API tools faq
paste
Advertisement
Guest User

WebM transcode w/ incorrect frame rate

a guest
Apr 15th, 2014
76
0
Never
Add comment
Not a member of Pastebin yet? Sign Up, it unlocks many cool features!
C 22.08 KB | None | 0 0
raw download clone embed print report
  1. /*
  2.  * Copyright (c) 2010 Nicolas George
  3.  * Copyright (c) 2011 Stefano Sabatini
  4.  * Copyright (c) 2014 Andrey Utkin
  5.  *
  6.  * Permission is hereby granted, free of charge, to any person obtaining a copy
  7.  * of this software and associated documentation files (the "Software"), to deal
  8.  * in the Software without restriction, including without limitation the rights
  9.  * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
  10.  * copies of the Software, and to permit persons to whom the Software is
  11.  * furnished to do so, subject to the following conditions:
  12.  *
  13.  * The above copyright notice and this permission notice shall be included in
  14.  * all copies or substantial portions of the Software.
  15.  *
  16.  * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
  17.  * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
  18.  * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
  19.  * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
  20.  * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
  21.  * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
  22.  * THE SOFTWARE.
  23.  */
  24.  
  25. /**
  26.  * @file
  27.  * API example for demuxing, decoding, filtering, encoding and muxing
  28.  * @example doc/examples/transcoding.c
  29.  */
  30.  
  31. #include <libavcodec/avcodec.h>
  32. #include <libavformat/avformat.h>
  33. #include <libavfilter/avfiltergraph.h>
  34. #include <libavfilter/avcodec.h>
  35. #include <libavfilter/buffersink.h>
  36. #include <libavfilter/buffersrc.h>
  37. #include <libavutil/opt.h>
  38. #include <libavutil/pixdesc.h>
  39.  
  40. #define STATS_LOG "stats.log"
  41.  
  42. static AVFormatContext *ifmt_ctx;
  43. static AVFormatContext *ofmt_ctx;
  44. typedef struct FilteringContext {
  45.   AVFilterContext *buffersink_ctx;
  46.   AVFilterContext *buffersrc_ctx;
  47.   AVFilterGraph *filter_graph;
  48. } FilteringContext;
  49. static FilteringContext *filter_ctx;
  50.  
  51. static int open_input_file(const char *filename) {
  52.   int ret;
  53.   unsigned int i;
  54.  
  55.   ifmt_ctx = NULL;
  56.   if ((ret = avformat_open_input(&ifmt_ctx, filename, NULL, NULL)) < 0) {
  57.     av_log(NULL, AV_LOG_ERROR, "Cannot open input file\n");
  58.     return ret;
  59.   }
  60.  
  61.   if ((ret = avformat_find_stream_info(ifmt_ctx, NULL)) < 0) {
  62.     av_log(NULL, AV_LOG_ERROR, "Cannot find stream information\n");
  63.     return ret;
  64.   }
  65.  
  66.   for (i = 0; i < ifmt_ctx->nb_streams; i++) {
  67.     AVStream *stream;
  68.     AVCodecContext *codec_ctx;
  69.     stream = ifmt_ctx->streams[i];
  70.     codec_ctx = stream->codec;
  71.     /* Reencode video & audio and remux subtitles etc. */
  72.     if (codec_ctx->codec_type == AVMEDIA_TYPE_VIDEO
  73.     || codec_ctx->codec_type == AVMEDIA_TYPE_AUDIO) {
  74.       /* Open decoder */
  75.       ret = avcodec_open2(codec_ctx,
  76.               avcodec_find_decoder(codec_ctx->codec_id), NULL);
  77.       if (ret < 0) {
  78.     av_log(NULL, AV_LOG_ERROR, "Failed to open decoder for stream #%u\n", i);
  79.     return ret;
  80.       }
  81.     }
  82.   }
  83.  
  84.   av_dump_format(ifmt_ctx, 0, filename, 0);
  85.   return 0;
  86. }
  87.  
  88. static int init_output_context(char* filename) {
  89.   int ret;
  90.   ofmt_ctx = NULL;
  91.  
  92.   avformat_alloc_output_context2(&ofmt_ctx, NULL, NULL, filename);
  93.   if (!ofmt_ctx) {
  94.     av_log(NULL, AV_LOG_ERROR, "Could not create output context\n");
  95.     return AVERROR_UNKNOWN;
  96.   }
  97.  
