You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.

478 lines
17KB

  1. /*
  2. * Copyright (c) 2011 Stefano Sabatini
  3. * This file is part of FFmpeg.
  4. *
  5. * FFmpeg is free software; you can redistribute it and/or
  6. * modify it under the terms of the GNU Lesser General Public
  7. * License as published by the Free Software Foundation; either
  8. * version 2.1 of the License, or (at your option) any later version.
  9. *
  10. * FFmpeg is distributed in the hope that it will be useful,
  11. * but WITHOUT ANY WARRANTY; without even the implied warranty of
  12. * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
  13. * Lesser General Public License for more details.
  14. *
  15. * You should have received a copy of the GNU Lesser General Public
  16. * License along with FFmpeg; if not, write to the Free Software
  17. * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
  18. */
  19. /**
  20. * @file
  21. * filter for showing textual video frame information
  22. */
  23. #include <inttypes.h>
  24. #include "libavutil/bswap.h"
  25. #include "libavutil/adler32.h"
  26. #include "libavutil/display.h"
  27. #include "libavutil/imgutils.h"
  28. #include "libavutil/internal.h"
  29. #include "libavutil/opt.h"
  30. #include "libavutil/pixdesc.h"
  31. #include "libavutil/spherical.h"
  32. #include "libavutil/stereo3d.h"
  33. #include "libavutil/timestamp.h"
  34. #include "libavutil/timecode.h"
  35. #include "libavutil/mastering_display_metadata.h"
  36. #include "libavutil/video_enc_params.h"
  37. #include "avfilter.h"
  38. #include "internal.h"
  39. #include "video.h"
  40. typedef struct ShowInfoContext {
  41. const AVClass *class;
  42. int calculate_checksums;
  43. } ShowInfoContext;
  44. #define OFFSET(x) offsetof(ShowInfoContext, x)
  45. #define VF AV_OPT_FLAG_VIDEO_PARAM|AV_OPT_FLAG_FILTERING_PARAM
  46. static const AVOption showinfo_options[] = {
  47. { "checksum", "calculate checksums", OFFSET(calculate_checksums), AV_OPT_TYPE_BOOL, {.i64=1}, 0, 1, VF },
  48. { NULL }
  49. };
  50. AVFILTER_DEFINE_CLASS(showinfo);
  51. static void dump_spherical(AVFilterContext *ctx, AVFrame *frame, const AVFrameSideData *sd)
  52. {
  53. const AVSphericalMapping *spherical = (const AVSphericalMapping *)sd->data;
  54. double yaw, pitch, roll;
  55. av_log(ctx, AV_LOG_INFO, "spherical information: ");
  56. if (sd->size < sizeof(*spherical)) {
  57. av_log(ctx, AV_LOG_ERROR, "invalid data\n");
  58. return;
  59. }
  60. if (spherical->projection == AV_SPHERICAL_EQUIRECTANGULAR)
  61. av_log(ctx, AV_LOG_INFO, "equirectangular ");
  62. else if (spherical->projection == AV_SPHERICAL_CUBEMAP)
  63. av_log(ctx, AV_LOG_INFO, "cubemap ");
  64. else if (spherical->projection == AV_SPHERICAL_EQUIRECTANGULAR_TILE)
  65. av_log(ctx, AV_LOG_INFO, "tiled equirectangular ");
  66. else {
  67. av_log(ctx, AV_LOG_WARNING, "unknown\n");
