ret < 0)
return ret;
if (codec->width || codec->height) {
char args[255];
AVFilterContext *filter;
snprintf(args, sizeof(args), "%d:%d:flags=0x%X",
codec->width,
codec->height,
(unsigned)ost->sws_flags);
snprintf(name, sizeof(name), "scaler for output stream %d:%d",
ost->file_index, ost->index);
if ((ret = avfilter_graph_create_filter(&filter, avfilter_get_by_name("scale"),
name, args, NULL, fg->graph)) < 0)
return ret;
if ((ret = avfilter_link(last_filter, pad_idx, filter, 0)) < 0)
return ret;
last_filter = filter;
pad_idx = 0;
}
if ((pix_fmts = choose_pix_fmts(ost))) {
AVFilterContext *filter;
snprintf(name, sizeof(name), "pixel format for output stream %d:%d",
ost->file_index, ost->index);
if ((ret = avfilter_graph_create_filter(&filter,
avfilter_get_by_name("format"),
"format", pix_fmts, NULL,
fg->graph)) < 0)
return ret;
if ((ret = avfilter_link(last_filter, pad_idx, filter, 0)) < 0)
return ret;
last_filter = filter;
pad_idx = 0;
av_freep(&pix_fmts);
}
if (ost->frame_rate.num && 0) {
AVFilterContext *fps;
char args[255];
snprintf(args, sizeof(args), "fps=%d/%d", ost->frame_rate.num,
ost->frame_rate.den);
snprintf(name, sizeof(name), "fps for output stream %d:%d",
ost->file_index, ost->index);
ret = avfilter_graph_create_filter(&fps, avfilter_get_by_name("fps"),
name, args, NULL, fg->graph);
if (ret < 0)
return ret;
ret = avfilter_link(last_filter, pad_idx, fps, 0);
if (ret < 0)
return ret;
last_filter = fps;
pad_idx = 0;
}
if ((ret = avfilter_link(last_filter, pad_idx, ofilter->filter, 0)) < 0)
return ret;
return 0;
}
static int configure_output_audio_filter(FilterGraph *fg, OutputFilter *ofilter, AVFilterInOut *out)
{
OutputStream *ost = ofilter->ost;
AVCodecContext *codec = ost->st->codec;
AVFilterContext *last_filter = out->filter_ctx;
int pad_idx = out->pad_idx;
char *sample_fmts, *sample_rates, *channel_layouts;
char name[255];
int ret;
snprintf(name, sizeof(name), "output stream %d:%d", ost->file_index, ost->index);
ret = avfilter_graph_create_filter(&ofilter->filter,
avfilter_get_by_name("ffabuffersink"),
name, NULL, NULL, fg->graph);
if (ret < 0)
return ret;
#define AUTO_INSERT_FILTER(opt_name, filter_name, arg) do { \
AVFilterContext *filt_ctx; \
\
av_log(NULL, AV_LOG_INFO, opt_name " is forwarded to lavfi " \
"similarly to -af " filter_name "=%s.\n", arg); \
\
ret = avfilter_graph_create_filter(&filt_ctx, \
avfilter_get_by_name(filter_name), \
filter_name, arg, NULL, fg->graph); \
if (ret < 0) \
return ret; \
\
ret = avfilter_link(last_filter, pad_idx, filt_ctx, 0); \
if (ret < 0) \
return ret; \
\
last_filter = filt_ctx; \
pad_idx = 0; \
} while (0)
if (ost->audio_channels_mapped) {
int i;
AVBPrint pan_buf;
av_bprint_init(&pan_buf, 256, 8192);
av_bprintf(&pan_buf, "0x%"PRIx64,
av_get_default_channel_layout(ost->audio_channels_mapped));
for (i = 0; i < ost->audio_channels_mapped; i++)
if (ost->audio_channels_map[i] != -1)
av_bprintf(&pan_buf, ":c%d=c%d", i, ost->audio_channels_map[i]);
AUTO_INSERT_FILTER("-map_channel", "pan", pan_buf.str);
av_bprint_finalize(&pan_buf, NULL);
}
if (codec->channels && !codec->channel_layout)
codec->channel_layout = av_get_default_channel_layout(codec->channels);
sample_fmts = choose_sample_fmts(ost);
sample_rates = choose_sample_rates(ost);
channel_layouts = choose_channel_layouts(ost);
if (sample_fmts || sample_rates || channel_layouts) {
AVFilterContext *format;
char args[256];
args[0] = 0;
if (sample_fmts)
av_strlcatf(args, sizeof(args), "sample_fmts=%s:",
sample_fmts);
if (sample_rates)
av_strlcatf(args, sizeof(args), "sample_rates=%s:",
sample_rates);
if (channel_layouts)