  98.   return 0;
  99. }
  100.  
  101. static int init_webm_encoders(int audioBitRate, int crf, int videoMaxBitRate, int threads,
  102.                   char* quality, int speed, int pass, char* stats) {
  103.   AVStream *out_stream;
  104.   AVStream *in_stream;
  105.   AVCodecContext *dec_ctx, *enc_ctx;
  106.   AVCodec *encoder;
  107.   int ret;
  108.   unsigned int i;
  109.  
  110.   for (i = 0; i < ifmt_ctx->nb_streams; i++) {
  111.     in_stream = ifmt_ctx->streams[i];
  112.     dec_ctx = in_stream->codec;
  113.     if (dec_ctx->codec_type == AVMEDIA_TYPE_VIDEO || dec_ctx->codec_type == AVMEDIA_TYPE_AUDIO) {
  114.  
  115.       AVDictionary *opts = NULL;
  116.       if (dec_ctx->codec_type == AVMEDIA_TYPE_VIDEO) {
  117.     encoder = avcodec_find_encoder(AV_CODEC_ID_VP8);
  118.     out_stream = avformat_new_stream(ofmt_ctx, encoder);
  119.     if (!out_stream) {
  120.       av_log(NULL, AV_LOG_ERROR, "Failed allocating output stream\n");
  121.       return AVERROR_UNKNOWN;
  122.     }
  123.  
  124.     enc_ctx = out_stream->codec;
  125.     enc_ctx->height = dec_ctx->height;
  126.     enc_ctx->width = dec_ctx->width;
  127.     enc_ctx->sample_aspect_ratio = dec_ctx->sample_aspect_ratio;
  128.     /* take first format from list of supported formats */
  129.     enc_ctx->pix_fmt = encoder->pix_fmts[0];
  130.     /* video time_base can be set to whatever is handy and supported by encoder */
  131.     /* enc_ctx->time_base = dec_ctx->time_base; */
  132.     enc_ctx->time_base.num = 1;
  133.     enc_ctx->time_base.den = 24;
  134.  
  135.     enc_ctx->bit_rate = videoMaxBitRate;
  136.     enc_ctx->thread_count = threads;
  137.     switch (pass) {
  138.     case 1:
  139.       enc_ctx->flags |= CODEC_FLAG_PASS1;
  140.       break;
  141.     case 2:
  142.       enc_ctx->flags |= CODEC_FLAG_PASS2;
  143.       if (stats) {
  144.         enc_ctx->stats_in = stats;
  145.       }
  146.       break;
  147.     }
  148.  
  149.     char crfString[3];
  150.     snprintf(crfString, 3, "%d", crf);
  151.     av_dict_set(&opts, "crf", crfString, 0);
  152.     av_dict_set(&opts, "quality", quality, 0);
  153.     char speedString[3];
  154.     snprintf(speedString, 3, "%d", speed);
  155.     av_dict_set(&opts, "speed", speedString, 0);
  156.       } else {
  157.     encoder = avcodec_find_encoder(AV_CODEC_ID_VORBIS);
  158.     out_stream = avformat_new_stream(ofmt_ctx, encoder);
  159.     if (!out_stream) {
  160.       av_log(NULL, AV_LOG_ERROR, "Failed allocating output stream\n");
  161.       return AVERROR_UNKNOWN;
  162.     }
  163.  
  164.     /* in_stream = ifmt_ctx->streams[i]; */
  165.     /* dec_ctx = in_stream->codec; */
  166.     enc_ctx = out_stream->codec;
  167.     /* encoder = out_stream->codec->codec; */
  168.  
  169.     enc_ctx->sample_rate = dec_ctx->sample_rate;
  170.     enc_ctx->channel_layout = dec_ctx->channel_layout;
  171.     enc_ctx->channels = av_get_channel_layout_nb_channels(enc_ctx->channel_layout);
  172.     /* take first format from list of supported formats */
  173.     enc_ctx->sample_fmt = encoder->sample_fmts[0];
  174.     enc_ctx->time_base = (AVRational){1, enc_ctx->sample_rate};
  175.     enc_ctx->bit_rate = audioBitRate;
  176.       }
  177.  
