对pts、dts、duration的处理主要集中在两大函数里面
1、process_input()读入数据并处理,放到滤镜里面
2、reap_filters()从滤镜读出数据,处理后写入文件

process_input()中对时间的处理

获取数据包

ret = get_input_packet(ifile, &pkt);
1
1
通过这条语句实际里面调用

return av_read_frame(f->ctx, pkt);
1
1
将读取到的帧放到pkt中。

处理获取的pkt中的时间

//如果pkt.dts或pkt.pts有值,而不是AV_NOPTS_VALUE时,做处理。
/*
ifile->ts_offset在open_input_file函数中有处理,涉及到参数"copyts"、"start_at_zero"、"itsoffset"、"ss"
语句如下。
f->ts_offset = o->input_ts_offset - (copy_ts ? (start_at_zero && ic->start_time != AV_NOPTS_VALUE ? ic->start_time : 0) : timestamp);
如果没有这些设置。f->ts_offset=0
*/
if (pkt.dts != AV_NOPTS_VALUE)
pkt.dts += av_rescale_q(ifile->ts_offset, AV_TIME_BASE_Q, ist->st->time_base);
if (pkt.pts != AV_NOPTS_VALUE)
pkt.pts += av_rescale_q(ifile->ts_offset, AV_TIME_BASE_Q, ist->st->time_base);

//ist->ts_scale在add_input_streams函数中设置
/*
ist->ts_scale = 1.0;//默认是1
MATCH_PER_STREAM_OPT(ts_scale, dbl, ist->ts_scale, ic, st);
涉及到的参数是"itsscale":解释为"set the input ts scale"
貌似是单位
*/
if (pkt.pts != AV_NOPTS_VALUE)
pkt.pts *= ist->ts_scale;
if (pkt.dts != AV_NOPTS_VALUE)
pkt.dts *= ist->ts_scale;

//做时间基的转行
pkt_dts = av_rescale_q_rnd(pkt.dts, ist->st->time_base, AV_TIME_BASE_Q, (enum AVRounding)(AV_ROUND_NEAR_INF|AV_ROUND_PASS_MINMAX));
if ((ist->dec_ctx->codec_type == AVMEDIA_TYPE_VIDEO ||
ist->dec_ctx->codec_type == AVMEDIA_TYPE_AUDIO) &&
pkt_dts != AV_NOPTS_VALUE && ist->next_dts == AV_NOPTS_VALUE && !copy_ts
&& (is->iformat->flags & AVFMT_TS_DISCONT) && ifile->last_ts != AV_NOPTS_VALUE) {
int64_t delta = pkt_dts - ifile->last_ts;
if (delta < -1LL*dts_delta_threshold*AV_TIME_BASE ||
delta > 1LL*dts_delta_threshold*AV_TIME_BASE){
ifile->ts_offset -= delta;
av_log(NULL, AV_LOG_DEBUG,
"Inter stream timestamp discontinuity %"PRId64", new offset= %"PRId64"\n",
delta, ifile->ts_offset);
pkt.dts -= av_rescale_q(delta, AV_TIME_BASE_Q, ist->st->time_base);
if (pkt.pts != AV_NOPTS_VALUE)
pkt.pts -= av_rescale_q(delta, AV_TIME_BASE_Q, ist->st->time_base);
}
}

duration = av_rescale_q(ifile->duration, ifile->time_base, ist->st->time_base);
if (pkt.pts != AV_NOPTS_VALUE) {
pkt.pts += duration;
ist->max_pts = FFMAX(pkt.pts, ist->max_pts);
ist->min_pts = FFMIN(pkt.pts, ist->min_pts);
}

if (pkt.dts != AV_NOPTS_VALUE)
pkt.dts += duration;

pkt_dts = av_rescale_q_rnd(pkt.dts, ist->st->time_base, AV_TIME_BASE_Q, (enum AVRounding)(AV_ROUND_NEAR_INF|AV_ROUND_PASS_MINMAX));
if ((ist->dec_ctx->codec_type == AVMEDIA_TYPE_VIDEO ||
ist->dec_ctx->codec_type == AVMEDIA_TYPE_AUDIO) &&
pkt_dts != AV_NOPTS_VALUE && ist->next_dts != AV_NOPTS_VALUE &&
!copy_ts) {
int64_t delta = pkt_dts - ist->next_dts;
if (is->iformat->flags & AVFMT_TS_DISCONT) {
if (delta < -1LL*dts_delta_threshold*AV_TIME_BASE ||
delta > 1LL*dts_delta_threshold*AV_TIME_BASE ||
pkt_dts + AV_TIME_BASE/10 < FFMAX(ist->pts, ist->dts)) {
ifile->ts_offset -= delta;
av_log(NULL, AV_LOG_DEBUG,
"timestamp discontinuity %"PRId64", new offset= %"PRId64"\n",
delta, ifile->ts_offset);
pkt.dts -= av_rescale_q(delta, AV_TIME_BASE_Q, ist->st->time_base);
if (pkt.pts != AV_NOPTS_VALUE)
pkt.pts -= av_rescale_q(delta, AV_TIME_BASE_Q, ist->st->time_base);
}
} else {
if ( delta < -1LL*dts_error_threshold*AV_TIME_BASE ||
delta > 1LL*dts_error_threshold*AV_TIME_BASE) {
av_log(NULL, AV_LOG_WARNING, "DTS %"PRId64", next:%"PRId64" st:%d invalid dropping\n", pkt.dts, ist->next_dts, pkt.stream_index);
pkt.dts = AV_NOPTS_VALUE;
}
if (pkt.pts != AV_NOPTS_VALUE){
int64_t pkt_pts = av_rescale_q(pkt.pts, ist->st->time_base, AV_TIME_BASE_Q);
delta = pkt_pts - ist->next_dts;
if ( delta < -1LL*dts_error_threshold*AV_TIME_BASE ||
delta > 1LL*dts_error_threshold*AV_TIME_BASE) {
av_log(NULL, AV_LOG_WARNING, "PTS %"PRId64", next:%"PRId64" invalid dropping st:%d\n", pkt.pts, ist->next_dts, pkt.stream_index);
pkt.pts = AV_NOPTS_VALUE;
}
}
}
}

