| 1 | /* |
| 2 | * Copyright (c) 2010 Stefano Sabatini |
| 3 | * Copyright (c) 2008 Victor Paesa |
| 4 | * |
| 5 | * This file is part of FFmpeg. |
| 6 | * |
| 7 | * FFmpeg is free software; you can redistribute it and/or |
| 8 | * modify it under the terms of the GNU Lesser General Public |
| 9 | * License as published by the Free Software Foundation; either |
| 10 | * version 2.1 of the License, or (at your option) any later version. |
| 11 | * |
| 12 | * FFmpeg is distributed in the hope that it will be useful, |
| 13 | * but WITHOUT ANY WARRANTY; without even the implied warranty of |
| 14 | * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU |
| 15 | * Lesser General Public License for more details. |
| 16 | * |
| 17 | * You should have received a copy of the GNU Lesser General Public |
| 18 | * License along with FFmpeg; if not, write to the Free Software |
| 19 | * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA |
| 20 | */ |
| 21 | |
| 22 | /** |
| 23 | * @file |
| 24 | * video presentation timestamp (PTS) modification filter |
| 25 | */ |
| 26 | |
| 27 | #include <inttypes.h> |
| 28 | |
| 29 | #include "libavutil/eval.h" |
| 30 | #include "libavutil/internal.h" |
| 31 | #include "libavutil/mathematics.h" |
| 32 | #include "libavutil/opt.h" |
| 33 | #include "libavutil/time.h" |
| 34 | #include "audio.h" |
| 35 | #include "avfilter.h" |
| 36 | #include "internal.h" |
| 37 | #include "video.h" |
| 38 | |
| 39 | static const char *const var_names[] = { |
| 40 | "FRAME_RATE", ///< defined only for constant frame-rate video |
| 41 | "INTERLACED", ///< tell if the current frame is interlaced |
| 42 | "N", ///< frame / sample number (starting at zero) |
| 43 | "NB_CONSUMED_SAMPLES", ///< number of samples consumed by the filter (only audio) |
| 44 | "NB_SAMPLES", ///< number of samples in the current frame (only audio) |
| 45 | "POS", ///< original position in the file of the frame |
| 46 | "PREV_INPTS", ///< previous input PTS |
| 47 | "PREV_INT", ///< previous input time in seconds |
| 48 | "PREV_OUTPTS", ///< previous output PTS |
| 49 | "PREV_OUTT", ///< previous output time in seconds |
| 50 | "PTS", ///< original pts in the file of the frame |
| 51 | "SAMPLE_RATE", ///< sample rate (only audio) |
| 52 | "STARTPTS", ///< PTS at start of movie |
| 53 | "STARTT", ///< time at start of movie |
| 54 | "T", ///< original time in the file of the frame |
| 55 | "TB", ///< timebase |
| 56 | "RTCTIME", ///< wallclock (RTC) time in micro seconds |
| 57 | "RTCSTART", ///< wallclock (RTC) time at the start of the movie in micro seconds |
| 58 | "S", // Number of samples in the current frame |
| 59 | "SR", // Audio sample rate |
| 60 | NULL |
| 61 | }; |
| 62 | |
| 63 | enum var_name { |
| 64 | VAR_FRAME_RATE, |
| 65 | VAR_INTERLACED, |
| 66 | VAR_N, |