  178.       /* Open codec with the set options */
  179.       ret = avcodec_open2(enc_ctx, encoder, &opts);
  180.       if (ret < 0) {
  181.     av_log(NULL, AV_LOG_ERROR, "Cannot open video encoder for stream #%u\n", i);
  182.     return ret;
  183.       }
  184.       int unused = av_dict_count(opts);
  185.       if (unused > 0) {
  186.     av_log(NULL, AV_LOG_WARNING, "%d unused options\n", unused);
  187.       }
  188.     /* } else if (dec_ctx->codec_type == AVMEDIA_TYPE_UNKNOWN) { */
  189.     } else {
  190.       av_log(NULL, AV_LOG_FATAL, "Elementary stream #%d is of unknown type, cannot proceed\n", i);
  191.       return AVERROR_INVALIDDATA;
  192.     } /* else { */
  193.     /*   /\* if this stream must be remuxed *\/ */
  194.     /*   ret = avcodec_copy_context(ofmt_ctx->streams[i]->codec, */
  195.     /*               ifmt_ctx->streams[i]->codec); */
  196.     /*   if (ret < 0) { */
  197.     /*  av_log(NULL, AV_LOG_ERROR, "Copying stream context failed\n"); */
  198.     /*  return ret; */
  199.     /*   } */
  200.     /* } */
  201.  
  202.     if (ofmt_ctx->oformat->flags & AVFMT_GLOBALHEADER)
  203.       enc_ctx->flags |= CODEC_FLAG_GLOBAL_HEADER;
  204.   }
  205.  
  206.   return 0;
  207. }
  208.  
  209. static int open_output_file(const char *filename) {
  210.   int ret;
  211.  
  212.   av_dump_format(ofmt_ctx, 0, filename, 1);
  213.  
  214.   if (!(ofmt_ctx->oformat->flags & AVFMT_NOFILE)) {
  215.     ret = avio_open(&ofmt_ctx->pb, filename, AVIO_FLAG_WRITE);
  216.     if (ret < 0) {
  217.       av_log(NULL, AV_LOG_ERROR, "Could not open output file '%s'", filename);
  218.       return ret;
  219.     }
  220.   }
  221.  
  222.   /* init muxer, write output file header */
  223.   ret = avformat_write_header(ofmt_ctx, NULL);
  224.   if (ret < 0) {
  225.     av_log(NULL, AV_LOG_ERROR, "Error occurred when opening output file\n");
  226.     return ret;
  227.   }
  228.  
  229.   return 0;
  230. }
  231.  
  232. static int init_filter(FilteringContext* fctx, AVCodecContext *dec_ctx,
  233.                AVCodecContext *enc_ctx, const char *filter_spec) {
  234.   char args[512];
  235.   int ret = 0;
  236.   AVFilter *buffersrc = NULL;
  237.   AVFilter *buffersink = NULL;
  238.   AVFilterContext *buffersrc_ctx = NULL;
  239.   AVFilterContext *buffersink_ctx = NULL;
  240.   AVFilterInOut *outputs = avfilter_inout_alloc();
  241.   AVFilterInOut *inputs  = avfilter_inout_alloc();
  242.   AVFilterGraph *filter_graph = avfilter_graph_alloc();
  243.  
  244.   if (!outputs || !inputs || !filter_graph) {
  245.     ret = AVERROR(ENOMEM);
  246.     goto end;
  247.   }
  248.  
  249.   if (dec_ctx->codec_type == AVMEDIA_TYPE_VIDEO) {
  250.     buffersrc = avfilter_get_by_name("buffer");
  251.     buffersink = avfilter_get_by_name("buffersink");
  252.     if (!buffersrc || !buffersink) {
  253.       av_log(NULL, AV_LOG_ERROR, "filtering source or sink element not found\n");
  254.       ret = AVERROR_UNKNOWN;
  255.       goto end;
  256.     }
  257.  
  258.     snprintf(args, sizeof(args),
  259.          "video_size=%dx%d:pix_fmt=%d:time_base=%d/%d:pixel_aspect=%d/%d",
  260.          dec_ctx->width, dec_ctx->height, dec_ctx->pix_fmt,
  261.          dec_ctx->time_base.num, dec_ctx->time_base.den,
  262.          dec_ctx->sample_aspect_ratio.num,
  263.          dec_ctx->sample_aspect_ratio.den);
  264.  
