From: Fabrice Bellard Date: Mon, 10 Nov 2003 18:54:17 +0000 (+0000) Subject: use av_read_frame() - added initial seek support (rename -start to -ss (mplayer.... X-Git-Tag: v0.5~15700 X-Git-Url: http://review.tizen.org/git/?a=commitdiff_plain;h=254abc2ec6b6d0bff590bf113095debd5a03687e;p=platform%2Fupstream%2Flibav.git use av_read_frame() - added initial seek support (rename -start to -ss (mplayer...)) - added '-dump' option to dump input packets Originally committed as revision 2504 to svn://svn.ffmpeg.org/ffmpeg/trunk --- diff --git a/ffmpeg.c b/ffmpeg.c index ebe0688..a802bc4 100644 --- a/ffmpeg.c +++ b/ffmpeg.c @@ -143,6 +143,7 @@ static char *str_copyright = NULL; static char *str_comment = NULL; static int do_benchmark = 0; static int do_hex_dump = 0; +static int do_pkt_dump = 0; static int do_psnr = 0; static int do_vstats = 0; static int do_pass = 0; @@ -202,11 +203,11 @@ typedef struct AVInputStream { int discard; /* true if stream data should be discarded */ int decoding_needed; /* true if the packets must be decoded in 'raw_fifo' */ int64_t sample_index; /* current sample */ - int frame_decoded; /* true if a video or audio frame has been decoded */ int64_t start; /* time when read started */ unsigned long frame; /* current frame */ - AVFrac next_pts; /* synthetic pts for cases where pkt.pts == 0 */ + int64_t next_pts; /* synthetic pts for cases where pkt.pts + is not defined */ int64_t pts; /* current pts */ } AVInputStream; @@ -518,15 +519,17 @@ static void do_video_out(AVFormatContext *s, } #if defined(AVSYNC_DEBUG) - static char *action[] = { "drop frame", "copy frame", "dup frame" }; - if (audio_sync) - fprintf(stderr, "Input APTS %12.6f, output APTS %12.6f, ", - (double) audio_sync->sync_ipts, - (double) audio_sync->st->pts.val * s->pts_num / s->pts_den); - fprintf(stderr, "Input VPTS %12.6f, output VPTS %12.6f: %s\n", - (double) ost->sync_ipts, - (double) ost->st->pts.val * s->pts_num / s->pts_den, - action[nb_frames]); + { + static char *action[] = { "drop frame", "copy frame", "dup frame" }; + if (audio_sync) + fprintf(stderr, "Input APTS %12.6f, output APTS %12.6f, ", + (double) audio_sync->sync_ipts, + (double) audio_sync->st->pts.val * s->pts_num / s->pts_den); + fprintf(stderr, "Input VPTS %12.6f, output VPTS %12.6f: %s\n", + (double) ost->sync_ipts, + (double) ost->st->pts.val * s->pts_num / s->pts_den, + action[nb_frames]); + } #endif if (nb_frames <= 0) @@ -1139,7 +1142,6 @@ static int av_encode(AVFormatContext **output_files, } //if (ist->st->codec.codec_type == CODEC_TYPE_VIDEO) // ist->st->codec.flags |= CODEC_FLAG_REPEAT_FIELD; - ist->frame_decoded = 1; } } @@ -1148,20 +1150,7 @@ static int av_encode(AVFormatContext **output_files, ist = ist_table[i]; is = input_files[ist->file_index]; ist->pts = 0; - if (ist->decoding_needed) { - switch (ist->st->codec.codec_type) { - case CODEC_TYPE_AUDIO: - av_frac_init(&ist->next_pts, - 0, 0, (uint64_t)is->pts_num * ist->st->codec.sample_rate); - break; - case CODEC_TYPE_VIDEO: - av_frac_init(&ist->next_pts, - 0, 0, (uint64_t)is->pts_num * ist->st->codec.frame_rate); - break; - default: - break; - } - } + ist->next_pts = 0; } /* compute buffer size max (should use a complete heuristic) */ @@ -1234,9 +1223,9 @@ static int av_encode(AVFormatContext **output_files, if (recording_time > 0 && pts_min >= (recording_time / 1000000.0)) break; - /* read a packet from it and output it in the fifo */ + /* read a frame from it and output it in the fifo */ is = input_files[file_index]; - if (av_read_packet(is, &pkt) < 0) { + if (av_read_frame(is, &pkt) < 0) { file_table[file_index].eof_reached = 1; continue; } @@ -1245,9 +1234,8 @@ static int av_encode(AVFormatContext **output_files, } else { stream_no_data = 0; } - if (do_hex_dump) { - printf("stream #%d, size=%d:\n", pkt.stream_index, pkt.size); - av_hex_dump(pkt.data, pkt.size); + if (do_pkt_dump) { + av_pkt_dump(stdout, &pkt, do_hex_dump); } /* the following test is needed in case new streams appear dynamically in stream : we ignore them */ @@ -1259,32 +1247,20 @@ static int av_encode(AVFormatContext **output_files, goto discard_packet; //fprintf(stderr,"read #%d.%d size=%d\n", ist->file_index, ist->index, pkt.size); + + if (pkt.pts != AV_NOPTS_VALUE) { + ist->pts = pkt.pts; + } else { + ist->pts = ist->next_pts; + } len = pkt.size; ptr = pkt.data; while (len > 0) { - int ipts; - /* decode the packet if needed */ + /* decode the packet if needed */ data_buf = NULL; /* fail safe */ data_size = 0; if (ist->decoding_needed) { - /* NOTE1: we only take into account the PTS if a new - frame has begun (MPEG semantics) */ - /* NOTE2: even if the fraction is not initialized, - av_frac_set can be used to set the integer part */ - if (ist->frame_decoded) { - /* If pts is unavailable -- we have to use synthetic one */ - if( pkt.pts != AV_NOPTS_VALUE ) - { - ist->pts = ist->next_pts.val = pkt.pts; - } - else - { - ist->pts = ist->next_pts.val; - } - ist->frame_decoded = 0; - } - switch(ist->st->codec.codec_type) { case CODEC_TYPE_AUDIO: /* XXX: could avoid copy if PCM 16 bits with same @@ -1293,17 +1269,17 @@ static int av_encode(AVFormatContext **output_files, ptr, len); if (ret < 0) goto fail_decode; + ptr += ret; + len -= ret; /* Some bug in mpeg audio decoder gives */ /* data_size < 0, it seems they are overflows */ if (data_size <= 0) { /* no audio frame */ - ptr += ret; - len -= ret; continue; } data_buf = (uint8_t *)samples; - av_frac_add(&ist->next_pts, - (uint64_t)is->pts_den * data_size / (2 * ist->st->codec.channels)); + ist->next_pts += ((int64_t)AV_TIME_BASE * data_size) / + (2 * ist->st->codec.channels); break; case CODEC_TYPE_VIDEO: { @@ -1323,12 +1299,14 @@ static int av_encode(AVFormatContext **output_files, } if (!got_picture) { /* no picture yet */ - ptr += ret; - len -= ret; - continue; + goto discard_packet; } - av_frac_add(&ist->next_pts, - (uint64_t)is->pts_den * ist->st->codec.frame_rate_base); + if (ist->st->codec.frame_rate_base != 0) { + ist->next_pts += ((int64_t)AV_TIME_BASE * + ist->st->codec.frame_rate_base) / + ist->st->codec.frame_rate; + } + len = 0; } break; default: @@ -1338,17 +1316,14 @@ static int av_encode(AVFormatContext **output_files, data_buf = ptr; data_size = len; ret = len; + len = 0; } - ptr += ret; - len -= ret; - buffer_to_free = 0; + buffer_to_free = NULL; if (ist->st->codec.codec_type == CODEC_TYPE_VIDEO) { pre_process_video_frame(ist, &picture, &buffer_to_free); } - ist->frame_decoded = 1; - /* frame rate emulation */ if (ist->st->codec.rate_emu) { int64_t