| 67 | VAR_NB_CONSUMED_SAMPLES, |
| 68 | VAR_NB_SAMPLES, |
| 69 | VAR_POS, |
| 70 | VAR_PREV_INPTS, |
| 71 | VAR_PREV_INT, |
| 72 | VAR_PREV_OUTPTS, |
| 73 | VAR_PREV_OUTT, |
| 74 | VAR_PTS, |
| 75 | VAR_SAMPLE_RATE, |
| 76 | VAR_STARTPTS, |
| 77 | VAR_STARTT, |
| 78 | VAR_T, |
| 79 | VAR_TB, |
| 80 | VAR_RTCTIME, |
| 81 | VAR_RTCSTART, |
| 82 | VAR_S, |
| 83 | VAR_SR, |
| 84 | VAR_VARS_NB |
| 85 | }; |
| 86 | |
| 87 | typedef struct SetPTSContext { |
| 88 | const AVClass *class; |
| 89 | char *expr_str; |
| 90 | AVExpr *expr; |
| 91 | double var_values[VAR_VARS_NB]; |
| 92 | enum AVMediaType type; |
| 93 | } SetPTSContext; |
| 94 | |
| 95 | static av_cold int init(AVFilterContext *ctx) |
| 96 | { |
| 97 | SetPTSContext *setpts = ctx->priv; |
| 98 | int ret; |
| 99 | |
| 100 | if ((ret = av_expr_parse(&setpts->expr, setpts->expr_str, |
| 101 | var_names, NULL, NULL, NULL, NULL, 0, ctx)) < 0) { |
| 102 | av_log(ctx, AV_LOG_ERROR, "Error while parsing expression '%s'\n", setpts->expr_str); |
| 103 | return ret; |
| 104 | } |
| 105 | |
| 106 | setpts->var_values[VAR_N] = 0.0; |
| 107 | setpts->var_values[VAR_S] = 0.0; |
| 108 | setpts->var_values[VAR_PREV_INPTS] = NAN; |
| 109 | setpts->var_values[VAR_PREV_INT] = NAN; |
| 110 | setpts->var_values[VAR_PREV_OUTPTS] = NAN; |
| 111 | setpts->var_values[VAR_PREV_OUTT] = NAN; |
| 112 | setpts->var_values[VAR_STARTPTS] = NAN; |
| 113 | setpts->var_values[VAR_STARTT] = NAN; |
| 114 | return 0; |
| 115 | } |
| 116 | |
| 117 | static int config_input(AVFilterLink *inlink) |
| 118 | { |
| 119 | AVFilterContext *ctx = inlink->dst; |
| 120 | SetPTSContext *setpts = ctx->priv; |
| 121 | |
| 122 | setpts->type = inlink->type; |
| 123 | setpts->var_values[VAR_TB] = av_q2d(inlink->time_base); |
| 124 | setpts->var_values[VAR_RTCSTART] = av_gettime(); |
| 125 | |
| 126 | setpts->var_values[VAR_SR] = |
| 127 | setpts->var_values[VAR_SAMPLE_RATE] = |
| 128 | setpts->type == AVMEDIA_TYPE_AUDIO ? inlink->sample_rate : NAN; |
| 129 | |
| 130 | setpts->var_values[VAR_FRAME_RATE] = inlink->frame_rate.num && inlink->frame_rate.den ? |
| 131 | av_q2d(inlink->frame_rate) : NAN; |
| 132 | |
| 133 | av_log(inlink->src, AV_LOG_VERBOSE, "TB:%f FRAME_RATE:%f SAMPLE_RATE:%f\n", |
| 134 | setpts->var_values[VAR_TB], |
| 135 | setpts->var_values[VAR_FRAME_RATE], |
| 136 | setpts->var_values[VAR_SAMPLE_RATE]); |
| 137 | return 0; |
| 138 | } |
| 139 | |
| 140 | #define D2TS(d) (isnan(d) ? AV_NOPTS_VALUE : (int64_t)(d)) |
| 141 | #define TS2D(ts) ((ts) == AV_NOPTS_VALUE ? NAN : (double)(ts)) |
| 142 | #define TS2T(ts, tb) ((ts) == AV_NOPTS_VALUE ? NAN : (double)(ts)*av_q2d(tb)) |
| 143 | |
| 144 | #define BUF_SIZE 64 |
| 145 | |
| 146 | static inline char *double2int64str(char *buf, double v) |
| 147 | { |
| 148 | if (isnan(v)) snprintf(buf, BUF_SIZE, "nan"); |