  265.     ret = avfilter_graph_create_filter(&buffersrc_ctx, buffersrc, "in",
  266.                        args, NULL, filter_graph);
  267.     if (ret < 0) {
  268.       av_log(NULL, AV_LOG_ERROR, "Cannot create buffer source\n");
  269.       goto end;
  270.     }
  271.  
  272.     ret = avfilter_graph_create_filter(&buffersink_ctx, buffersink, "out",
  273.                        NULL, NULL, filter_graph);
  274.     if (ret < 0) {
  275.       av_log(NULL, AV_LOG_ERROR, "Cannot create buffer sink\n");
  276.       goto end;
  277.     }
  278.  
  279.     ret = av_opt_set_bin(buffersink_ctx, "pix_fmts",
  280.              (uint8_t*)&enc_ctx->pix_fmt, sizeof(enc_ctx->pix_fmt),
  281.              AV_OPT_SEARCH_CHILDREN);
  282.     if (ret < 0) {
  283.       av_log(NULL, AV_LOG_ERROR, "Cannot set output pixel format\n");
  284.       goto end;
  285.     }
  286.   } else if (dec_ctx->codec_type == AVMEDIA_TYPE_AUDIO) {
  287.     buffersrc = avfilter_get_by_name("abuffer");
  288.     buffersink = avfilter_get_by_name("abuffersink");
  289.     if (!buffersrc || !buffersink) {
  290.       av_log(NULL, AV_LOG_ERROR, "filtering source or sink element not found\n");
  291.       ret = AVERROR_UNKNOWN;
  292.       goto end;
  293.     }
  294.  
  295.     if (!dec_ctx->channel_layout)
  296.       dec_ctx->channel_layout =
  297.     av_get_default_channel_layout(dec_ctx->channels);
  298.     snprintf(args, sizeof(args),
  299.          "time_base=%d/%d:sample_rate=%d:sample_fmt=%s:channel_layout=0x%"PRIx64,
  300.          dec_ctx->time_base.num, dec_ctx->time_base.den, dec_ctx->sample_rate,
  301.          av_get_sample_fmt_name(dec_ctx->sample_fmt),
  302.          dec_ctx->channel_layout);
  303.     ret = avfilter_graph_create_filter(&buffersrc_ctx, buffersrc, "in",
  304.                        args, NULL, filter_graph);
  305.     if (ret < 0) {
  306.       av_log(NULL, AV_LOG_ERROR, "Cannot create audio buffer source\n");
  307.       goto end;
  308.     }
  309.  
  310.     ret = avfilter_graph_create_filter(&buffersink_ctx, buffersink, "out",
  311.                        NULL, NULL, filter_graph);
  312.     if (ret < 0) {
  313.       av_log(NULL, AV_LOG_ERROR, "Cannot create audio buffer sink\n");
  314.       goto end;
  315.     }
  316.  
  317.     ret = av_opt_set_bin(buffersink_ctx, "sample_fmts",
  318.              (uint8_t*)&enc_ctx->sample_fmt, sizeof(enc_ctx->sample_fmt),
  319.              AV_OPT_SEARCH_CHILDREN);
  320.     if (ret < 0) {
  321.       av_log(NULL, AV_LOG_ERROR, "Cannot set output sample format\n");
  322.       goto end;
  323.     }
  324.  
  325.     ret = av_opt_set_bin(buffersink_ctx, "channel_layouts",
  326.              (uint8_t*)&enc_ctx->channel_layout,
  327.              sizeof(enc_ctx->channel_layout), AV_OPT_SEARCH_CHILDREN);
  328.     if (ret < 0) {
  329.       av_log(NULL, AV_LOG_ERROR, "Cannot set output channel layout\n");
  330.       goto end;
  331.     }
  332.  
  333.     ret = av_opt_set_bin(buffersink_ctx, "sample_rates",
  334.              (uint8_t*)&enc_ctx->sample_rate, sizeof(enc_ctx->sample_rate),
  335.              AV_OPT_SEARCH_CHILDREN);
  336.     if (ret < 0) {
  337.       av_log(NULL, AV_LOG_ERROR, "Cannot set output sample rate\n");
  338.       goto end;
  339.     }
  340.   } else {
  341.     ret = AVERROR_UNKNOWN;
  342.     goto end;
  343.   }
  344.  
