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.

442 lines
16KB

  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, AVFrameSideData *sd)
  52. {
  53. AVSphericalMapping *spherical = (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");
  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");
  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, AVFrameSideData *sd)
  86. {
  87. 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");
  91. return;
  92. }
  93. stereo = (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_roi(AVFilterContext *ctx, AVFrameSideData *sd)
  99. {
  100. int nb_rois;
  101. const AVRegionOfInterest *roi;
  102. uint32_t roi_size;
  103. roi = (const AVRegionOfInterest *)sd->data;
  104. roi_size = roi->self_size;
  105. if (!roi_size || sd->size % roi_size != 0) {
  106. av_log(ctx, AV_LOG_ERROR, "Invalid AVRegionOfInterest.self_size.");
  107. return;
  108. }
  109. nb_rois = sd->size / roi_size;
  110. av_log(ctx, AV_LOG_INFO, "Regions Of Interest(RoI) information: ");
  111. for (int i = 0; i < nb_rois; i++) {
  112. roi = (const AVRegionOfInterest *)(sd->data + roi_size * i);
  113. av_log(ctx, AV_LOG_INFO, "index: %d, region: (%d, %d)/(%d, %d), qp offset: %d/%d.\n",
  114. i, roi->left, roi->top, roi->right, roi->bottom, roi->qoffset.num, roi->qoffset.den);
  115. }
  116. }
  117. static void dump_mastering_display(AVFilterContext *ctx, AVFrameSideData *sd)
  118. {
  119. AVMasteringDisplayMetadata *mastering_display;
  120. av_log(ctx, AV_LOG_INFO, "mastering display: ");
  121. if (sd->size < sizeof(*mastering_display)) {
  122. av_log(ctx, AV_LOG_ERROR, "invalid data");
  123. return;
  124. }
  125. mastering_display = (AVMasteringDisplayMetadata *)sd->data;
  126. av_log(ctx, AV_LOG_INFO, "has_primaries:%d has_luminance:%d "
  127. "r(%5.4f,%5.4f) g(%5.4f,%5.4f) b(%5.4f %5.4f) wp(%5.4f, %5.4f) "
  128. "min_luminance=%f, max_luminance=%f",
  129. mastering_display->has_primaries, mastering_display->has_luminance,
  130. av_q2d(mastering_display->display_primaries[0][0]),
  131. av_q2d(mastering_display->display_primaries[0][1]),
  132. av_q2d(mastering_display->display_primaries[1][0]),
  133. av_q2d(mastering_display->display_primaries[1][1]),
  134. av_q2d(mastering_display->display_primaries[2][0]),
  135. av_q2d(mastering_display->display_primaries[2][1]),
  136. av_q2d(mastering_display->white_point[0]), av_q2d(mastering_display->white_point[1]),
  137. av_q2d(mastering_display->min_luminance), av_q2d(mastering_display->max_luminance));
  138. }
  139. static void dump_content_light_metadata(AVFilterContext *ctx, AVFrameSideData *sd)
  140. {
  141. AVContentLightMetadata* metadata = (AVContentLightMetadata*)sd->data;
  142. av_log(ctx, AV_LOG_INFO, "Content Light Level information: "
  143. "MaxCLL=%d, MaxFALL=%d",
  144. metadata->MaxCLL, metadata->MaxFALL);
  145. }
  146. static void dump_video_enc_params(AVFilterContext *ctx, AVFrameSideData *sd)
  147. {
  148. AVVideoEncParams *par = (AVVideoEncParams*)sd->data;
  149. int plane, acdc;
  150. av_log(ctx, AV_LOG_INFO, "video encoding parameters: type %d; ", par->type);
