| Line | Branch | Exec | Source |
|---|---|---|---|
| 1 | /* | ||
| 2 | * Avid DNxUncomressed / SMPTE RDD 50 parser | ||
| 3 | * Copyright (c) 2024 Martin Schitter | ||
| 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 | This parser for DNxUncompressed video data is mostly based on | ||
| 24 | reverse engineering of output generated by DaVinci Resolve 19 | ||
| 25 | but was later also checked against the SMPTE RDD 50 specification. | ||
| 26 | |||
| 27 | Limitations: Multiple image planes are not supported. | ||
| 28 | */ | ||
| 29 | |||
| 30 | #include "avcodec.h" | ||
| 31 | #include "libavutil/intreadwrite.h" | ||
| 32 | |||
| 33 | typedef struct DNxUcParseContext { | ||
| 34 | uint32_t fourcc_tag; | ||
| 35 | uint32_t width; | ||
| 36 | uint32_t height; | ||
| 37 | uint32_t nr_bytes; | ||
| 38 | } DNxUcParseContext; | ||
| 39 | |||
| 40 | /* | ||
| 41 | DNxUncompressed frame data comes wrapped in nested boxes of metadata | ||
| 42 | (box structure: len + fourcc marker + data): | ||
| 43 | |||
| 44 | [0-4] len of outer essence unit box (typically 37 bytes of header + frame data) | ||
| 45 | [4-7] fourcc 'pack' | ||
| 46 | |||
| 47 | [8-11] len of "signal info box" (always 21) | ||
| 48 | [12-15] fourcc 'sinf' | ||
| 49 | [16-19] frame width / line packing size | ||
| 50 | [20-23] frame hight / nr of lines | ||
| 51 | [24-27] fourcc pixel format indicator | ||
| 52 | [28] frame_layout (0: progressive, 1: interlaced) | ||
| 53 | |||
| 54 | [29-32] len of "signal data box" (nr of frame data bytes + 8) | ||
| 55 | [33-36] fourcc 'sdat' | ||
| 56 | [37-..] frame data | ||
| 57 | |||
| 58 | A sequence of 'signal info'+'signal data' box pairs wrapped in | ||
| 59 | 'icmp'(=image component) boxes can be utilized to compose more | ||
| 60 | complex multi plane images. | ||
| 61 | This feature is only partially supported in the present implementation. | ||
| 62 | We never pick more than the first pair of info and image data enclosed | ||
| 63 | in this way. | ||
| 64 | */ | ||
| 65 | |||
| 66 | ✗ | static int dnxuc_parse(AVCodecParserContext *s, | |
| 67 | AVCodecContext *avctx, | ||
| 68 | const uint8_t **poutbuf, int *poutbuf_size, | ||
| 69 | const uint8_t *buf, int buf_size) | ||
| 70 | { | ||
| 71 | ✗ | const int HEADER_SIZE = 37; | |
| 72 | ✗ | int icmp_offset = 0; | |
| 73 | |||
| 74 | DNxUcParseContext *pc; | ||
| 75 | ✗ | pc = (DNxUcParseContext *) s->priv_data; | |
| 76 | |||
| 77 | ✗ | if (!buf_size) { | |
| 78 | ✗ | return 0; | |
| 79 | } | ||
| 80 | ✗ | if (buf_size > 16 && MKTAG('i','c','m','p') == AV_RL32(buf+12)){ | |
| 81 | ✗ | icmp_offset += 8; | |
| 82 | } | ||
| 83 | ✗ | if ( buf_size < 37 + icmp_offset /* check metadata structure expectations */ | |
| 84 | ✗ | || MKTAG('p','a','c','k') != AV_RL32(buf+4+icmp_offset) | |
| 85 | ✗ | || MKTAG('s','i','n','f') != AV_RL32(buf+12+icmp_offset) | |
| 86 | ✗ | || MKTAG('s','d','a','t') != AV_RL32(buf+33+icmp_offset)){ | |
| 87 | ✗ | av_log(avctx, AV_LOG_ERROR, "can't read DNxUncompressed metadata.\n"); | |
| 88 | ✗ | *poutbuf_size = 0; | |
| 89 | ✗ | return buf_size; | |
| 90 | } | ||
| 91 | |||
| 92 | ✗ | pc->fourcc_tag = AV_RL32(buf+24+icmp_offset); | |
| 93 | ✗ | pc->width = AV_RL32(buf+16+icmp_offset); | |
| 94 | ✗ | pc->height = AV_RL32(buf+20+icmp_offset); | |
| 95 | ✗ | pc->nr_bytes = AV_RL32(buf+29+icmp_offset) - 8; | |
| 96 | |||
| 97 | ✗ | if (!avctx->codec_tag) { | |
| 98 | ✗ | av_log(avctx, AV_LOG_INFO, "dnxuc_parser: '%s' %dx%d %dbpp %d\n", | |
| 99 | ✗ | av_fourcc2str(pc->fourcc_tag), | |
| 100 | pc->width, pc->height, | ||
| 101 | ✗ | (pc->nr_bytes*8)/(pc->width*pc->height), | |
| 102 | pc->nr_bytes); | ||
| 103 | ✗ | avctx->codec_tag = pc->fourcc_tag; | |
| 104 | } | ||
| 105 | |||
| 106 | ✗ | if (pc->nr_bytes > buf_size - HEADER_SIZE + icmp_offset){ | |
| 107 | ✗ | av_log(avctx, AV_LOG_ERROR, "Insufficient size of image essence data.\n"); | |
| 108 | ✗ | *poutbuf_size = 0; | |
| 109 | ✗ | return buf_size; | |
| 110 | } | ||
| 111 | |||
| 112 | ✗ | *poutbuf = buf + HEADER_SIZE + icmp_offset; | |
| 113 | ✗ | *poutbuf_size = pc->nr_bytes; | |
| 114 | |||
| 115 | ✗ | return buf_size; | |
| 116 | } | ||
| 117 | |||
| 118 | const AVCodecParser ff_dnxuc_parser = { | ||
| 119 | .codec_ids = { AV_CODEC_ID_DNXUC }, | ||
| 120 | .priv_data_size = sizeof(DNxUcParseContext), | ||
| 121 | .parser_parse = dnxuc_parse, | ||
| 122 | }; | ||
| 123 |