  345.   /* Endpoints for the filter graph. */
  346.   outputs->name       = av_strdup("in");
  347.   outputs->filter_ctx = buffersrc_ctx;
  348.   outputs->pad_idx    = 0;
  349.   outputs->next       = NULL;
  350.  
  351.   inputs->name       = av_strdup("out");
  352.   inputs->filter_ctx = buffersink_ctx;
  353.   inputs->pad_idx    = 0;
  354.   inputs->next       = NULL;
  355.  
  356.   if (!outputs->name || !inputs->name) {
  357.     ret = AVERROR(ENOMEM);
  358.     goto end;
  359.   }
  360.  
  361.   if ((ret = avfilter_graph_parse_ptr(filter_graph, filter_spec,
  362.                       &inputs, &outputs, NULL)) < 0)
  363.     goto end;
  364.  
  365.   if ((ret = avfilter_graph_config(filter_graph, NULL)) < 0)
  366.     goto end;
  367.  
  368.   /* Fill FilteringContext */
  369.   fctx->buffersrc_ctx = buffersrc_ctx;
  370.   fctx->buffersink_ctx = buffersink_ctx;
  371.   fctx->filter_graph = filter_graph;
  372.  
  373.  end:
  374.   avfilter_inout_free(&inputs);
  375.   avfilter_inout_free(&outputs);
  376.  
  377.   return ret;
  378. }
  379.  
  380. static int init_filters(enum AVCodecID audioCodec) {
  381.   const char *filter_spec;
  382.   unsigned int i;
  383.   int ret;
  384.   filter_ctx = av_malloc_array(ifmt_ctx->nb_streams, sizeof(*filter_ctx));
  385.   if (!filter_ctx)
  386.     return AVERROR(ENOMEM);
  387.  
  388.   for (i = 0; i < ifmt_ctx->nb_streams; i++) {
  389.     filter_ctx[i].buffersrc_ctx  = NULL;
  390.     filter_ctx[i].buffersink_ctx = NULL;
  391.     filter_ctx[i].filter_graph   = NULL;
  392.     /* Skip streams that are neither audio nor video */
  393.     if (!(ifmt_ctx->streams[i]->codec->codec_type == AVMEDIA_TYPE_AUDIO
  394.       || ifmt_ctx->streams[i]->codec->codec_type == AVMEDIA_TYPE_VIDEO))
  395.       continue;
  396.  
  397.  
  398.     if (ifmt_ctx->streams[i]->codec->codec_type == AVMEDIA_TYPE_VIDEO)
  399.       filter_spec = "null"; /* passthrough (dummy) filter for video */
  400.     else
  401.       /* TODO: make this more general */
  402.       if (audioCodec == AV_CODEC_ID_VORBIS) {
  403.     filter_spec = "asetnsamples=n=64";
  404.       } else {
  405.     filter_spec = "null"; /* passthrough (dummy) filter for audio */
  406.     /* filter_spec = "fps=24"; */
  407.     /* filter_spec = "settb=expr=1/24"; */
  408.       }
  409.     ret = init_filter(&filter_ctx[i], ifmt_ctx->streams[i]->codec,
  410.               ofmt_ctx->streams[i]->codec, filter_spec);
  411.     if (ret)
  412.       return ret;
  413.   }
  414.   return 0;
  415. }
  416.  
  417. static int encode_write_frame(AVFrame *filt_frame, unsigned int stream_index, int *got_frame) {
  418.   int ret;
  419.   int got_frame_local;
  420.   AVPacket enc_pkt;
  421.   int (*enc_func)(AVCodecContext *, AVPacket *, const AVFrame *, int *) =
  422.     (ifmt_ctx->streams[stream_index]->codec->codec_type ==
  423.      AVMEDIA_TYPE_VIDEO) ? avcodec_encode_video2 : avcodec_encode_audio2;
  424.  
  425.   if (!got_frame)
  426.     got_frame = &got_frame_local;
  427.  
  428.   /* av_log(NULL, AV_LOG_INFO, "Encoding frame\n"); */
  429.   /* encode filtered frame */
  430.   enc_pkt.data = NULL;
  431.   enc_pkt.size = 0;
  432.   av_init_packet(&enc_pkt);
  433.   ret = enc_func(ofmt_ctx->streams[stream_index]->codec, &enc_pkt,
  434.          filt_frame, got_frame);
  435.   av_frame_free(&filt_frame);
  436.   if (ret < 0)
  437.     return ret;
  438.   if (!(*got_frame))
  439.     return 0;
  440.  