if (pkt.dts != AV_NOPTS_VALUE)
ifile->last_ts = av_rescale_q(pkt.dts, ist->st->time_base, AV_TIME_BASE_Q);
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
如果pkt.dts == AV_NOPTS_VALUE 并且pkt.dts == AV_NOPTS_VALUE,以上这些语句都没有起作用。

之后调用如下语句

process_input_packet(ist, &pkt, 0);
1
1
process_input_packet

/* pkt = NULL means EOF (needed to flush decoder buffers) */
static int process_input_packet(InputStream *ist, const AVPacket *pkt, int no_eof)
{
int ret = 0, i;
int got_output = 0;

AVPacket avpkt;
//如果是文件的第一次读取,就处理
if (!ist->saw_first_ts) {
//有B帧的情况下dts和pts是不同的。
ist->dts = ist->st->avg_frame_rate.num ? - ist->dec_ctx->has_b_frames * AV_TIME_BASE / av_q2d(ist->st->avg_frame_rate) : 0;
ist->pts = 0;
if (pkt && pkt->pts != AV_NOPTS_VALUE && !ist->decoding_needed) {
ist->dts += av_rescale_q(pkt->pts, ist->st->time_base, AV_TIME_BASE_Q);
ist->pts = ist->dts; //unused but better to set it to a value thats not totally wrong
}
ist->saw_first_ts = 1;
}
//此时在ist中dts和pts都是有值的,第一次进来是初始值,之后进来的值
//初始化next_dts和next_pts
if (ist->next_dts == AV_NOPTS_VALUE)
ist->next_dts = ist->dts;
if (ist->next_pts == AV_NOPTS_VALUE)
ist->next_pts = ist->pts;
/*如果pkt为NULL,表示流结束了。给avpkt赋值为此后在调用decode_video()->avcodec_decode_video2()时会将剩余的帧都返回。
在process_input()函数的出错处理里面有调用process_input_packet(ist, NULL, 0);就在此处处理
*/
if (!pkt) {
/* EOF handling */
av_init_packet(&avpkt);
avpkt.data = NULL;
avpkt.size = 0;
goto handle_eof;
} else {
avpkt = *pkt;
}

if (pkt->dts != AV_NOPTS_VALUE) {
ist->next_dts = ist->dts = av_rescale_q(pkt->dts, ist->st->time_base, AV_TIME_BASE_Q);
if (ist->dec_ctx->codec_type != AVMEDIA_TYPE_VIDEO || !ist->decoding_needed)
ist->next_pts = ist->pts = ist->dts;
}

// while we have more to decode or while the decoder did output something on EOF
/*pkt里为视频是为一帧,为音频是可能为多帧。所以要一个循环*/
while (ist->decoding_needed && (avpkt.size > 0 || (!pkt && got_output))) {
int duration;
handle_eof:
//
ist->pts = ist->next_pts;
ist->dts = ist->next_dts;

//没什么用。
if (avpkt.size && avpkt.size != pkt->size &&
!(ist->dec->capabilities & AV_CODEC_CAP_SUBFRAMES)) {
av_log(NULL, ist->showed_multi_packet_warning ? AV_LOG_VERBOSE : AV_LOG_WARNING,
"Multiple frames in a packet from stream %d\n", pkt->stream_index);
ist->showed_multi_packet_warning = 1;
}
//解码一个pkt
switch (ist->dec_ctx->codec_type) {
case AVMEDIA_TYPE_AUDIO:
ret = decode_audio (ist, &avpkt, &got_output);
break;
case AVMEDIA_TYPE_VIDEO:
ret = decode_video (ist, &avpkt, &got_output);
if (avpkt.duration) {
//返回的pkt中已经计算出了时长
duration = av_rescale_q(avpkt.duration, ist->st->time_base, AV_TIME_BASE_Q);
} else if(ist->dec_ctx->framerate.num != 0 && ist->dec_ctx->framerate.den != 0) {
//根据帧率计算时长
int ticks= av_stream_get_parser(ist->st) ? av_stream_get_parser(ist->st)->repeat_pict+1 : ist->dec_ctx->ticks_per_frame;
duration = ((int64_t)AV_TIME_BASE *
ist->dec_ctx->framerate.den * ticks) /
ist->dec_ctx->framerate.num / ist->dec_ctx->ticks_per_frame;
} else
duration = 0;
//dts和pts加时长,就是下一个帧的dts和pts,注意pkt中的都是临时的。只是在解码的时候用,过程当中的时间都保持在ist中。
if(ist->dts != AV_NOPTS_VALUE && duration) {
ist->next_dts += duration;
}else
ist->next_dts = AV_NOPTS_VALUE;

if (got_output)
ist->next_pts += duration; //FIXME the duration is not correct in some cases
break;
case AVMEDIA_TYPE_SUBTITLE:
ret = transcode_subtitles(ist, &avpkt, &got_output);
break;
default:
return -1;
}

if (ret < 0) {
av_log(NULL, AV_LOG_ERROR, "Error while decoding stream #%d:%d: %s\n",
ist->file_index, ist->st->index, av_err2str(ret));
if (exit_on_error)
exit_program(1);
break;
}

avpkt.dts=
avpkt.pts= AV_NOPTS_VALUE;