pts = av_rescale((int64_t) ist->frame * ist->st->codec.frame_rate_base, 1000000, ist->st->codec.frame_rate); @@ -1372,75 +1347,73 @@ static int av_encode(AVFormatContext **output_files, } } #endif - ipts = (double)ist->pts * is->pts_num / is->pts_den; - //fprintf(stderr,"decoded ipts=%ld %d\n",ipts,ist->pts); - - /* if output time reached then transcode raw format, + /* if output time reached then transcode raw format, encode packets and output them */ - if (start_time == 0 || ipts > (start_time / 1000000.0)) - for(i=0;ipts >= start_time) + for(i=0;isource_index == ist_index) { - os = output_files[ost->file_index]; + ost = ost_table[i]; + if (ost->source_index == ist_index) { + os = output_files[ost->file_index]; #if 0 - printf("%d: got pts=%f %f\n", i, pkt.pts / 90000.0, - (ist->pts - ost->st->pts.val) / 90000.0); + printf("%d: got pts=%0.3f %0.3f\n", i, + (double)pkt.pts / AV_TIME_BASE, + ((double)ist->pts / AV_TIME_BASE) - + ((double)ost->st->pts.val * os->pts_num / os->pts_den)); #endif - /* set the input output pts pairs */ - ost->sync_ipts = (double)ist->pts * is->pts_num / - is->pts_den; - /* XXX: take into account the various fifos, - in particular for audio */ - ost->sync_opts = ost->st->pts.val; - //printf("ipts=%lld sync_ipts=%f sync_opts=%lld pts.val=%lld pkt.pts=%lld\n", ist->pts, ost->sync_ipts, ost->sync_opts, ost->st->pts.val, pkt.pts); - - if (ost->encoding_needed) { - switch(ost->st->codec.codec_type) { - case CODEC_TYPE_AUDIO: - do_audio_out(os, ost, ist, data_buf, data_size); - break; - case CODEC_TYPE_VIDEO: - /* find an audio stream for synchro */ - { - int i; - AVOutputStream *audio_sync, *ost1; - audio_sync = NULL; - for(i=0;ifile_index == ost->file_index && - ost1->st->codec.codec_type == CODEC_TYPE_AUDIO) { - audio_sync = ost1; - break; + /* set the input output pts pairs */ + ost->sync_ipts = (double)ist->pts / AV_TIME_BASE; + /* XXX: take into account the various fifos, + in particular for audio */ + ost->sync_opts = ost->st->pts.val; + //printf("ipts=%lld sync_ipts=%f sync_opts=%lld pts.val=%lld pkt.pts=%lld\n", ist->pts, ost->sync_ipts, ost->sync_opts, ost->st->pts.val, pkt.pts); + + if (ost->encoding_needed) { + switch(ost->st->codec.codec_type) { + case CODEC_TYPE_AUDIO: + do_audio_out(os, ost, ist, data_buf, data_size); + break; + case CODEC_TYPE_VIDEO: + /* find an audio stream for synchro */ + { + int i; + AVOutputStream *audio_sync, *ost1; + audio_sync = NULL; + for(i=0;ifile_index == ost->file_index && + ost1->st->codec.codec_type == CODEC_TYPE_AUDIO) { + audio_sync = ost1; + break; + } } - } - do_video_out(os, ost, ist, &picture, &frame_size, audio_sync); - if (do_vstats && frame_size) - do_video_stats(os, ost, frame_size); + do_video_out(os, ost, ist, &picture, &frame_size, audio_sync); + if (do_vstats && frame_size) + do_video_stats(os, ost, frame_size); + } + break; + default: + av_abort(); } - break; - default: - av_abort(); - } - } else { - AVFrame avframe; + } else { + AVFrame avframe; - /* no reencoding needed : output the packet directly */ - /* force the input stream PTS */ + /* no reencoding needed : output the packet directly */ + /* force the input stream PTS */ - memset(&avframe, 0, sizeof(AVFrame)); - ost->st->codec.coded_frame= &avframe; - avframe.key_frame = pkt.flags & PKT_FLAG_KEY; + memset(&avframe, 0, sizeof(AVFrame)); + ost->st->codec.coded_frame= &avframe; + avframe.key_frame = pkt.flags & PKT_FLAG_KEY; - av_write_frame(os, ost->index, data_buf, data_size); - ost->st->codec.frame_number++; - ost->frame_number++; + av_write_frame(os, ost->index, data_buf, data_size); + ost->st->codec.frame_number++; + ost->frame_number++; + } } } - } av_free(buffer_to_free); } discard_packet: @@ -2079,6 +2052,23 @@ static void opt_input_file(const char *filename) exit(1); } + /* if seeking requested, we execute it */ + if (start_time != 0) { + int64_t timestamp; + + timestamp = start_time; + /* add the stream start time */ + if (ic->start_time != AV_NOPTS_VALUE) + timestamp += ic->start_time; + ret = av_seek_frame(ic, -1, timestamp); + if (ret < 0) { + fprintf(stderr, "%s: could not seek to position %0.3f\n", + filename, (double)timestamp / AV_TIME_BASE); + } + /* reset seek info */ + start_time = 0; + } + /* update the current parameters so that they match the one of the input stream */ for(i=0;inb_streams;i++) { AVCodecContext *enc = &ic->streams[i]->codec; @@ -2100,11 +2090,6 @@ static void opt_input_file(const char *filename) enc->error_concealment = error_concealment; enc->idct_algo= idct_algo; enc->debug= debug; -/* if(enc->codec->capabilities & CODEC_CAP_TRUNCATED) - enc->flags|= CODEC_FLAG_TRUNCATED; */ - if(/*enc->codec_id==CODEC_ID_MPEG4 || */enc->codec_id==CODEC_ID_MPEG1VIDEO || enc->codec_id==CODEC_ID_H264) - enc->flags|= CODEC_FLAG_TRUNCATED; - if(bitexact) enc->flags|= CODEC_FLAG_BITEXACT; @@ -2120,6 +2105,8 @@ static void opt_input_file(const char *filename) enc->rate_emu = rate_emu; break; + case CODEC_TYPE_DATA: + break; default: av_abort(); } @@ -2824,7 +2811,7 @@ const OptionDef options[] = { { "y", OPT_BOOL, {(void*)&file_overwrite}, "overwrite output files" }, { "map", HAS_ARG | OPT_EXPERT, {(void*)opt_map}, "set input stream mapping", "file:stream" }, { "t", HAS_ARG, {(void*)opt_recording_time}, "set the recording time", "duration" }, - { "start", HAS_ARG, {(void*)opt_start_time}, "set the start time offset", "time_off" }, + { "ss", HAS_ARG, {(void*)opt_start_time}, "set the start time offset", "time_off" }, { "title", HAS_ARG | OPT_STRING, {(void*)&str_title}, "set the title", "string" }, { "author", HAS_ARG | OPT_STRING, {(void*)&str_author}, "set the author", "string" }, { "copyright", HAS_ARG | OPT_STRING, {(void*)&str_copyright}, "set the copyright", "string" }, @@ -2832,8 +2819,10 @@ const OptionDef options[] = { { "debug", HAS_ARG | OPT_EXPERT, {(void*)opt_debug}, "print specific debug info", "" }, { "benchmark", OPT_BOOL | OPT_EXPERT, {(void*)&do_benchmark}, "add timings for benchmarking" }, - { "hex", OPT_BOOL | OPT_EXPERT, {(void*)&do_hex_dump}, + { "dump", OPT_BOOL | OPT_EXPERT, {(void*)&do_pkt_dump}, "dump each input packet" }, + { "hex", OPT_BOOL | OPT_EXPERT, {(void*)&do_hex_dump}, + "when dumping packets, also dump the payload" }, { "bitexact", OPT_EXPERT, {(void*)opt_bitexact}, "only use bit exact algorithms (for codec testing)" }, { "re", OPT_BOOL | OPT_EXPERT, {(void*)&rate_emu}, "read input at native frame rate", "" }, { "loop", OPT_BOOL | OPT_EXPERT, {(void*)&loop_input}, "loop (current only works with images)" },