  441.   /* prepare packet for muxing */
  442.   enc_pkt.stream_index = stream_index;
  443.   enc_pkt.dts = av_rescale_q_rnd(enc_pkt.dts,
  444.                  ofmt_ctx->streams[stream_index]->codec->time_base,
  445.                  ofmt_ctx->streams[stream_index]->time_base,
  446.                  AV_ROUND_NEAR_INF|AV_ROUND_PASS_MINMAX);
  447.   enc_pkt.pts = av_rescale_q_rnd(enc_pkt.pts,
  448.                  ofmt_ctx->streams[stream_index]->codec->time_base,
  449.                  ofmt_ctx->streams[stream_index]->time_base,
  450.                  AV_ROUND_NEAR_INF|AV_ROUND_PASS_MINMAX);
  451.   enc_pkt.duration = av_rescale_q(enc_pkt.duration,
  452.                   ofmt_ctx->streams[stream_index]->codec->time_base,
  453.                   ofmt_ctx->streams[stream_index]->time_base);
  454.  
  455.   /* av_log(NULL, AV_LOG_DEBUG, "Muxing frame\n"); */
  456.   /* mux encoded frame */
  457.   ret = av_interleaved_write_frame(ofmt_ctx, &enc_pkt);
  458.   return ret;
  459. }
  460.  
  461. static int filter_encode_write_frame(AVFrame *frame, unsigned int stream_index) {
  462.   int ret;
  463.   AVFrame *filt_frame;
  464.  
  465.   /* av_log(NULL, AV_LOG_INFO, "Pushing decoded frame to filters\n"); */
  466.   /* push the decoded frame into the filtergraph */
  467.   ret = av_buffersrc_add_frame_flags(filter_ctx[stream_index].buffersrc_ctx,
  468.                      frame, 0);
  469.   if (ret < 0) {
  470.     av_log(NULL, AV_LOG_ERROR, "Error while feeding the filtergraph\n");
  471.     return ret;
  472.   }
  473.  
  474.   /* pull filtered frames from the filtergraph */
  475.   while (1) {
  476.     filt_frame = av_frame_alloc();
  477.     if (!filt_frame) {
  478.       ret = AVERROR(ENOMEM);
  479.       break;
  480.     }
  481.     /* av_log(NULL, AV_LOG_INFO, "Pulling filtered frame from filters\n"); */
  482.     ret = av_buffersink_get_frame(filter_ctx[stream_index].buffersink_ctx,
  483.                   filt_frame);
  484.     if (ret < 0) {
  485.       /* if no more frames for output - returns AVERROR(EAGAIN)
  486.        * if flushed and no more frames for output - returns AVERROR_EOF
  487.        * rewrite retcode to 0 to show it as normal procedure completion
  488.        */
  489.       if (ret == AVERROR(EAGAIN) || ret == AVERROR_EOF)
  490.     ret = 0;
  491.       av_frame_free(&filt_frame);
  492.       break;
  493.     }
  494.  
  495.     filt_frame->pict_type = AV_PICTURE_TYPE_NONE;
  496.     ret = encode_write_frame(filt_frame, stream_index, NULL);
  497.     if (ret < 0)
  498.       break;
  499.   }
  500.  
  501.   return ret;
  502. }
  503.  
  504. static int flush_encoder(unsigned int stream_index) {
  505.   int ret;
  506.   int got_frame;
  507.  
  508.   if (!(ofmt_ctx->streams[stream_index]->codec->codec->capabilities &
  509.     CODEC_CAP_DELAY))
  510.     return 0;
  511.  
  512.   while (1) {
  513.     av_log(NULL, AV_LOG_INFO, "Flushing stream #%u encoder\n", stream_index);
  514.     ret = encode_write_frame(NULL, stream_index, &got_frame);
  515.     if (ret < 0)
  516.       break;
  517.     if (!got_frame)
  518.       return 0;
  519.   }
  520.   return ret;
  521. }
  522.  