// touch data and size only if not EOF
if (pkt) {
if(ist->dec_ctx->codec_type != AVMEDIA_TYPE_AUDIO)
ret = avpkt.size;
avpkt.data += ret;
avpkt.size -= ret;
}
if (!got_output) {
continue;
}
if (got_output && !pkt)
break;
}
//pkt 传入是NULL时,给滤镜传入EOF信息
/* after flushing, send an EOF on all the filter inputs attached to the stream */
/* except when looping we need to flush but not to send an EOF */
if (!pkt && ist->decoding_needed && !got_output && !no_eof) {
int ret = send_filter_eof(ist);
if (ret < 0) {
av_log(NULL, AV_LOG_FATAL, "Error marking filters as finished\n");
exit_program(1);
}
}

/* handle stream copy */
if (!ist->decoding_needed) {
ist->dts = ist->next_dts;
switch (ist->dec_ctx->codec_type) {
case AVMEDIA_TYPE_AUDIO:
ist->next_dts += ((int64_t)AV_TIME_BASE * ist->dec_ctx->frame_size) /
ist->dec_ctx->sample_rate;
break;
case AVMEDIA_TYPE_VIDEO:
if (ist->framerate.num) {
// TODO: Remove work-around for c99-to-c89 issue 7
AVRational time_base_q = AV_TIME_BASE_Q;
int64_t next_dts = av_rescale_q(ist->next_dts, time_base_q, av_inv_q(ist->framerate));
ist->next_dts = av_rescale_q(next_dts + 1, av_inv_q(ist->framerate), time_base_q);
} else if (pkt->duration) {
ist->next_dts += av_rescale_q(pkt->duration, ist->st->time_base, AV_TIME_BASE_Q);
} else if(ist->dec_ctx->framerate.num != 0) {
int ticks= av_stream_get_parser(ist->st) ? av_stream_get_parser(ist->st)->repeat_pict + 1 : ist->dec_ctx->ticks_per_frame;
ist->next_dts += ((int64_t)AV_TIME_BASE *
ist->dec_ctx->framerate.den * ticks) /
ist->dec_ctx->framerate.num / ist->dec_ctx->ticks_per_frame;
}
break;
}
ist->pts = ist->dts;
ist->next_pts = ist->next_dts;
}
for (i = 0; pkt && i < nb_output_streams; i++) {
OutputStream *ost = output_streams[i];

if (!check_output_constraints(ist, ost) || ost->encoding_needed)
continue;
//如果不需要重新编解码,就走到这里。ost->encoding_needed判断。
do_streamcopy(ist, ost, pkt);
}

return got_output;
}
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
decode_video

针对时间的处理只有一句

pkt->dts = av_rescale_q(ist->dts, AV_TIME_BASE_Q, ist->st->time_base);
1
1
将在ist中保存的dts给pkt中

static int decode_video(InputStream *ist, AVPacket *pkt, int *got_output)
{
AVFrame *decoded_frame, *f;
int i, ret = 0, err = 0, resample_changed;
int64_t best_effort_timestamp;
AVRational *frame_sample_aspect;

if (!ist->decoded_frame && !(ist->decoded_frame = av_frame_alloc()))
return AVERROR(ENOMEM);
if (!ist->filter_frame && !(ist->filter_frame = av_frame_alloc()))
return AVERROR(ENOMEM);
//解码后的帧存放处
decoded_frame = ist->decoded_frame;
//开始给pkt中的时间赋值。只对dts赋值。
pkt->dts = av_rescale_q(ist->dts, AV_TIME_BASE_Q, ist->st->time_base);

update_benchmark(NULL);
//实际的解码函数,got_output不为0表示解码出了帧,解码出的帧放到decoded_frame中。
ret = avcodec_decode_video2(ist->dec_ctx,
decoded_frame, got_output, pkt);
update_benchmark("decode_video %d.%d", ist->file_index, ist->st->index);

// The following line may be required in some cases where there is no parser
// or the parser does not has_b_frames correctly
if (ist->st->codec->has_b_frames < ist->dec_ctx->has_b_frames) {
if (ist->dec_ctx->codec_id == AV_CODEC_ID_H264) {
ist->st->codec->has_b_frames = ist->dec_ctx->has_b_frames;
} else
av_log(ist->dec_ctx, AV_LOG_WARNING,
"has_b_frames is larger in decoder than demuxer %d > %d.\n"
"If you want to help, upload a sample "
"of this file to ftp://upload.ffmpeg.org/incoming/ "
"and contact the ffmpeg-devel mailing list. (ffmpeg-devel@ffmpeg.org)",
ist->dec_ctx->has_b_frames,
ist->st->codec->has_b_frames);
}
//判断输出结果,涉及到参数"xerror"。
check_decode_result(ist, got_output, ret);
//如果解码返回成功,做个判断,看是否解码后的长宽、像素格式和最初分析的一样,不一样就输出。没啥用
if (*got_output && ret >= 0) {
if (ist->dec_ctx->width != decoded_frame->width ||
ist->dec_ctx->height != decoded_frame->height ||
ist->dec_ctx->pix_fmt != decoded_frame->format) {
av_log(NULL, AV_LOG_DEBUG, "Frame parameters mismatch context %d,%d,%d != %d,%d,%d\n",
decoded_frame->width,
decoded_frame->height,
decoded_frame->format,
ist->dec_ctx->width,
ist->dec_ctx->height,
ist->dec_ctx->pix_fmt);
}
}