av_strlcatf(args, sizeof(args), "channel_layouts=%s:",
channel_layouts);
av_freep(&sample_fmts);
av_freep(&sample_rates);
av_freep(&channel_layouts);
snprintf(name, sizeof(name), "audio format for output stream %d:%d",
ost->file_index, ost->index);
ret = avfilter_graph_create_filter(&format,
avfilter_get_by_name("aformat"),
name, args, NULL, fg->graph);
if (ret < 0)
return ret;
ret = avfilter_link(last_filter, pad_idx, format, 0);
if (ret < 0)
return ret;
last_filter = format;
pad_idx = 0;
}
if (audio_volume != 256 && 0) {
char args[256];
snprintf(args, sizeof(args), "%f", audio_volume / 256.);
AUTO_INSERT_FILTER("-vol", "volume", args);
}
if ((ret = avfilter_link(last_filter, pad_idx, ofilter->filter, 0)) < 0)
return ret;
return 0;
}
#define DESCRIBE_FILTER_LINK(f, inout, in) \
{ \
AVFilterContext *ctx = inout->filter_ctx; \
AVFilterPad *pads = in ? ctx->input_pads : ctx->output_pads; \
int nb_pads = in ? ctx->input_count : ctx->output_count; \
AVIOContext *pb; \
\
if (avio_open_dyn_buf(&pb) < 0) \
exit_program(1); \
\
avio_printf(pb, "%s", ctx->filter->name); \
if (nb_pads > 1) \
avio_printf(pb, ":%s", avfilter_pad_get_name(pads, inout->pad_idx));\
avio_w8(pb, 0); \
avio_close_dyn_buf(pb, &f->name); \
}
int configure_output_filter(FilterGraph *fg, OutputFilter *ofilter, AVFilterInOut *out)
{
av_freep(&ofilter->name);
DESCRIBE_FILTER_LINK(ofilter, out, 0);
switch (avfilter_pad_get_type(out->filter_ctx->output_pads, out->pad_idx)) {
case AVMEDIA_TYPE_VIDEO: return configure_output_video_filter(fg, ofilter, out);
case AVMEDIA_TYPE_AUDIO: return configure_output_audio_filter(fg, ofilter, out);
default: av_assert0(0);
}
}
static int sub2video_prepare(InputStream *ist)
{
AVFormatContext *avf = input_files[ist->file_index]->ctx;
int i, ret, w, h;
uint8_t *image[4];
int linesize[4];
/* Compute the size of the canvas for the subtitles stream.
If the subtitles codec has set a size, use it. Otherwise use the
maximum dimensions of the video streams in the same file. */
w = ist->st->codec->width;
h = ist->st->codec->height;
if (!(w && h)) {
for (i = 0; i < avf->nb_streams; i++) {
if (avf->streams[i]->codec->codec_type == AVMEDIA_TYPE_VIDEO) {
w = FFMAX(w, avf->streams[i]->codec->width);
h = FFMAX(h, avf->streams[i]->codec->height);
}
}
if (!(w && h)) {
w = FFMAX(w, 720);
h = FFMAX(h, 576);
}
av_log(avf, AV_LOG_INFO, "sub2video: using %dx%d canvas\n", w, h);
}
ist->sub2video.w = ist->st->codec->width = w;
ist->sub2video.h = ist->st->codec->height = h;
/* rectangles are PIX_FMT_PAL8, but we have no guarantee that the
palettes for all rectangles are identical or compatible */
ist->st->codec->pix_fmt = PIX_FMT_RGB32;
ret = av_image_alloc(image, linesize, w, h, PIX_FMT_RGB32, 32);
if (ret < 0)
return ret;
memset(image[0], 0, h * linesize[0]);
ist->sub2video.ref = avfilter_get_video_buffer_ref_from_arrays(
image, linesize, AV_PERM_READ | AV_PERM_PRESERVE,
w, h, PIX_FMT_RGB32);
if (!ist->sub2video.ref) {
av_free(image[0]);
return AVERROR(ENOMEM);
}
return 0;
}
static int configure_input_video_filter(FilterGraph *fg, InputFilter *ifilter,
AVFilterInOut *in)
{
AVFilterContext *first_filter = in->filter_ctx;
AVFilter *filter = avfilter_get_by_name("buffer");
InputStream *ist = ifilter->ist;
AVRational tb = ist->framerate.num ? av_inv_q(ist->framerate) :
ist->st->time_base;
AVRational fr = ist->framerate.num ? ist->framerate :
ist->st->r_frame_rate;
AVRational sar;
AVBPrint args;
char name[255];
int pad_idx = in->pad_idx;
int ret;
if (ist->st->codec->codec_type == AVMEDIA_TYPE_SUBTITLE) {
ret = sub2video_prepare(ist);
if (ret < 0)
return ret;
}
sar = ist->st->sample_aspect_ratio.num ?