  151. if (par->qp)
  152. av_log(ctx, AV_LOG_INFO, "qp=%d; ", par->qp);
  153. for (plane = 0; plane < FF_ARRAY_ELEMS(par->delta_qp); plane++)
  154. for (acdc = 0; acdc < FF_ARRAY_ELEMS(par->delta_qp[plane]); acdc++) {
  155. int delta_qp = par->delta_qp[plane][acdc];
  156. if (delta_qp)
  157. av_log(ctx, AV_LOG_INFO, "delta_qp[%d][%d]=%d; ",
  158. plane, acdc, delta_qp);
  159. }
  160. if (par->nb_blocks)
  161. av_log(ctx, AV_LOG_INFO, "%u blocks; ", par->nb_blocks);
  162. }
  163. static void dump_color_property(AVFilterContext *ctx, AVFrame *frame)
  164. {
  165. const char *color_range_str = av_color_range_name(frame->color_range);
  166. const char *colorspace_str = av_color_space_name(frame->colorspace);
  167. const char *color_primaries_str = av_color_primaries_name(frame->color_primaries);
  168. const char *color_trc_str = av_color_transfer_name(frame->color_trc);
  169. if (!color_range_str || frame->color_range == AVCOL_RANGE_UNSPECIFIED) {
  170. av_log(ctx, AV_LOG_INFO, "color_range:unknown");
  171. } else {
  172. av_log(ctx, AV_LOG_INFO, "color_range:%s", color_range_str);
  173. }
  174. if (!colorspace_str || frame->colorspace == AVCOL_SPC_UNSPECIFIED) {
  175. av_log(ctx, AV_LOG_INFO, " color_space:unknown");
  176. } else {
  177. av_log(ctx, AV_LOG_INFO, " color_space:%s", colorspace_str);
  178. }
  179. if (!color_primaries_str || frame->color_primaries == AVCOL_PRI_UNSPECIFIED) {
  180. av_log(ctx, AV_LOG_INFO, " color_primaries:unknown");
  181. } else {
  182. av_log(ctx, AV_LOG_INFO, " color_primaries:%s", color_primaries_str);
  183. }
  184. if (!color_trc_str || frame->color_trc == AVCOL_TRC_UNSPECIFIED) {
  185. av_log(ctx, AV_LOG_INFO, " color_trc:unknown");
  186. } else {
  187. av_log(ctx, AV_LOG_INFO, " color_trc:%s", color_trc_str);
  188. }
  189. av_log(ctx, AV_LOG_INFO, "\n");
  190. }
  191. static void update_sample_stats_8(const uint8_t *src, int len, int64_t *sum, int64_t *sum2)
  192. {
  193. int i;
  194. for (i = 0; i < len; i++) {
  195. *sum += src[i];
  196. *sum2 += src[i] * src[i];
  197. }
  198. }
  199. static void update_sample_stats_16(int be, const uint8_t *src, int len, int64_t *sum, int64_t *sum2)
  200. {
  201. const uint16_t *src1 = (const uint16_t *)src;
  202. int i;
  203. for (i = 0; i < len / 2; i++) {
  204. if ((HAVE_BIGENDIAN && !be) || (!HAVE_BIGENDIAN && be)) {
  205. *sum += av_bswap16(src1[i]);
  206. *sum2 += (uint32_t)av_bswap16(src1[i]) * (uint32_t)av_bswap16(src1[i]);
  207. } else {
  208. *sum += src1[i];
  209. *sum2 += (uint32_t)src1[i] * (uint32_t)src1[i];
  210. }
  211. }
  212. }
  213. static void update_sample_stats(int depth, int be, const uint8_t *src, int len, int64_t *sum, int64_t *sum2)
  214. {
  215. if (depth <= 8)
  216. update_sample_stats_8(src, len, sum, sum2);
  217. else
  218. update_sample_stats_16(be, src, len, sum, sum2);
  219. }
  220. static int filter_frame(AVFilterLink *inlink, AVFrame *frame)
  221. {
  222. AVFilterContext *ctx = inlink->dst;
  223. ShowInfoContext *s = ctx->priv;
  224. const AVPixFmtDescriptor *desc = av_pix_fmt_desc_get(inlink->format);