if (!*got_output || ret < 0)
return ret;
//对应参数"top",和场有关
if(ist->top_field_first>=0)
decoded_frame->top_field_first = ist->top_field_first;
//记录解码的帧数
ist->frames_decoded++;
//和硬件相关
if (ist->hwaccel_retrieve_data && decoded_frame->format == ist->hwaccel_pix_fmt) {
err = ist->hwaccel_retrieve_data(ist->dec_ctx, decoded_frame);
if (err < 0)
goto fail;
}
ist->hwaccel_retrieved_pix_fmt = decoded_frame->format;
//此句很重要,是解码计算出的帧时间戳
best_effort_timestamp= av_frame_get_best_effort_timestamp(decoded_frame);
if(best_effort_timestamp != AV_NOPTS_VALUE) {
//ist中的时间单位是AV_TIME_BASE_Q,decoded_frame中的时间单位是ist->st->time_base
int64_t ts = av_rescale_q(decoded_frame->pts = best_effort_timestamp, ist->st->time_base, AV_TIME_BASE_Q);

if (ts != AV_NOPTS_VALUE)
ist->next_pts = ist->pts = ts;
}

if (debug_ts) {
av_log(NULL, AV_LOG_INFO, "decoder -> ist_index:%d type:video "
"frame_pts:%s frame_pts_time:%s best_effort_ts:%"PRId64" best_effort_ts_time:%s keyframe:%d frame_type:%d time_base:%d/%d\n",
ist->st->index, av_ts2str(decoded_frame->pts),
av_ts2timestr(decoded_frame->pts, &ist->st->time_base),
best_effort_timestamp,
av_ts2timestr(best_effort_timestamp, &ist->st->time_base),
decoded_frame->key_frame, decoded_frame->pict_type,
ist->st->time_base.num, ist->st->time_base.den);
}
/*表示视频就一帧。解完这次就不解了。处理音频时就没有这一句。在process_input_packet中会对此值做处理。
如果处理完还有值就需要在解码*/
pkt->size = 0;

if (ist->st->sample_aspect_ratio.num)
decoded_frame->sample_aspect_ratio = ist->st->sample_aspect_ratio;
//判断判断为真,需要重新设置滤镜链表。
resample_changed = ist->resample_width != decoded_frame->width ||
ist->resample_height != decoded_frame->height ||
ist->resample_pix_fmt != decoded_frame->format;
if (resample_changed) {
av_log(NULL, AV_LOG_INFO,
"Input stream #%d:%d frame changed from size:%dx%d fmt:%s to size:%dx%d fmt:%s\n",
ist->file_index, ist->st->index,
ist->resample_width, ist->resample_height, av_get_pix_fmt_name(ist->resample_pix_fmt),
decoded_frame->width, decoded_frame->height, av_get_pix_fmt_name(decoded_frame->format));

ist->resample_width = decoded_frame->width;
ist->resample_height = decoded_frame->height;
ist->resample_pix_fmt = decoded_frame->format;

for (i = 0; i < nb_filtergraphs; i++) {
if (ist_in_filtergraph(filtergraphs[i], ist) && ist->reinit_filters &&
configure_filtergraph(filtergraphs[i]) < 0) {
av_log(NULL, AV_LOG_FATAL, "Error reinitializing filters!\n");
exit_program(1);
}
}
}

frame_sample_aspect= av_opt_ptr(avcodec_get_frame_class(), decoded_frame, "sample_aspect_ratio");
for (i = 0; i < ist->nb_filters; i++) {
if (!frame_sample_aspect->num)
*frame_sample_aspect = ist->st->sample_aspect_ratio;

if (i < ist->nb_filters - 1) {
f = ist->filter_frame;
err = av_frame_ref(f, decoded_frame);
if (err < 0)
break;
} else
f = decoded_frame;
//放到滤镜里面
ret = av_buffersrc_add_frame_flags(ist->filters[i]->filter, f, AV_BUFFERSRC_FLAG_PUSH);
if (ret == AVERROR_EOF) {
ret = 0; /* ignore */
} else if (ret < 0) {
av_log(NULL, AV_LOG_FATAL,
"Failed to inject frame into filter network: %s\n", av_err2str(ret));
exit_program(1);
}
}

fail:
av_frame_unref(ist->filter_frame);
av_frame_unref(decoded_frame);
return err < 0 ? err : ret;
}
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
reap_filters

/**
* Get and encode new output from any of the filtergraphs, without causing
* activity.
*
* @return 0 for success, <0 for severe errors
*/
static int reap_filters(int flush)
{
AVFrame *filtered_frame = NULL;
int i;