  523. static int transcode() {
  524.   int ret;
  525.   AVPacket packet = { .data = NULL, .size = 0 };
  526.   AVFrame *frame = NULL;
  527.   enum AVMediaType type;
  528.   unsigned int stream_index;
  529.   unsigned int i;
  530.   int got_frame;
  531.   int (*dec_func)(AVCodecContext *, AVFrame *, int *, const AVPacket *);
  532.  
  533.   /* read all packets */
  534.   while (1) {
  535.     if ((ret = av_read_frame(ifmt_ctx, &packet)) < 0)
  536.       break;
  537.     stream_index = packet.stream_index;
  538.     type = ifmt_ctx->streams[packet.stream_index]->codec->codec_type;
  539.     av_log(NULL, AV_LOG_DEBUG, "Demuxer gave frame of stream_index %u\n",
  540.        stream_index);
  541.  
  542.     if (filter_ctx[stream_index].filter_graph) {
  543.       av_log(NULL, AV_LOG_DEBUG, "Going to reencode&filter the frame\n");
  544.       frame = av_frame_alloc();
  545.       if (!frame) {
  546.     ret = AVERROR(ENOMEM);
  547.     break;
  548.       }
  549.       packet.dts = av_rescale_q_rnd(packet.dts,
  550.                     ifmt_ctx->streams[stream_index]->time_base,
  551.                     ifmt_ctx->streams[stream_index]->codec->time_base,
  552.                     AV_ROUND_NEAR_INF|AV_ROUND_PASS_MINMAX);
  553.       packet.pts = av_rescale_q_rnd(packet.pts,
  554.                     ifmt_ctx->streams[stream_index]->time_base,
  555.                     ifmt_ctx->streams[stream_index]->codec->time_base,
  556.                     AV_ROUND_NEAR_INF|AV_ROUND_PASS_MINMAX);
  557.       dec_func = (type == AVMEDIA_TYPE_VIDEO) ? avcodec_decode_video2 :
  558.     avcodec_decode_audio4;
  559.       ret = dec_func(ifmt_ctx->streams[stream_index]->codec, frame,
  560.              &got_frame, &packet);
  561.       if (ret < 0) {
  562.     av_frame_free(&frame);
  563.     av_log(NULL, AV_LOG_ERROR, "Decoding failed\n");
  564.     break;
  565.       }
  566.  
  567.       if (got_frame) {
  568.     frame->pts = av_frame_get_best_effort_timestamp(frame);
  569.     ret = filter_encode_write_frame(frame, stream_index);
  570.     av_frame_free(&frame);
  571.     if (ret < 0)
  572.       goto end;
  573.       } else {
  574.     av_frame_free(&frame);
  575.       }
  576.     } else {
  577.       /* remux this frame without reencoding */
  578.       packet.dts = av_rescale_q_rnd(packet.dts,
  579.                     ifmt_ctx->streams[stream_index]->time_base,
  580.                     ofmt_ctx->streams[stream_index]->time_base,
  581.                     AV_ROUND_NEAR_INF|AV_ROUND_PASS_MINMAX);
  582.       packet.pts = av_rescale_q_rnd(packet.pts,
  583.                     ifmt_ctx->streams[stream_index]->time_base,
  584.                     ofmt_ctx->streams[stream_index]->time_base,
  585.                     AV_ROUND_NEAR_INF|AV_ROUND_PASS_MINMAX);
  586.  
  587.       ret = av_interleaved_write_frame(ofmt_ctx, &packet);
  588.       if (ret < 0)
  589.     goto end;
  590.     }
  591.     av_free_packet(&packet);
  592.   }
  593.  
  594.   /* flush filters and encoders */
  595.   for (i = 0; i < ifmt_ctx->nb_streams; i++) {
  596.     /* flush filter */
  597.     if (!filter_ctx[i].filter_graph)
  598.       continue;
  599.     ret = filter_encode_write_frame(NULL, i);
  600.     if (ret < 0) {
  601.       av_log(NULL, AV_LOG_ERROR, "Flushing filter failed\n");
  602.       goto end;
  603.     }
  604.  
  605.     /* flush encoder */
  606.     ret = flush_encoder(i);
  607.     if (ret < 0) {
  608.       av_log(NULL, AV_LOG_ERROR, "Flushing encoder failed\n");
  609.       goto end;
  610.     }
  611.   }
  612.  
  613.   av_write_trailer(ofmt_ctx);
  614.  