  68. return;
  69. }
  70. yaw = ((double)spherical->yaw) / (1 << 16);
  71. pitch = ((double)spherical->pitch) / (1 << 16);
  72. roll = ((double)spherical->roll) / (1 << 16);
  73. av_log(ctx, AV_LOG_INFO, "(%f/%f/%f) ", yaw, pitch, roll);
  74. if (spherical->projection == AV_SPHERICAL_EQUIRECTANGULAR_TILE) {
  75. size_t l, t, r, b;
  76. av_spherical_tile_bounds(spherical, frame->width, frame->height,
  77. &l, &t, &r, &b);
  78. av_log(ctx, AV_LOG_INFO,
  79. "[%"SIZE_SPECIFIER", %"SIZE_SPECIFIER", %"SIZE_SPECIFIER", %"SIZE_SPECIFIER"] ",
  80. l, t, r, b);
  81. } else if (spherical->projection == AV_SPHERICAL_CUBEMAP) {
  82. av_log(ctx, AV_LOG_INFO, "[pad %"PRIu32"] ", spherical->padding);
  83. }
  84. }
  85. static void dump_stereo3d(AVFilterContext *ctx, const AVFrameSideData *sd)
  86. {
  87. const AVStereo3D *stereo;
  88. av_log(ctx, AV_LOG_INFO, "stereoscopic information: ");
  89. if (sd->size < sizeof(*stereo)) {
  90. av_log(ctx, AV_LOG_ERROR, "invalid data\n");
  91. return;
  92. }
  93. stereo = (const AVStereo3D *)sd->data;
  94. av_log(ctx, AV_LOG_INFO, "type - %s", av_stereo3d_type_name(stereo->type));
  95. if (stereo->flags & AV_STEREO3D_FLAG_INVERT)
  96. av_log(ctx, AV_LOG_INFO, " (inverted)");
  97. }
  98. static void dump_s12m_timecode(AVFilterContext *ctx, AVRational frame_rate, const AVFrameSideData *sd)
  99. {
  100. const uint32_t *tc = (const uint32_t *)sd->data;
  101. if ((sd->size != sizeof(uint32_t) * 4) || (tc[0] > 3)) {
  102. av_log(ctx, AV_LOG_ERROR, "invalid data\n");
  103. return;
  104. }
  105. for (int j = 1; j <= tc[0]; j++) {
  106. char tcbuf[AV_TIMECODE_STR_SIZE];
  107. av_timecode_make_smpte_tc_string2(tcbuf, frame_rate, tc[j], 0, 0);
  108. av_log(ctx, AV_LOG_INFO, "timecode - %s%s", tcbuf, j != tc[0] ? ", " : "");
  109. }
  110. }
  111. static void dump_roi(AVFilterContext *ctx, const AVFrameSideData *sd)
  112. {
  113. int nb_rois;
  114. const AVRegionOfInterest *roi;
  115. uint32_t roi_size;
  116. roi = (const AVRegionOfInterest *)sd->data;
  117. roi_size = roi->self_size;
  118. if (!roi_size || sd->size % roi_size != 0) {
  119. av_log(ctx, AV_LOG_ERROR, "Invalid AVRegionOfInterest.self_size.\n");
  120. return;
  121. }
  122. nb_rois = sd->size / roi_size;
  123. av_log(ctx, AV_LOG_INFO, "Regions Of Interest(RoI) information: ");
  124. for (int i = 0; i < nb_rois; i++) {
  125. roi = (const AVRegionOfInterest *)(sd->data + roi_size * i);
  126. av_log(ctx, AV_LOG_INFO, "index: %d, region: (%d, %d)/(%d, %d), qp offset: %d/%d.\n",
  127. i, roi->left, roi->top, roi->right, roi->bottom, roi->qoffset.num, roi->qoffset.den);
  128. }
  129. }
  130. static void dump_mastering_display(AVFilterContext *ctx, const AVFrameSideData *sd)
  131. {
  132. const AVMasteringDisplayMetadata *mastering_display;
  133. av_log(ctx, AV_LOG_INFO, "mastering display: ");
  134. if (sd->size < sizeof(*mastering_display)) {