/* Reap all buffers present in the buffer sinks */
for (i = 0; i < nb_output_streams; i++) {
OutputStream *ost = output_streams[i];
OutputFile *of = output_files[ost->file_index];
AVFilterContext *filter;
AVCodecContext *enc = ost->enc_ctx;
int ret = 0;

if (!ost->filter)
continue;
filter = ost->filter->filter;

if (!ost->filtered_frame && !(ost->filtered_frame = av_frame_alloc())) {
return AVERROR(ENOMEM);
}
//从滤镜处理后的帧存放在这里
filtered_frame = ost->filtered_frame;

while (1) {
double float_pts = AV_NOPTS_VALUE; // this is identical to filtered_frame.pts but with higher precision
//从滤镜获取帧
ret = av_buffersink_get_frame_flags(filter, filtered_frame,
AV_BUFFERSINK_FLAG_NO_REQUEST);
if (ret < 0) {
if (ret != AVERROR(EAGAIN) && ret != AVERROR_EOF) {
av_log(NULL, AV_LOG_WARNING,
"Error in av_buffersink_get_frame_flags(): %s\n", av_err2str(ret));
} else if (flush && ret == AVERROR_EOF) {
//没有数据了。已经到了滤镜的结束。
if (filter->inputs[0]->type == AVMEDIA_TYPE_VIDEO)
do_video_out(of->ctx, ost, NULL, AV_NOPTS_VALUE);
}
break;
}
//如果设置了此值,获取到的帧丢弃
if (ost->finished) {
av_frame_unref(filtered_frame);
continue;
}
//下面这段不知道啥意思
if (filtered_frame->pts != AV_NOPTS_VALUE) {
int64_t start_time = (of->start_time == AV_NOPTS_VALUE) ? 0 : of->start_time;
AVRational tb = enc->time_base;
int extra_bits = av_clip(29 - av_log2(tb.den), 0, 16);

tb.den <<= extra_bits;
float_pts =
av_rescale_q(filtered_frame->pts, filter->inputs[0]->time_base, tb) -
av_rescale_q(start_time, AV_TIME_BASE_Q, tb);
float_pts /= 1 << extra_bits;
// avoid exact midoints to reduce the chance of rounding differences, this can be removed in case the fps code is changed to work with integers
float_pts += FFSIGN(float_pts) * 1.0 / (1<<17);

filtered_frame->pts =
av_rescale_q(filtered_frame->pts, filter->inputs[0]->time_base, enc->time_base) -
av_rescale_q(start_time, AV_TIME_BASE_Q, enc->time_base);
}
//if (ost->source_index >= 0)
// *filtered_frame= *input_streams[ost->source_index]->decoded_frame; //for me_threshold

switch (filter->inputs[0]->type) {
case AVMEDIA_TYPE_VIDEO:
//设置宽高比
if (!ost->frame_aspect_ratio.num)
enc->sample_aspect_ratio = filtered_frame->sample_aspect_ratio;

if (debug_ts) {
av_log(NULL, AV_LOG_INFO, "filter -> pts:%s pts_time:%s exact:%f time_base:%d/%d\n",
av_ts2str(filtered_frame->pts), av_ts2timestr(filtered_frame->pts, &enc->time_base),
float_pts,
enc->time_base.num, enc->time_base.den);
}
//编码并输出到文件
do_video_out(of->ctx, ost, filtered_frame, float_pts);
break;
case AVMEDIA_TYPE_AUDIO:
if (!(enc->codec->capabilities & AV_CODEC_CAP_PARAM_CHANGE) &&
enc->channels != av_frame_get_channels(filtered_frame)) {
av_log(NULL, AV_LOG_ERROR,
"Audio filter graph output is not normalized and encoder does not support parameter changes\n");
break;
}
do_audio_out(of->ctx, ost, filtered_frame);
break;
default:
// TODO support subtitle filters
av_assert0(0);
}

av_frame_unref(filtered_frame);
}
}

return 0;
}
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
do_video_out

static void do_video_out(AVFormatContext *s,
OutputStream *ost,
AVFrame *next_picture,
double sync_ipts)
{
int ret, format_video_sync;
AVPacket pkt;
AVCodecContext *enc = ost->enc_ctx;
AVCodecContext *mux_enc = ost->st->codec;
int nb_frames, nb0_frames, i;
double delta, delta0;
double duration = 0;
int frame_size = 0;
InputStream *ist = NULL;
AVFilterContext *filter = ost->filter->filter;

if (ost->source_index >= 0)
ist = input_streams[ost->source_index];

if (filter->inputs[0]->frame_rate.num > 0 &&
filter->inputs[0]->frame_rate.den > 0)
duration = 1/(av_q2d(filter->inputs[0]->frame_rate) * av_q2d(enc->time_base));

if(ist && ist->st->start_time != AV_NOPTS_VALUE && ist->st->first_dts != AV_NOPTS_VALUE && ost->frame_rate.num)
duration = FFMIN(duration, 1/(av_q2d(ost->frame_rate) * av_q2d(enc->time_base)));

if (!ost->filters_script &&
!ost->filters &&
next_picture &&
ist &&
lrintf(av_frame_get_pkt_duration(next_picture) * av_q2d(ist->st->time_base) / av_q2d(enc->time_base)) > 0) {
duration = lrintf(av_frame_get_pkt_duration(next_picture) * av_q2d(ist->st->time_base) / av_q2d(enc->time_base));
}

if (!next_picture) {
//end, flushing
nb0_frames = nb_frames = mid_pred(ost->last_nb0_frames[0],
ost->last_nb0_frames[1],
ost->last_nb0_frames[2]);
} else {
delta0 = sync_ipts - ost->sync_opts; // delta0 is the "drift" between the input frame (next_picture) and where it would fall in the output.
delta = delta0 + duration;

