libavcodec/av1_parser.c

   1 /*
   2  * AV1 parser
   3  *
   4  * Copyright (C) 2018 James Almer <jamrial@gmail.com>
   5  *
   6  * This file is part of FFmpeg.
   7  *
   8  * FFmpeg is free software; you can redistribute it and/or
   9  * modify it under the terms of the GNU Lesser General Public
  10  * License as published by the Free Software Foundation; either
  11  * version 2.1 of the License, or (at your option) any later version.
  12  *
  13  * FFmpeg is distributed in the hope that it will be useful,
  14  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  15  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
  16  * Lesser General Public License for more details.
  17  *
  18  * You should have received a copy of the GNU Lesser General Public
  19  * License along with FFmpeg; if not, write to the Free Software
  20  * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
  21  */
  22
  23 #include "libavutil/avassert.h"
  24 #include "cbs.h"
  25 #include "cbs_av1.h"
  26 #include "parser.h"
  27
  28 typedef struct AV1ParseContext {
  29     CodedBitstreamContext *cbc;
  30     CodedBitstreamFragment temporal_unit;
  31     int parsed_extradata;
  32 } AV1ParseContext;
  33
  34 static const enum AVPixelFormat pix_fmts_8bit[2][2] = {
  35     { AV_PIX_FMT_YUV444P, AV_PIX_FMT_NONE },
  36     { AV_PIX_FMT_YUV422P, AV_PIX_FMT_YUV420P },
  37 };
  38 static const enum AVPixelFormat pix_fmts_10bit[2][2] = {
  39     { AV_PIX_FMT_YUV444P10, AV_PIX_FMT_NONE },
  40     { AV_PIX_FMT_YUV422P10, AV_PIX_FMT_YUV420P10 },
  41 };
  42 static const enum AVPixelFormat pix_fmts_12bit[2][2] = {
  43     { AV_PIX_FMT_YUV444P12, AV_PIX_FMT_NONE },
  44     { AV_PIX_FMT_YUV422P12, AV_PIX_FMT_YUV420P12 },
  45 };
  46
  47 static const enum AVPixelFormat pix_fmts_rgb[3] = {
  48     AV_PIX_FMT_GBRP, AV_PIX_FMT_GBRP10, AV_PIX_FMT_GBRP12,
  49 };
  50
  51 static int av1_parser_parse(AVCodecParserContext *ctx,
  52                             AVCodecContext *avctx,
  53                             const uint8_t **out_data, int *out_size,
  54                             const uint8_t *data, int size)
  55 {
  56     AV1ParseContext *s = ctx->priv_data;
  57     CodedBitstreamFragment *td = &s->temporal_unit;
  58     const CodedBitstreamAV1Context *av1 = s->cbc->priv_data;
  59     const AV1RawSequenceHeader *seq;
  60     const AV1RawColorConfig *color;
  61     int ret;
  62
  63     *out_data = data;
  64     *out_size = size;
  65
  66     ctx->key_frame         = -1;
  67     ctx->pict_type         = AV_PICTURE_TYPE_NONE;
  68     ctx->picture_structure = AV_PICTURE_STRUCTURE_UNKNOWN;
  69
  70     s->cbc->log_ctx = avctx;
  71
  72     if (avctx->extradata_size && !s->parsed_extradata) {
  73         s->parsed_extradata = 1;
  74
  75         ret = ff_cbs_read_extradata_from_codec(s->cbc, td, avctx);
  76         if (ret < 0) {
  77             av_log(avctx, AV_LOG_WARNING, "Failed to parse extradata.\n");
  78         }
  79
  80         ff_cbs_fragment_reset(td);
  81     }
  82
  83     ret = ff_cbs_read(s->cbc, td, data, size);
  84     if (ret < 0) {
  85         av_log(avctx, AV_LOG_ERROR, "Failed to parse temporal unit.\n");
  86         goto end;
  87     }
  88
  89     if (!av1->sequence_header) {
  90         av_log(avctx, AV_LOG_ERROR, "No sequence header available\n");
  91         goto end;
  92     }
  93
  94     seq = av1->sequence_header;
  95     color = &seq->color_config;
  96
  97     for (int i = 0; i < td->nb_units; i++) {
  98         const CodedBitstreamUnit *unit = &td->units[i];
  99         const AV1RawOBU *obu = unit->content;
 100         const AV1RawFrameHeader *frame;
 101
 102         if (unit->type == AV1_OBU_FRAME)
 103             frame = &obu->obu.frame.header;
 104         else if (unit->type == AV1_OBU_FRAME_HEADER)
 105             frame = &obu->obu.frame_header;
 106         else
 107             continue;