  135. av_log(ctx, AV_LOG_ERROR, "invalid data\n");
  136. return;
  137. }
  138. mastering_display = (const AVMasteringDisplayMetadata *)sd->data;
  139. av_log(ctx, AV_LOG_INFO, "has_primaries:%d has_luminance:%d "
  140. "r(%5.4f,%5.4f) g(%5.4f,%5.4f) b(%5.4f %5.4f) wp(%5.4f, %5.4f) "
  141. "min_luminance=%f, max_luminance=%f",
  142. mastering_display->has_primaries, mastering_display->has_luminance,
  143. av_q2d(mastering_display->display_primaries[0][0]),
  144. av_q2d(mastering_display->display_primaries[0][1]),
  145. av_q2d(mastering_display->display_primaries[1][0]),
  146. av_q2d(mastering_display->display_primaries[1][1]),
  147. av_q2d(mastering_display->display_primaries[2][0]),
  148. av_q2d(mastering_display->display_primaries[2][1]),
  149. av_q2d(mastering_display->white_point[0]), av_q2d(mastering_display->white_point[1]),
  150. av_q2d(mastering_display->min_luminance), av_q2d(mastering_display->max_luminance));
  151. }
  152. static void dump_content_light_metadata(AVFilterContext *ctx, const AVFrameSideData *sd)
  153. {
  154. const AVContentLightMetadata *metadata = (const AVContentLightMetadata *)sd->data;
  155. av_log(ctx, AV_LOG_INFO, "Content Light Level information: "
  156. "MaxCLL=%d, MaxFALL=%d",
  157. metadata->MaxCLL, metadata->MaxFALL);
  158. }
  159. static void dump_video_enc_params(AVFilterContext *ctx, const AVFrameSideData *sd)
  160. {
  161. const AVVideoEncParams *par = (const AVVideoEncParams *)sd->data;
  162. int plane, acdc;
  163. av_log(ctx, AV_LOG_INFO, "video encoding parameters: type %d; ", par->type);
  164. if (par->qp)
  165. av_log(ctx, AV_LOG_INFO, "qp=%d; ", par->qp);
  166. for (plane = 0; plane < FF_ARRAY_ELEMS(par->delta_qp); plane++)
  167. for (acdc = 0; acdc < FF_ARRAY_ELEMS(par->delta_qp[plane]); acdc++) {
  168. int delta_qp = par->delta_qp[plane][acdc];
  169. if (delta_qp)
  170. av_log(ctx, AV_LOG_INFO, "delta_qp[%d][%d]=%d; ",
  171. plane, acdc, delta_qp);
  172. }
  173. if (par->nb_blocks)
  174. av_log(ctx, AV_LOG_INFO, "%u blocks; ", par->nb_blocks);
  175. }
  176. static void dump_sei_unregistered_metadata(AVFilterContext *ctx, const AVFrameSideData *sd)
  177. {
  178. const int uuid_size = 16;
  179. const uint8_t *user_data = sd->data;
  180. int i;
  181. if (sd->size < uuid_size) {
  182. av_log(ctx, AV_LOG_ERROR, "invalid data(%d < UUID(%d-bytes))\n", sd->size, uuid_size);
  183. return;
  184. }
  185. av_log(ctx, AV_LOG_INFO, "User Data Unregistered:\n");
  186. av_log(ctx, AV_LOG_INFO, "UUID=");
  187. for (i = 0; i < uuid_size; i++) {
  188. av_log(ctx, AV_LOG_INFO, "%02x", user_data[i]);
  189. if (i == 3 || i == 5 || i == 7 || i == 9)
  190. av_log(ctx, AV_LOG_INFO, "-");
  191. }
  192. av_log(ctx, AV_LOG_INFO, "\n");
  193. av_log(ctx, AV_LOG_INFO, "User Data=");
  194. for (; i < sd->size; i++) {
  195. av_log(ctx, AV_LOG_INFO, "%02x", user_data[i]);
  196. }