  615.   // Retrieve and store the first instance of codec statistics
  616.   // TODO: less naive, deal with multiple instances of statistics
  617.   for (i = 0; i < ofmt_ctx->nb_streams; i++) {
  618.     AVCodecContext* codec = ofmt_ctx->streams[i]->codec;
  619.     if ((codec->flags & CODEC_FLAG_PASS1) && (codec->stats_out)){
  620.     FILE* logfile = fopen(STATS_LOG, "wb");
  621.     fprintf(logfile, "%s", codec->stats_out);
  622.     if (fclose(logfile) < 0) {
  623.       av_log(NULL, AV_LOG_ERROR, "Error closing log file.\n");
  624.     }
  625.     break;
  626.       }
  627.   }
  628.  
  629.   av_log(NULL, AV_LOG_INFO, "output duration = %" PRId64 "\n", ofmt_ctx->duration);
  630.  
  631.  end:
  632.   av_free_packet(&packet);
  633.   av_frame_free(&frame);
  634.   for (i = 0; i < ifmt_ctx->nb_streams; i++) {
  635.     avcodec_close(ifmt_ctx->streams[i]->codec);
  636.     if (ofmt_ctx && ofmt_ctx->nb_streams > i && ofmt_ctx->streams[i] && ofmt_ctx->streams[i]->codec)
  637.       avcodec_close(ofmt_ctx->streams[i]->codec);
  638.     if (filter_ctx && filter_ctx[i].filter_graph)
  639.       avfilter_graph_free(&filter_ctx[i].filter_graph);
  640.   }
  641.   av_free(filter_ctx);
  642.   avformat_close_input(&ifmt_ctx);
  643.   if (ofmt_ctx && !(ofmt_ctx->oformat->flags & AVFMT_NOFILE))
  644.     avio_close(ofmt_ctx->pb);
  645.   avformat_free_context(ofmt_ctx);
  646.  
  647.   if (ret < 0)
  648.     av_log(NULL, AV_LOG_ERROR, "Error occurred: %s\n", av_err2str(ret));
  649.  
  650.   return ret ? 1 : 0;
  651. }
  652.  
  653. int TranscodeToWebM(char* inputPath, char* outputPath, int audioBitRate, int crf, int videoMaxBitRate, int threads,
  654.             char* quality, int speed) {
  655.   int ret;
  656.   unsigned int pass;
  657.   char* stats = NULL;
  658.  
  659.   av_register_all();
  660.   avfilter_register_all();
  661.  
  662.   for (pass = 1; pass <= 2; pass++) {
  663.     if ((ret = open_input_file(inputPath)) < 0)
  664.       goto end;
  665.  
  666.     if ((ret = init_output_context(outputPath)) < 0)
  667.       goto end;
  668.  
  669.     if (pass == 2) {
  670.       size_t stats_length;
  671.       if (cmdutils_read_file(STATS_LOG, &stats, &stats_length) < 0) {
  672.     av_log(NULL, AV_LOG_ERROR, "Error reading stats file.\n");
  673.     break;
  674.       }
  675.     }
  676.  
  677.     if ((ret = init_webm_encoders(audioBitRate, crf, videoMaxBitRate, threads, quality, speed, pass, stats)) < 0)
  678.       goto end;
  679.  
  680.     if ((ret = open_output_file(outputPath)) < 0)
  681.       goto end;
  682.  
  683.     if ((ret = init_filters(AV_CODEC_ID_VORBIS)) < 0)
  684.       goto end;
  685.  
  686.     if ((ret = transcode()) < 0)
  687.       goto end;
  688.   }
  689.  
  690.   if (remove(STATS_LOG) != 0) {
  691.     av_log(NULL, AV_LOG_ERROR, "Failed to remove %s\n", STATS_LOG);
  692.   }
  693.  
  694.  end:
  695.   if (ret < 0) {
  696.     av_log(NULL, AV_LOG_ERROR, "Error occurred: %s\n", av_err2str(ret));
  697.     return ret;
  698.   }
  699.  
  700.   return 0;
  701. }
Advertisement
Add Comment
Please, Sign In to add comment
Advertisement
Public Pastes
Advertisement
We use cookies for various purposes including analytics. By continuing to use Pastebin, you agree to our use of cookies as described in the Cookies Policy.  OK, I Understand
Not a member of Pastebin yet?
Sign Up, it unlocks many cool features!