/* by default, we output a single frame */
nb0_frames = 0; // tracks the number of times the PREVIOUS frame should be duplicated, mostly for variable framerate (VFR)
nb_frames = 1;

format_video_sync = video_sync_method;
if (format_video_sync == VSYNC_AUTO) {
if(!strcmp(s->oformat->name, "avi")) {
format_video_sync = VSYNC_VFR;
} else
format_video_sync = (s->oformat->flags & AVFMT_VARIABLE_FPS) ? ((s->oformat->flags & AVFMT_NOTIMESTAMPS) ? VSYNC_PASSTHROUGH : VSYNC_VFR) : VSYNC_CFR;
if ( ist
&& format_video_sync == VSYNC_CFR
&& input_files[ist->file_index]->ctx->nb_streams == 1
&& input_files[ist->file_index]->input_ts_offset == 0) {
format_video_sync = VSYNC_VSCFR;
}
if (format_video_sync == VSYNC_CFR && copy_ts) {
format_video_sync = VSYNC_VSCFR;
}
}
ost->is_cfr = (format_video_sync == VSYNC_CFR || format_video_sync == VSYNC_VSCFR);

if (delta0 < 0 &&
delta > 0 &&
format_video_sync != VSYNC_PASSTHROUGH &&
format_video_sync != VSYNC_DROP) {
if (delta0 < -0.6) {
av_log(NULL, AV_LOG_WARNING, "Past duration %f too large\n", -delta0);
} else
av_log(NULL, AV_LOG_DEBUG, "Clipping frame in rate conversion by %f\n", -delta0);
sync_ipts = ost->sync_opts;
duration += delta0;
delta0 = 0;
}

switch (format_video_sync) {
case VSYNC_VSCFR:
if (ost->frame_number == 0 && delta0 >= 0.5) {
av_log(NULL, AV_LOG_DEBUG, "Not duplicating %d initial frames\n", (int)lrintf(delta0));
delta = duration;
delta0 = 0;
ost->sync_opts = lrint(sync_ipts);
}
case VSYNC_CFR:
// FIXME set to 0.5 after we fix some dts/pts bugs like in avidec.c
if (frame_drop_threshold && delta < frame_drop_threshold && ost->frame_number) {
nb_frames = 0;
} else if (delta < -1.1)
nb_frames = 0;
else if (delta > 1.1) {
nb_frames = lrintf(delta);
if (delta0 > 1.1)
nb0_frames = lrintf(delta0 - 0.6);
}
break;
case VSYNC_VFR:
if (delta <= -0.6)
nb_frames = 0;
else if (delta > 0.6)
ost->sync_opts = lrint(sync_ipts);
break;
case VSYNC_DROP:
case VSYNC_PASSTHROUGH:
ost->sync_opts = lrint(sync_ipts);
break;
default:
av_assert0(0);
}
}

nb_frames = FFMIN(nb_frames, ost->max_frames - ost->frame_number);
nb0_frames = FFMIN(nb0_frames, nb_frames);

memmove(ost->last_nb0_frames + 1,
ost->last_nb0_frames,
sizeof(ost->last_nb0_frames[0]) * (FF_ARRAY_ELEMS(ost->last_nb0_frames) - 1));
ost->last_nb0_frames[0] = nb0_frames;

if (nb0_frames == 0 && ost->last_dropped) {
nb_frames_drop++;
av_log(NULL, AV_LOG_VERBOSE,
"*** dropping frame %d from stream %d at ts %"PRId64"\n",
ost->frame_number, ost->st->index, ost->last_frame->pts);
}
if (nb_frames > (nb0_frames && ost->last_dropped)www.huacairen88.cn + (nb_frames > nb0_frames)) {
if (nb_frames > dts_error_threshold * 30) {
av_log(NULL, AV_LOG_ERROR, "%d frame duplication too large, skipping\n", nb_frames - 1);
nb_frames_drop++;
return;
}
nb_frames_dup += nb_frames - (nb0_frames && ost->last_dropped) - (nb_frames > nb0_frames);
av_log(NULL, AV_LOG_VERBOSE, "*** %d dup!\n", nb_frames - 1);
}
ost->last_dropped = nb_frames == nb0_frames && next_picture;

/* duplicates frame if needed */
for (i = 0; i < nb_frames; i++) {
AVFrame *in_picture;
av_init_packet(&pkt);
pkt.data = NULL;
pkt.size = 0;

if (i < nb0_frames && ost->last_ www.feiyunyl.cn frame) {
in_picture = ost->last_frame;
} else
in_picture = next_picture;

if (!in_picture)
return;

in_picture->pts = ost->sync_opts;

#if 1
if (!check_recording_time(ost))
#else
if (ost->frame_number >= ost->max_frames)
#endif
return;