| 149 | else snprintf(buf, BUF_SIZE, "%"PRId64, (int64_t)v); |
| 150 | return buf; |
| 151 | } |
| 152 | |
| 153 | #define d2istr(v) double2int64str((char[BUF_SIZE]){0}, v) |
| 154 | |
| 155 | static int filter_frame(AVFilterLink *inlink, AVFrame *frame) |
| 156 | { |
| 157 | SetPTSContext *setpts = inlink->dst->priv; |
| 158 | int64_t in_pts = frame->pts; |
| 159 | double d; |
| 160 | |
| 161 | if (isnan(setpts->var_values[VAR_STARTPTS])) { |
| 162 | setpts->var_values[VAR_STARTPTS] = TS2D(frame->pts); |
| 163 | setpts->var_values[VAR_STARTT ] = TS2T(frame->pts, inlink->time_base); |
| 164 | } |
| 165 | setpts->var_values[VAR_PTS ] = TS2D(frame->pts); |
| 166 | setpts->var_values[VAR_T ] = TS2T(frame->pts, inlink->time_base); |
| 167 | setpts->var_values[VAR_POS ] = av_frame_get_pkt_pos(frame) == -1 ? NAN : av_frame_get_pkt_pos(frame); |
| 168 | setpts->var_values[VAR_RTCTIME ] = av_gettime(); |
| 169 | |
| 170 | if (inlink->type == AVMEDIA_TYPE_VIDEO) { |
| 171 | setpts->var_values[VAR_INTERLACED] = frame->interlaced_frame; |
| 172 | } else if (inlink->type == AVMEDIA_TYPE_AUDIO) { |
| 173 | setpts->var_values[VAR_S] = frame->nb_samples; |
| 174 | setpts->var_values[VAR_NB_SAMPLES] = frame->nb_samples; |
| 175 | } |
| 176 | |
| 177 | d = av_expr_eval(setpts->expr, setpts->var_values, NULL); |
| 178 | frame->pts = D2TS(d); |
| 179 | |
| 180 | av_dlog(inlink->dst, |
| 181 | "N:%"PRId64" PTS:%s T:%f POS:%s", |
| 182 | (int64_t)setpts->var_values[VAR_N], |
| 183 | d2istr(setpts->var_values[VAR_PTS]), |
| 184 | setpts->var_values[VAR_T], |
| 185 | d2istr(setpts->var_values[VAR_POS])); |
| 186 | switch (inlink->type) { |
| 187 | case AVMEDIA_TYPE_VIDEO: |
| 188 | av_dlog(inlink->dst, " INTERLACED:%"PRId64, |
| 189 | (int64_t)setpts->var_values[VAR_INTERLACED]); |
| 190 | break; |
| 191 | case AVMEDIA_TYPE_AUDIO: |
| 192 | av_dlog(inlink->dst, " NB_SAMPLES:%"PRId64" NB_CONSUMED_SAMPLES:%"PRId64, |
| 193 | (int64_t)setpts->var_values[VAR_NB_SAMPLES], |
| 194 | (int64_t)setpts->var_values[VAR_NB_CONSUMED_SAMPLES]); |
| 195 | break; |
| 196 | } |
| 197 | av_dlog(inlink->dst, " -> PTS:%s T:%f\n", d2istr(d), TS2T(d, inlink->time_base)); |
| 198 | |
| 199 | if (inlink->type == AVMEDIA_TYPE_VIDEO) { |
| 200 | setpts->var_values[VAR_N] += 1.0; |
| 201 | } else { |
| 202 | setpts->var_values[VAR_N] += frame->nb_samples; |
| 203 | } |
| 204 | |
| 205 | setpts->var_values[VAR_PREV_INPTS ] = TS2D(in_pts); |
| 206 | setpts->var_values[VAR_PREV_INT ] = TS2T(in_pts, inlink->time_base); |
| 207 | setpts->var_values[VAR_PREV_OUTPTS] = TS2D(frame->pts); |
| 208 | setpts->var_values[VAR_PREV_OUTT] = TS2T(frame->pts, inlink->time_base); |
| 209 | if (setpts->type == AVMEDIA_TYPE_AUDIO) { |
| 210 | setpts->var_values[VAR_NB_CONSUMED_SAMPLES] += frame->nb_samples; |