  197. av_log(ctx, AV_LOG_INFO, "\n");
  198. }
  199. static void dump_color_property(AVFilterContext *ctx, AVFrame *frame)
  200. {
  201. const char *color_range_str = av_color_range_name(frame->color_range);
  202. const char *colorspace_str = av_color_space_name(frame->colorspace);
  203. const char *color_primaries_str = av_color_primaries_name(frame->color_primaries);
  204. const char *color_trc_str = av_color_transfer_name(frame->color_trc);
  205. if (!color_range_str || frame->color_range == AVCOL_RANGE_UNSPECIFIED) {
  206. av_log(ctx, AV_LOG_INFO, "color_range:unknown");
  207. } else {
  208. av_log(ctx, AV_LOG_INFO, "color_range:%s", color_range_str);
  209. }
  210. if (!colorspace_str || frame->colorspace == AVCOL_SPC_UNSPECIFIED) {
  211. av_log(ctx, AV_LOG_INFO, " color_space:unknown");
  212. } else {
  213. av_log(ctx, AV_LOG_INFO, " color_space:%s", colorspace_str);
  214. }
  215. if (!color_primaries_str || frame->color_primaries == AVCOL_PRI_UNSPECIFIED) {
  216. av_log(ctx, AV_LOG_INFO, " color_primaries:unknown");
  217. } else {
  218. av_log(ctx, AV_LOG_INFO, " color_primaries:%s", color_primaries_str);
  219. }
  220. if (!color_trc_str || frame->color_trc == AVCOL_TRC_UNSPECIFIED) {
  221. av_log(ctx, AV_LOG_INFO, " color_trc:unknown");
  222. } else {
  223. av_log(ctx, AV_LOG_INFO, " color_trc:%s", color_trc_str);
  224. }
  225. av_log(ctx, AV_LOG_INFO, "\n");
  226. }
  227. static void update_sample_stats_8(const uint8_t *src, int len, int64_t *sum, int64_t *sum2)
  228. {
  229. int i;
  230. for (i = 0; i < len; i++) {
  231. *sum += src[i];
  232. *sum2 += src[i] * src[i];
  233. }
  234. }
  235. static void update_sample_stats_16(int be, const uint8_t *src, int len, int64_t *sum, int64_t *sum2)
  236. {
  237. const uint16_t *src1 = (const uint16_t *)src;
  238. int i;
  239. for (i = 0; i < len / 2; i++) {
  240. if ((HAVE_BIGENDIAN && !be) || (!HAVE_BIGENDIAN && be)) {
  241. *sum += av_bswap16(src1[i]);
  242. *sum2 += (uint32_t)av_bswap16(src1[i]) * (uint32_t)av_bswap16(src1[i]);
  243. } else {
  244. *sum += src1[i];
  245. *sum2 += (uint32_t)src1[i] * (uint32_t)src1[i];
  246. }
  247. }
  248. }
  249. static void update_sample_stats(int depth, int be, const uint8_t *src, int len, int64_t *sum, int64_t *sum2)
  250. {
  251. if (depth <= 8)
  252. update_sample_stats_8(src, len, sum, sum2);
  253. else
  254. update_sample_stats_16(be, src, len, sum, sum2);
  255. }
  256. static int filter_frame(AVFilterLink *inlink, AVFrame *frame)
  257. {
  258. AVFilterContext *ctx = inlink->dst;
  259. ShowInfoContext *s = ctx->priv;
  260. const AVPixFmtDescriptor *desc = av_pix_fmt_desc_get(inlink->format);
  261. uint32_t plane_checksum[4] = {0}, checksum = 0;
  262. int64_t sum[4] = {0}, sum2[4] = {0};
  263. int32_t pixelcount[4] = {0};
  264. int bitdepth = desc->comp[0].depth;
  265. int be = desc->flags & AV_PIX_FMT_FLAG_BE;