ist->st->sample_aspect_ratio :
ist->st->codec->sample_aspect_ratio;
if(!sar.den)
sar = (AVRational){0,1};
av_bprint_init(&args, 0, 1);
av_bprintf(&args,
"video_size=%dx%d:pix_fmt=%d:time_base=%d/%d:"
"pixel_aspect=%d/%d:sws_param=flags=%d", ist->st->codec->width,
ist->st->codec->height, ist->st->codec->pix_fmt,
tb.num, tb.den, sar.num, sar.den,
SWS_BILINEAR + ((ist->st->codec->flags&CODEC_FLAG_BITEXACT) ? SWS_BITEXACT:0));
if (fr.num && fr.den)
av_bprintf(&args, ":frame_rate=%d/%d", fr.num, fr.den);
snprintf(name, sizeof(name), "graph %d input from stream %d:%d", fg->index,
ist->file_index, ist->st->index);
if ((ret = avfilter_graph_create_filter(&ifilter->filter, filter, name,
args.str, NULL, fg->graph)) < 0)
return ret;
if (ist->framerate.num) {
AVFilterContext *setpts;
snprintf(name, sizeof(name), "force CFR for input from stream %d:%d",
ist->file_index, ist->st->index);
if ((ret = avfilter_graph_create_filter(&setpts,
avfilter_get_by_name("setpts"),
name, "N", NULL,
fg->graph)) < 0)
return ret;
if ((ret = avfilter_link(setpts, 0, first_filter, pad_idx)) < 0)
return ret;
first_filter = setpts;
pad_idx = 0;
}
if ((ret = avfilter_link(ifilter->filter, 0, first_filter, pad_idx)) < 0)
return ret;
return 0;
}
static int configure_input_audio_filter(FilterGraph *fg, InputFilter *ifilter,
AVFilterInOut *in)
{
AVFilterContext *first_filter = in->filter_ctx;
AVFilter *filter = avfilter_get_by_name("abuffer");
InputStream *ist = ifilter->ist;
int pad_idx = in->pad_idx;
char args[255], name[255];
int ret;
snprintf(args, sizeof(args), "time_base=%d/%d:sample_rate=%d:sample_fmt=%s"
":channel_layout=0x%"PRIx64,
1, ist->st->codec->sample_rate,
ist->st->codec->sample_rate,
av_get_sample_fmt_name(ist->st->codec->sample_fmt),
ist->st->codec->channel_layout);
snprintf(name, sizeof(name), "graph %d input from stream %d:%d", fg->index,
ist->file_index, ist->st->index);
if ((ret = avfilter_graph_create_filter(&ifilter->filter, filter,
name, args, NULL,
fg->graph)) < 0)
return ret;
#define AUTO_INSERT_FILTER_INPUT(opt_name, filter_name, arg) do { \
AVFilterContext *filt_ctx; \
\
av_log(NULL, AV_LOG_INFO, opt_name " is forwarded to lavfi " \
"similarly to -af " filter_name "=%s.\n", arg); \
\
snprintf(name, sizeof(name), "graph %d %s for input stream %d:%d", \
fg->index, filter_name, ist->file_index, ist->st->index); \
ret = avfilter_graph_create_filter(&filt_ctx, \
avfilter_get_by_name(filter_name), \
name, arg, NULL, fg->graph); \
if (ret < 0) \
return ret; \
\
ret = avfilter_link(filt_ctx, 0, first_filter, pad_idx); \
if (ret < 0) \
return ret; \
\
first_filter = filt_ctx; \
} while (0)
if (audio_sync_method > 0) {
char args[256] = {0};
av_strlcatf(args, sizeof(args), "min_comp=0.001:min_hard_comp=%f", audio_drift_threshold);
if (audio_sync_method > 1)
av_strlcatf(args, sizeof(args), ":max_soft_comp=%f", audio_sync_method/(double)ist->st->codec->sample_rate);
AUTO_INSERT_FILTER_INPUT("-async", "aresample", args);
}
// if (ost->audio_channels_mapped) {
// int i;
// AVBPrint pan_buf;
// av_bprint_init(&pan_buf, 256, 8192);