| 211 | } |
| 212 | return ff_filter_frame(inlink->dst->outputs[0], frame); |
| 213 | } |
| 214 | |
| 215 | static av_cold void uninit(AVFilterContext *ctx) |
| 216 | { |
| 217 | SetPTSContext *setpts = ctx->priv; |
| 218 | av_expr_free(setpts->expr); |
| 219 | setpts->expr = NULL; |
| 220 | } |
| 221 | |
| 222 | #define OFFSET(x) offsetof(SetPTSContext, x) |
| 223 | #define FLAGS AV_OPT_FLAG_VIDEO_PARAM | AV_OPT_FLAG_AUDIO_PARAM | AV_OPT_FLAG_FILTERING_PARAM |
| 224 | static const AVOption options[] = { |
| 225 | { "expr", "Expression determining the frame timestamp", OFFSET(expr_str), AV_OPT_TYPE_STRING, { .str = "PTS" }, .flags = FLAGS }, |
| 226 | { NULL } |
| 227 | }; |
| 228 | |
| 229 | #if CONFIG_SETPTS_FILTER |
| 230 | #define setpts_options options |
| 231 | AVFILTER_DEFINE_CLASS(setpts); |
| 232 | |
| 233 | static const AVFilterPad avfilter_vf_setpts_inputs[] = { |
| 234 | { |
| 235 | .name = "default", |
| 236 | .type = AVMEDIA_TYPE_VIDEO, |
| 237 | .config_props = config_input, |
| 238 | .filter_frame = filter_frame, |
| 239 | }, |
| 240 | { NULL } |
| 241 | }; |
| 242 | |
| 243 | static const AVFilterPad avfilter_vf_setpts_outputs[] = { |
| 244 | { |
| 245 | .name = "default", |
| 246 | .type = AVMEDIA_TYPE_VIDEO, |
| 247 | }, |
| 248 | { NULL } |
| 249 | }; |
| 250 | |
| 251 | AVFilter ff_vf_setpts = { |
| 252 | .name = "setpts", |
| 253 | .description = NULL_IF_CONFIG_SMALL("Set PTS for the output video frame."), |
| 254 | .init = init, |
| 255 | .uninit = uninit, |
| 256 | |
| 257 | .priv_size = sizeof(SetPTSContext), |
| 258 | .priv_class = &setpts_class, |
| 259 | |
| 260 | .inputs = avfilter_vf_setpts_inputs, |
| 261 | .outputs = avfilter_vf_setpts_outputs, |
| 262 | }; |
| 263 | #endif /* CONFIG_SETPTS_FILTER */ |
| 264 | |
| 265 | #if CONFIG_ASETPTS_FILTER |
| 266 | |
| 267 | #define asetpts_options options |
| 268 | AVFILTER_DEFINE_CLASS(asetpts); |
| 269 | |
| 270 | static const AVFilterPad asetpts_inputs[] = { |
| 271 | { |
| 272 | .name = "default", |
| 273 | .type = AVMEDIA_TYPE_AUDIO, |
| 274 | .config_props = config_input, |
| 275 | .filter_frame = filter_frame, |
| 276 | }, |
| 277 | { NULL } |
| 278 | }; |
| 279 | |
| 280 | static const AVFilterPad asetpts_outputs[] = { |
| 281 | { |
| 282 | .name = "default", |
| 283 | .type = AVMEDIA_TYPE_AUDIO, |
| 284 | }, |
| 285 | { NULL } |
| 286 | }; |
| 287 | |
| 288 | AVFilter ff_af_asetpts = { |
| 289 | .name = "asetpts", |
| 290 | .description = NULL_IF_CONFIG_SMALL("Set PTS for the output audio frame."), |
| 291 | .init = init, |
| 292 | .uninit = uninit, |
| 293 | .priv_size = sizeof(SetPTSContext), |
| 294 | .priv_class = &asetpts_class, |
| 295 | .inputs = asetpts_inputs, |
| 296 | .outputs = asetpts_outputs, |
| 297 | }; |
| 298 | #endif /* CONFIG_ASETPTS_FILTER */ |