  266. int i, plane, vsub = desc->log2_chroma_h;
  267. for (plane = 0; plane < 4 && s->calculate_checksums && frame->data[plane] && frame->linesize[plane]; plane++) {
  268. uint8_t *data = frame->data[plane];
  269. int h = plane == 1 || plane == 2 ? AV_CEIL_RSHIFT(inlink->h, vsub) : inlink->h;
  270. int linesize = av_image_get_linesize(frame->format, frame->width, plane);
  271. int width = linesize >> (bitdepth > 8);
  272. if (linesize < 0)
  273. return linesize;
  274. for (i = 0; i < h; i++) {
  275. plane_checksum[plane] = av_adler32_update(plane_checksum[plane], data, linesize);
  276. checksum = av_adler32_update(checksum, data, linesize);
  277. update_sample_stats(bitdepth, be, data, linesize, sum+plane, sum2+plane);
  278. pixelcount[plane] += width;
  279. data += frame->linesize[plane];
  280. }
  281. }
  282. av_log(ctx, AV_LOG_INFO,
  283. "n:%4"PRId64" pts:%7s pts_time:%-7s pos:%9"PRId64" "
  284. "fmt:%s sar:%d/%d s:%dx%d i:%c iskey:%d type:%c ",
  285. inlink->frame_count_out,
  286. av_ts2str(frame->pts), av_ts2timestr(frame->pts, &inlink->time_base), frame->pkt_pos,
  287. desc->name,
  288. frame->sample_aspect_ratio.num, frame->sample_aspect_ratio.den,
  289. frame->width, frame->height,
  290. !frame->interlaced_frame ? 'P' : /* Progressive */
  291. frame->top_field_first ? 'T' : 'B', /* Top / Bottom */
  292. frame->key_frame,
  293. av_get_picture_type_char(frame->pict_type));
  294. if (s->calculate_checksums) {
  295. av_log(ctx, AV_LOG_INFO,
  296. "checksum:%08"PRIX32" plane_checksum:[%08"PRIX32,
  297. checksum, plane_checksum[0]);
  298. for (plane = 1; plane < 4 && frame->data[plane] && frame->linesize[plane]; plane++)
  299. av_log(ctx, AV_LOG_INFO, " %08"PRIX32, plane_checksum[plane]);
  300. av_log(ctx, AV_LOG_INFO, "] mean:[");
  301. for (plane = 0; plane < 4 && frame->data[plane] && frame->linesize[plane]; plane++)
  302. av_log(ctx, AV_LOG_INFO, "%"PRId64" ", (sum[plane] + pixelcount[plane]/2) / pixelcount[plane]);
  303. av_log(ctx, AV_LOG_INFO, "\b] stdev:[");
  304. for (plane = 0; plane < 4 && frame->data[plane] && frame->linesize[plane]; plane++)
  305. av_log(ctx, AV_LOG_INFO, "%3.1f ",
  306. sqrt((sum2[plane] - sum[plane]*(double)sum[plane]/pixelcount[plane])/pixelcount[plane]));
  307. av_log(ctx, AV_LOG_INFO, "\b]");
  308. }
  309. av_log(ctx, AV_LOG_INFO, "\n");
  310. for (i = 0; i < frame->nb_side_data; i++) {
  311. AVFrameSideData *sd = frame->side_data[i];
  312. av_log(ctx, AV_LOG_INFO, " side data - ");
  313. switch (sd->type) {
  314. case AV_FRAME_DATA_PANSCAN:
  315. av_log(ctx, AV_LOG_INFO, "pan/scan");
  316. break;
  317. case AV_FRAME_DATA_A53_CC:
  318. av_log(ctx, AV_LOG_INFO, "A/53 closed captions (%d bytes)", sd->size);
  319. break;
  320. case AV_FRAME_DATA_SPHERICAL:
  321. dump_spherical(ctx, frame, sd);
  322. break;
  323. case AV_FRAME_DATA_STEREO3D:
  324. dump_stereo3d(ctx, sd);
  325. break;