#if FF_API_LAVF_FMT_RAWPICTURE
if (s->oformat->flags & AVFMT_RAWPICTURE &&
enc->codec->id == AV_CODEC_ID_RAWVIDEO) {
/* raw pictures are written as AVPicture structure to
avoid any copies. We support temporarily the older
method. */
if (in_picture->interlaced_frame)
mux_enc->field_order = in_picture->top_field_first ? AV_FIELD_TB:AV_FIELD_BT;
else
mux_enc->field_order = AV_FIELD_PROGRESSIVE;
pkt.data = (uint8_t *)in_picture;
pkt.size = sizeof(AVPicture);
pkt.pts = av_rescale_q(in_picture->pts,www.feihuayl.cn enc->time_base, ost->st->time_base);
pkt.flags |= AV_PKT_FLAG_KEY;

write_frame(s, &pkt, ost);
} else
#endif
{
int got_packet, forced_keyframe = 0;
double pts_time;

if (enc->flags & (AV_CODEC_FLAG_INTERLACED_DCT | AV_CODEC_FLAG_INTERLACED_ME) &&
ost->top_field_first >= 0)
in_picture->top_field_first = !!ost->top_field_first;

if (in_picture->interlaced_frame) {
if (enc->codec->id == AV_CODEC_ID_MJPEG)
mux_enc->field_order = in_picture->www.fsmaxbuy.com/ top_field_first ? AV_FIELD_TT:AV_FIELD_BB;
else
mux_enc->field_order = in_picture->top_field_first ? AV_FIELD_TB:AV_FIELD_BT;
} else
mux_enc->field_order = AV_FIELD_PROGRESSIVE;

in_picture->quality = enc->global_quality;
in_picture->pict_type = 0;

pts_time = in_picture->pts != AV_NOPTS_VALUE ?
in_picture->pts * av_q2d(enc->time_base) : NAN;
if (ost->forced_kf_index < ost->forced_kf_count &&
in_picture->pts >= ost->forced_kf_pts[ost->forced_kf_index]) {
ost->forced_kf_index++;
forced_keyframe = 1;
} else if (ost->forced_keyframes_pexpr) {
double res;
ost->forced_keyframes_expr_const_values[FKF_T] = pts_time;
res = av_expr_eval(ost->forced_keyframes_pexpr,
ost->forced_keyframes_expr_const_values, NULL);
ff_dlog(NULL, "force_key_frame: n:%f n_forced:%f prev_forced_n:%f t:%f prev_forced_t:%f -> res:%f\n",
ost->forced_keyframes_expr_const_values[FKF_N],
ost->forced_keyframes_expr_const_values[FKF_N_FORCED],
ost->forced_keyframes_expr_const_values[FKF_PREV_FORCED_N],
ost->forced_keyframes_expr_const_vwww.yunfeizao.cn
www.feiyunfan.cn alues[FKF_T],
ost->forced_keyframes_expr_const_values[FKF_PREV_FORCED_T],
res);
if (res) {
forced_keyframe = 1;
ost->forced_keyframes_expr_const_values[FKF_PREV_FORCED_N] =
ost->forced_keyframes_expr_const_values[FKF_N];
ost->forced_keyframes_expr_const_values[FKF_PREV_FORCED_T] =
ost->forced_keyframes_expr_const_values[FKF_T];
ost->forced_keyframes_expr_const_values[FKF_N_FORCED] += 1;
}

ost->forced_keyframes_expr_const_values[FKF_N] += 1;
} else if ( ost->forced_keyframes
&& !strncmp(ost->forced_keyframes, "source", 6)
&& in_picture->key_frame==1) {
forced_keyframe = 1;
}

if (forced_keyframe) {
in_picture->pict_type = AV_PICTURE_TYPE_I;
av_log(NULL, AV_LOG_DEBUG, "Forced keyframe at time %f\n", pts_time);
}

update_benchmark(NULL);
if (debug_ts) {
av_log(NULL, AV_LOG_INFO, "encoder <- type:video "
"frame_pts:%s frame_pts_time:%s time_base:%d/%d\n",
av_ts2str(in_picture->pts), av_ts2timestr(in_picture->pts, &enc->time_base),
enc->time_base.num, enc->time_base.den);
}

ost->frames_encoded++;

ret = avcodec_encode_video2(enc, &pkt, in_picture, &got_packet);
update_benchmark("encode_video %d.%d", ost->file_index, ost->index);
if (ret < 0) {
av_log(NULL, AV_LOG_FATAL, "Video encoding failed\n");
exit_program(1);
}

if (got_packet) {
if (debug_ts) {
av_log(NULL, AV_LOG_INFO, "encoder -> type:video "
"pkt_pts:%s pkt_pts_time:%s pkt_dts:%s pkt_dts_time:%s\n",
av_ts2str(pkt.pts), av_ts2timestr(pkt.pts, &enc->time_base),
av_ts2str(pkt.dts), av_ts2timestr(pkt.dts, &enc->time_base));
}

if (pkt.pts == AV_NOPTS_VALUE && !(enc->codec->capabilities & AV_CODEC_CAP_DELAY))
pkt.pts = ost->sync_opts;

av_packet_rescale_ts(&pkt, enc->time_base, ost->st->time_base);

if (debug_ts) {
av_log(NULL, AV_LOG_INFO, "encoder -> type:video "
"pkt_pts:%s pkt_pts_time:%s pkt_dts:%s pkt_dts_time:%s\n",
av_ts2str(pkt.pts), av_ts2timestr(pkt.pts, &ost->st->time_base),
av_ts2str(pkt.dts), av_ts2timestr(pkt.dts, &ost->st->time_base));
}

frame_size = pkt.size;
write_frame(s, &pkt, ost);