  225. uint32_t plane_checksum[4] = {0}, checksum = 0;
  226. int64_t sum[4] = {0}, sum2[4] = {0};
  227. int32_t pixelcount[4] = {0};
  228. int bitdepth = desc->comp[0].depth;
  229. int be = desc->flags & AV_PIX_FMT_FLAG_BE;
  230. int i, plane, vsub = desc->log2_chroma_h;
  231. for (plane = 0; plane < 4 && s->calculate_checksums && frame->data[plane] && frame->linesize[plane]; plane++) {
  232. uint8_t *data = frame->data[plane];
  233. int h = plane == 1 || plane == 2 ? AV_CEIL_RSHIFT(inlink->h, vsub) : inlink->h;
  234. int linesize = av_image_get_linesize(frame->format, frame->width, plane);
  235. int width = linesize >> (bitdepth > 8);
  236. if (linesize < 0)
  237. return linesize;
  238. for (i = 0; i < h; i++) {
  239. plane_checksum[plane] = av_adler32_update(plane_checksum[plane], data, linesize);
  240. checksum = av_adler32_update(checksum, data, linesize);
  241. update_sample_stats(bitdepth, be, data, linesize, sum+plane, sum2+plane);
  242. pixelcount[plane] += width;
  243. data += frame->linesize[plane];
  244. }
  245. }
  246. av_log(ctx, AV_LOG_INFO,
  247. "n:%4"PRId64" pts:%7s pts_time:%-7s pos:%9"PRId64" "
  248. "fmt:%s sar:%d/%d s:%dx%d i:%c iskey:%d type:%c ",
  249. inlink->frame_count_out,
  250. av_ts2str(frame->pts), av_ts2timestr(frame->pts, &inlink->time_base), frame->pkt_pos,
  251. desc->name,
  252. frame->sample_aspect_ratio.num, frame->sample_aspect_ratio.den,
  253. frame->width, frame->height,
  254. !frame->interlaced_frame ? 'P' : /* Progressive */
  255. frame->top_field_first ? 'T' : 'B', /* Top / Bottom */
  256. frame->key_frame,
  257. av_get_picture_type_char(frame->pict_type));
  258. if (s->calculate_checksums) {
  259. av_log(ctx, AV_LOG_INFO,
  260. "checksum:%08"PRIX32" plane_checksum:[%08"PRIX32,
  261. checksum, plane_checksum[0]);
  262. for (plane = 1; plane < 4 && frame->data[plane] && frame->linesize[plane]; plane++)
  263. av_log(ctx, AV_LOG_INFO, " %08"PRIX32, plane_checksum[plane]);
  264. av_log(ctx, AV_LOG_INFO, "] mean:[");
  265. for (plane = 0; plane < 4 && frame->data[plane] && frame->linesize[plane]; plane++)
  266. av_log(ctx, AV_LOG_INFO, "%"PRId64" ", (sum[plane] + pixelcount[plane]/2) / pixelcount[plane]);
  267. av_log(ctx, AV_LOG_INFO, "\b] stdev:[");
  268. for (plane = 0; plane < 4 && frame->data[plane] && frame->linesize[plane]; plane++)
  269. av_log(ctx, AV_LOG_INFO, "%3.1f ",
  270. sqrt((sum2[plane] - sum[plane]*(double)sum[plane]/pixelcount[plane])/pixelcount[plane]));
  271. av_log(ctx, AV_LOG_INFO, "\b]");
  272. }
  273. av_log(ctx, AV_LOG_INFO, "\n");
  274. for (i = 0; i < frame->nb_side_data; i++) {
  275. AVFrameSideData *sd = frame->side_data[i];
  276. av_log(ctx, AV_LOG_INFO, " side data - ");
  277. switch (sd->type) {
  278. case AV_FRAME_DATA_PANSCAN:
  279. av_log(ctx, AV_LOG_INFO, "pan/scan");
  280. break;
  281. case AV_FRAME_DATA_A53_CC:
  282. av_log(ctx, AV_LOG_INFO, "A/53 closed captions (%d bytes)", sd->size);
  283. break;
  284. case AV_FRAME_DATA_SPHERICAL:
  285. dump_spherical(ctx, frame, sd);
  286. break;
  287. case AV_FRAME_DATA_STEREO3D:
  288. dump_stereo3d(ctx, sd);
  289. break;