  326. case AV_FRAME_DATA_S12M_TIMECODE: {
  327. dump_s12m_timecode(ctx, inlink->frame_rate, sd);
  328. break;
  329. }
  330. case AV_FRAME_DATA_DISPLAYMATRIX:
  331. av_log(ctx, AV_LOG_INFO, "displaymatrix: rotation of %.2f degrees",
  332. av_display_rotation_get((int32_t *)sd->data));
  333. break;
  334. case AV_FRAME_DATA_AFD:
  335. av_log(ctx, AV_LOG_INFO, "afd: value of %"PRIu8, sd->data[0]);
  336. break;
  337. case AV_FRAME_DATA_REGIONS_OF_INTEREST:
  338. dump_roi(ctx, sd);
  339. break;
  340. case AV_FRAME_DATA_MASTERING_DISPLAY_METADATA:
  341. dump_mastering_display(ctx, sd);
  342. break;
  343. case AV_FRAME_DATA_CONTENT_LIGHT_LEVEL:
  344. dump_content_light_metadata(ctx, sd);
  345. break;
  346. case AV_FRAME_DATA_GOP_TIMECODE: {
  347. char tcbuf[AV_TIMECODE_STR_SIZE];
  348. av_timecode_make_mpeg_tc_string(tcbuf, *(int64_t *)(sd->data));
  349. av_log(ctx, AV_LOG_INFO, "GOP timecode - %s", tcbuf);
  350. break;
  351. }
  352. case AV_FRAME_DATA_VIDEO_ENC_PARAMS:
  353. dump_video_enc_params(ctx, sd);
  354. break;
  355. case AV_FRAME_DATA_SEI_UNREGISTERED:
  356. dump_sei_unregistered_metadata(ctx, sd);
  357. break;
  358. default:
  359. av_log(ctx, AV_LOG_WARNING, "unknown side data type %d (%d bytes)\n",
  360. sd->type, sd->size);
  361. break;
  362. }
  363. av_log(ctx, AV_LOG_INFO, "\n");
  364. }
  365. dump_color_property(ctx, frame);
  366. return ff_filter_frame(inlink->dst->outputs[0], frame);
  367. }
  368. static int config_props(AVFilterContext *ctx, AVFilterLink *link, int is_out)
  369. {
  370. av_log(ctx, AV_LOG_INFO, "config %s time_base: %d/%d, frame_rate: %d/%d\n",
  371. is_out ? "out" : "in",
  372. link->time_base.num, link->time_base.den,
  373. link->frame_rate.num, link->frame_rate.den);
  374. return 0;
  375. }
  376. static int config_props_in(AVFilterLink *link)
  377. {
  378. AVFilterContext *ctx = link->dst;
  379. return config_props(ctx, link, 0);
  380. }
  381. static int config_props_out(AVFilterLink *link)
  382. {
  383. AVFilterContext *ctx = link->src;
  384. return config_props(ctx, link, 1);
  385. }
  386. static const AVFilterPad avfilter_vf_showinfo_inputs[] = {
  387. {
  388. .name = "default",
  389. .type = AVMEDIA_TYPE_VIDEO,
  390. .filter_frame = filter_frame,
  391. .config_props = config_props_in,
  392. },
  393. { NULL }
  394. };
  395. static const AVFilterPad avfilter_vf_showinfo_outputs[] = {
  396. {
  397. .name = "default",
  398. .type = AVMEDIA_TYPE_VIDEO,
  399. .config_props = config_props_out,
  400. },
  401. { NULL }
  402. };
  403. AVFilter ff_vf_showinfo = {
  404. .name = "showinfo",
  405. .description = NULL_IF_CONFIG_SMALL("Show textual information for each video frame."),
  406. .inputs = avfilter_vf_showinfo_inputs,
  407. .outputs = avfilter_vf_showinfo_outputs,
  408. .priv_size = sizeof(ShowInfoContext),
  409. .priv_class = &showinfo_class,
  410. };