 108
 109         if (obu->header.spatial_id > 0)
 110             continue;
 111
 112         if (!frame->show_frame && !frame->show_existing_frame)
 113             continue;
 114
 115         ctx->width  = frame->frame_width_minus_1 + 1;
 116         ctx->height = frame->frame_height_minus_1 + 1;
 117
 118         ctx->key_frame = frame->frame_type == AV1_FRAME_KEY && !frame->show_existing_frame;
 119
 120         switch (frame->frame_type) {
 121         case AV1_FRAME_KEY:
 122         case AV1_FRAME_INTRA_ONLY:
 123             ctx->pict_type = AV_PICTURE_TYPE_I;
 124             break;
 125         case AV1_FRAME_INTER:
 126             ctx->pict_type = AV_PICTURE_TYPE_P;
 127             break;
 128         case AV1_FRAME_SWITCH:
 129             ctx->pict_type = AV_PICTURE_TYPE_SP;
 130             break;
 131         }
 132         ctx->picture_structure = AV_PICTURE_STRUCTURE_FRAME;
 133     }
 134
 135     switch (av1->bit_depth) {
 136     case 8:
 137         ctx->format = color->mono_chrome ? AV_PIX_FMT_GRAY8
 138                                          : pix_fmts_8bit [color->subsampling_x][color->subsampling_y];
 139         break;
 140     case 10:
 141         ctx->format = color->mono_chrome ? AV_PIX_FMT_GRAY10
 142                                          : pix_fmts_10bit[color->subsampling_x][color->subsampling_y];
 143         break;
 144     case 12:
 145         ctx->format = color->mono_chrome ? AV_PIX_FMT_GRAY12
 146                                          : pix_fmts_12bit[color->subsampling_x][color->subsampling_y];
 147         break;
 148     }
 149     av_assert2(ctx->format != AV_PIX_FMT_NONE);
 150
 151     if (!color->subsampling_x && !color->subsampling_y &&
 152         color->matrix_coefficients       == AVCOL_SPC_RGB &&
 153         color->color_primaries           == AVCOL_PRI_BT709 &&
 154         color->transfer_characteristics  == AVCOL_TRC_IEC61966_2_1)
 155         ctx->format = pix_fmts_rgb[color->high_bitdepth + color->twelve_bit];
 156
 157     avctx->profile = seq->seq_profile;
 158     avctx->level   = seq->seq_level_idx[0];
 159
 160     avctx->colorspace = (enum AVColorSpace) color->matrix_coefficients;
 161     avctx->color_primaries = (enum AVColorPrimaries) color->color_primaries;
 162     avctx->color_trc = (enum AVColorTransferCharacteristic) color->transfer_characteristics;
 163     avctx->color_range = color->color_range ? AVCOL_RANGE_JPEG : AVCOL_RANGE_MPEG;
 164
 165     if (avctx->framerate.num)
 166         avctx->time_base = av_inv_q(av_mul_q(avctx->framerate, (AVRational){avctx->ticks_per_frame, 1}));
 167
 168 end:
 169     ff_cbs_fragment_reset(td);
 170
 171     s->cbc->log_ctx = NULL;
 172
 173     return size;
 174 }
 175
 176 static const CodedBitstreamUnitType decompose_unit_types[] = {
 177     AV1_OBU_TEMPORAL_DELIMITER,
 178     AV1_OBU_SEQUENCE_HEADER,
 179     AV1_OBU_FRAME_HEADER,
 180     AV1_OBU_TILE_GROUP,
 181     AV1_OBU_FRAME,
 182 };
 183
 184 static av_cold int av1_parser_init(AVCodecParserContext *ctx)
 185 {
 186     AV1ParseContext *s = ctx->priv_data;
 187     int ret;
 188
 189     ret = ff_cbs_init(&s->cbc, AV_CODEC_ID_AV1, NULL);
 190     if (ret < 0)
 191         return ret;
 192
 193     s->cbc->decompose_unit_types    = decompose_unit_types;
 194     s->cbc->nb_decompose_unit_types = FF_ARRAY_ELEMS(decompose_unit_types);
 195
 196     return 0;
 197 }
 198
 199 static void av1_parser_close(AVCodecParserContext *ctx)
 200 {
 201     AV1ParseContext *s = ctx->priv_data;
 202
 203     ff_cbs_fragment_free(&s->temporal_unit);
 204     ff_cbs_close(&s->cbc);
 205 }
 206
 207 const AVCodecParser ff_av1_parser = {
 208     .codec_ids      = { AV_CODEC_ID_AV1 },
 209     .priv_data_size = sizeof(AV1ParseContext),
 210     .parser_init    = av1_parser_init,
 211     .parser_close   = av1_parser_close,
 212     .parser_parse   = av1_parser_parse,
 213 };