  290. case AV_FRAME_DATA_S12M_TIMECODE: {
  291. uint32_t *tc = (uint32_t*)sd->data;
  292. int m = FFMIN(tc[0],3);
  293. if (sd->size != 16) {
  294. av_log(ctx, AV_LOG_ERROR, "invalid data");
  295. break;
  296. }
  297. for (int j = 1; j <= m; j++) {
  298. char tcbuf[AV_TIMECODE_STR_SIZE];
  299. av_timecode_make_smpte_tc_string(tcbuf, tc[j], 0);
  300. av_log(ctx, AV_LOG_INFO, "timecode - %s%s", tcbuf, j != m ? ", " : "");
  301. }
  302. break;
  303. }
  304. case AV_FRAME_DATA_DISPLAYMATRIX:
  305. av_log(ctx, AV_LOG_INFO, "displaymatrix: rotation of %.2f degrees",
  306. av_display_rotation_get((int32_t *)sd->data));
  307. break;
  308. case AV_FRAME_DATA_AFD:
  309. av_log(ctx, AV_LOG_INFO, "afd: value of %"PRIu8, sd->data[0]);
  310. break;
  311. case AV_FRAME_DATA_REGIONS_OF_INTEREST:
  312. dump_roi(ctx, sd);
  313. break;
  314. case AV_FRAME_DATA_MASTERING_DISPLAY_METADATA:
  315. dump_mastering_display(ctx, sd);
  316. break;
  317. case AV_FRAME_DATA_CONTENT_LIGHT_LEVEL:
  318. dump_content_light_metadata(ctx, sd);
  319. break;
  320. case AV_FRAME_DATA_GOP_TIMECODE: {
  321. char tcbuf[AV_TIMECODE_STR_SIZE];
  322. av_timecode_make_mpeg_tc_string(tcbuf, *(int64_t *)(sd->data));
  323. av_log(ctx, AV_LOG_INFO, "GOP timecode - %s", tcbuf);
  324. break;
  325. }
  326. case AV_FRAME_DATA_VIDEO_ENC_PARAMS:
  327. dump_video_enc_params(ctx, sd);
  328. break;
  329. default:
  330. av_log(ctx, AV_LOG_WARNING, "unknown side data type %d (%d bytes)",
  331. sd->type, sd->size);
  332. break;
  333. }
  334. av_log(ctx, AV_LOG_INFO, "\n");
  335. }
  336. dump_color_property(ctx, frame);
  337. return ff_filter_frame(inlink->dst->outputs[0], frame);
  338. }
  339. static int config_props(AVFilterContext *ctx, AVFilterLink *link, int is_out)
  340. {
  341. av_log(ctx, AV_LOG_INFO, "config %s time_base: %d/%d, frame_rate: %d/%d\n",
  342. is_out ? "out" : "in",
  343. link->time_base.num, link->time_base.den,
  344. link->frame_rate.num, link->frame_rate.den);
  345. return 0;
  346. }
  347. static int config_props_in(AVFilterLink *link)
  348. {
  349. AVFilterContext *ctx = link->dst;
  350. return config_props(ctx, link, 0);
  351. }
  352. static int config_props_out(AVFilterLink *link)
  353. {
  354. AVFilterContext *ctx = link->src;
  355. return config_props(ctx, link, 1);
  356. }
  357. static const AVFilterPad avfilter_vf_showinfo_inputs[] = {
  358. {
  359. .name = "default",
  360. .type = AVMEDIA_TYPE_VIDEO,
  361. .filter_frame = filter_frame,
  362. .config_props = config_props_in,
  363. },
  364. { NULL }
  365. };
  366. static const AVFilterPad avfilter_vf_showinfo_outputs[] = {
  367. {
  368. .name = "default",
  369. .type = AVMEDIA_TYPE_VIDEO,
  370. .config_props = config_props_out,
  371. },
  372. { NULL }
  373. };
  374. AVFilter ff_vf_showinfo = {
  375. .name = "showinfo",
  376. .description = NULL_IF_CONFIG_SMALL("Show textual information for each video frame."),
  377. .inputs = avfilter_vf_showinfo_inputs,
  378. .outputs = avfilter_vf_showinfo_outputs,
  379. .priv_size = sizeof(ShowInfoContext),
  380. .priv_class = &showinfo_class,
  381. };