/* if two pass, output log */
if (ost->logfile && enc->stats_out) {
fprintf(ost->logfile, "%s", enc->stats_out);
}
}
}
ost->sync_opts++;
/*
* For video, number of frames in == number of packets out.
* But there may be reordering, so we can't throw away frames on encoder
* flush, we need to limit them here, before they go into encoder.
*/
ost->frame_number++;

if (vstats_filename && frame_size)
do_video_stats(ost, frame_size);
}

if (!ost->last_frame)
ost->last_frame = av_frame_alloc();
av_frame_unref(ost->last_frame);
if (next_picture && ost->last_frame)
av_frame_ref(ost->last_frame, next_picture);
else
av_frame_free(&ost->last_frame);

transcode_step()在转码过程中对pts、dts、duration的处理的更多相关文章

  1. Ubuntu编译Android源码过程中的空间不足解决方法

    Android源码一般几十G,就拿Android5.0来说,下载下来大概也有44G左右,和编译产生的文件以及Ubuntu系统占用的空间加起来,源码双倍的空间都不够有.编译源码前能分配足够的空间再好不过 ...

  2. [原]编译Android源码过程中遇到的问题

    编译Android源码的过程参考Android官网介绍: 1.下载Android源码的步骤:https://source.android.com/source/downloading.html 2.编 ...

  3. nginx源码编译以及源码编译过程中遇到的问题

    本文主要讲nginx安装以及安装过程中遇到的问题. 谈到nginx 必须聊聊它的起源和发展. nginx是由俄罗斯工程师Igor Sysoev 用C语言开发的一个免费开源的Web服务器软件,于2004 ...

  4. mysql-5.5.28源码安装过程中错误总结

    介绍一下关于mysql-5.5.28源码安装过程中几大错误总结,希望此文章对各位同学有所帮助.系统centOS 6.3 mini (没有任何编译环境)预编译环境首先装了众所周知的 cmake(yum ...

  5. Linux系统源码安装过程中的prefix选项

    在linux和unix环境中,源码安装是最常用的软件安装方式,一些软件除了提供源码外,也提供各种发行版的二进制安装包(如基于redhat包管理工具的rpm包),但强烈建议使用源码安装方式.原因是:(1 ...

  6. OpenJDK源码研究笔记(十三):Javac编译过程中的上下文容器(Context)、单例(Singleton)和延迟创建(LazyCreation)3种模式

    在阅读Javac源码的过程中,发现一个上下文对象Context. 这个对象用来确保一次编译过程中的用到的类都只有一个实例,即实现我们经常提到的"单例模式". 今天,特意对这个上下文 ...

  7. 调试过程中发现按f5无法走进jdk源码

    debug 模式 ,在fis=new FileInputStream(file); 行打断点 调试过程中发现按f5无法走进jdk源码 package com.lzl.spring.test; impo ...

  8. 3.3 Spring5源码---循环依赖过程中spring读取不完整bean的最终解决方案

    根据之前解析的循环依赖的源码, 分析了一级缓存,二级缓存,三级缓存的作用以及如何解决循环依赖的. 然而在多线程的情况下, Spring在创建bean的过程中, 可能会读取到不完整的bean. 下面, ...

  9. live555源码研究(十)------在编译过程中遇到的问题及解决方法

    一.编译testOnDemandRTSPServer.cpp. 在testProgs项目中,加入testOnDemandRTSPServer.cpp进行编译,编译类型是编译成exe文件,在编译过程中会 ...

随机推荐

  1. Java-String之寻根问底

    Java-String之寻根问底 引言 在java编程中,几乎每天都会跟String打交道,因此,深入理解String及其用法十分有必要.下面分三方面来详细说明下String相关的特点及用法 •Imm ...

  2. 机器学习笔记1——Linear Regression with One Variable

    Linear Regression with One Variable Model Representation Recall that in *regression problems*, we ar ...

  3. PHP中使用函数array_merge()合并数组

    如果明白数组其实就是map的话,我想你就会明白array_merge为什么要这么实现了 PHP中合并数组分成两种情况 1.如果这两个数组中有相同的字符串键名: <?php header('Con ...

  4. Y2K Accounting Bug

    题目: Description Accounting for Computer Machinists (ACM) has sufferred from the Y2K bug and lost som ...

  5. HDU 4635 多校第四场 1004 强联通

    我还有什么好说,还有什么好说...... 我是SBSBSBSBSBSBSBSBSBSBSBSBBSBSBSBSBSBSBSBSBS........................ 题意 思路什么的都不 ...

  6. 大文件遍历shell脚本

    要求说明: 一.普通方法 无读写磁盘优化 有写磁盘优化 有读写磁盘优化 问题:脚本执行越来越慢 top ps aux vmstat 查看系统运行情况正常. 二.AWK 三.perl

  7. Java或web中解决所有路径问题

    Java开发中使用的路径,分为两种:绝对路径和相对路径.归根结底,Java本质上只能使用绝对路径来寻找资源.所有的相对路径寻找资源的方法,都不过是一些便利方法.不过是API在底层帮助我们构建了绝对路径 ...

  8. C#常用工具类——Excel操作类

    /// 常用工具类——Excel操作类 /// <para> ------------------------------------------------</para> / ...

  9. python中的lambda表达

    C++中的lambda表达式与C++11增加标准库,是一个简短的匿名的可调用对象,编译器会将其转化为一个匿名类的对象.lambda表达式的最大特点就是简短灵活.调用方便.它不须要处理非常复杂的逻辑.通 ...

  10. 追踪CPU跑满 堆栈调试

    http://blog.donghao.org/2014/04/24/%E8%BF%BD%E8%B8%AAcpu%E8%B7%91%E6%BB%A1/