// av_bprintf(&pan_buf, "0x%"PRIx64,
// av_get_default_channel_layout(ost->audio_channels_mapped));
// for (i = 0; i < ost->audio_channels_mapped; i++)
// if (ost->audio_channels_map[i] != -1)
// av_bprintf(&pan_buf, ":c%d=c%d", i, ost->audio_channels_map[i]);
// AUTO_INSERT_FILTER_INPUT("-map_channel", "pan", pan_buf.str);
// av_bprint_finalize(&pan_buf, NULL);
// }
if (audio_volume != 256) {
char args[256];
snprintf(args, sizeof(args), "%f", audio_volume / 256.);
AUTO_INSERT_FILTER_INPUT("-vol", "volume", args);
}
if ((ret = avfilter_link(ifilter->filter, 0, first_filter, pad_idx)) < 0)
return ret;
return 0;
}
static int configure_input_filter(FilterGraph *fg, InputFilter *ifilter,
AVFilterInOut *in)
{
av_freep(&ifilter->name);
DESCRIBE_FILTER_LINK(ifilter, in, 1);
switch (avfilter_pad_get_type(in->filter_ctx->input_pads, in->pad_idx)) {
case AVMEDIA_TYPE_VIDEO: return configure_input_video_filter(fg, ifilter, in);
case AVMEDIA_TYPE_AUDIO: return configure_input_audio_filter(fg, ifilter, in);
default: av_assert0(0);
}
}
int configure_filtergraph(FilterGraph *fg)
{
AVFilterInOut *inputs, *outputs, *cur;
int ret, i, init = !fg->graph, simple = !fg->graph_desc;
const char *graph_desc = simple ? fg->outputs[0]->ost->avfilter :
fg->graph_desc;
avfilter_graph_free(&fg->graph);
if (!(fg->graph = avfilter_graph_alloc()))
return AVERROR(ENOMEM);
if (simple) {
OutputStream *ost = fg->outputs[0]->ost;
char args[255];
snprintf(args, sizeof(args), "flags=0x%X", (unsigned)ost->sws_flags);
fg->graph->scale_sws_opts = av_strdup(args);
}
if ((ret = avfilter_graph_parse2(fg->graph, graph_desc, &inputs, &outputs)) < 0)
return ret;
if (simple && (!inputs || inputs->next || !outputs || outputs->next)) {
av_log(NULL, AV_LOG_ERROR, "Simple filtergraph '%s' does not have "
"exactly one input and output.\n", graph_desc);
return AVERROR(EINVAL);
}
for (cur = inputs; !simple && init && cur; cur = cur->next)
init_input_filter(fg, cur);
for (cur = inputs, i = 0; cur; cur = cur->next, i++)
if ((ret = configure_input_filter(fg, fg->inputs[i], cur)) < 0)
return ret;
avfilter_inout_free(&inputs);
if (!init || simple) {
/* we already know the mappings between lavfi outputs and output streams,
* so we can finish the setup */
for (cur = outputs, i = 0; cur; cur = cur->next, i++)
configure_output_filter(fg, fg->outputs[i], cur);
avfilter_inout_free(&outputs);
if ((ret = avfilter_graph_config(fg->graph, NULL)) < 0)
return ret;
} else {
/* wait until output mappings are processed */
for (cur = outputs; cur;) {
fg->outputs = grow_array(fg->outputs, sizeof(*fg->outputs),
&fg->nb_outputs, fg->nb_outputs + 1);
if (!(fg->outputs[fg->nb_outputs - 1] = av_mallocz(sizeof(*fg->outputs[0]))))
exit_program(1);
fg->outputs[fg->nb_outputs - 1]->graph = fg;
fg->outputs[fg->nb_outputs - 1]->out_tmp = cur;
cur = cur->next;
fg->outputs[fg->nb_outputs - 1]->out_tmp->next = NULL;
}
}
return 0;
}
int ist_in_filtergraph(FilterGraph *fg, InputStream *ist)
{
int i;
for (i = 0; i < fg->nb_inputs; i++)
if (fg->inputs[i]->ist == ist)
return 1;
return 0;
}