blob: 43ca4821cb1a3c841b25fd7f981a9d1d8243afba
1 | /* |
2 | * LATM/LOAS muxer |
3 | * Copyright (c) 2011 Kieran Kunhya <kieran@kunhya.com> |
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 | #include "libavcodec/get_bits.h" |
23 | #include "libavcodec/put_bits.h" |
24 | #include "libavcodec/avcodec.h" |
25 | #include "libavcodec/mpeg4audio.h" |
26 | #include "libavutil/opt.h" |
27 | #include "avformat.h" |
28 | #include "internal.h" |
29 | #include "rawenc.h" |
30 | |
31 | #define MAX_EXTRADATA_SIZE 1024 |
32 | |
33 | typedef struct LATMContext { |
34 | AVClass *av_class; |
35 | int off; |
36 | int channel_conf; |
37 | int object_type; |
38 | int counter; |
39 | int mod; |
40 | uint8_t buffer[0x1fff + MAX_EXTRADATA_SIZE + 1024]; |
41 | } LATMContext; |
42 | |
43 | static const AVOption options[] = { |
44 | {"smc-interval", "StreamMuxConfig interval.", |
45 | offsetof(LATMContext, mod), AV_OPT_TYPE_INT, {.i64 = 0x0014}, 0x0001, 0xffff, AV_OPT_FLAG_ENCODING_PARAM}, |
46 | {NULL}, |
47 | }; |
48 | |
49 | static const AVClass latm_muxer_class = { |
50 | .class_name = "LATM/LOAS muxer", |
51 | .item_name = av_default_item_name, |
52 | .option = options, |
53 | .version = LIBAVUTIL_VERSION_INT, |
54 | }; |
55 | |
56 | static int latm_decode_extradata(LATMContext *ctx, uint8_t *buf, int size) |
57 | { |
58 | MPEG4AudioConfig m4ac; |
59 | |
60 | if (size > MAX_EXTRADATA_SIZE) { |
61 | av_log(ctx, AV_LOG_ERROR, "Extradata is larger than currently supported.\n"); |
62 | return AVERROR_INVALIDDATA; |
63 | } |
64 | ctx->off = avpriv_mpeg4audio_get_config(&m4ac, buf, size * 8, 1); |
65 | if (ctx->off < 0) |
66 | return ctx->off; |
67 | |
68 | if (ctx->object_type == AOT_ALS && (ctx->off & 7)) { |
69 | // as long as avpriv_mpeg4audio_get_config works correctly this is impossible |
70 | av_log(ctx, AV_LOG_ERROR, "BUG: ALS offset is not byte-aligned\n"); |
71 | return AVERROR_INVALIDDATA; |
72 | } |
73 | /* FIXME: are any formats not allowed in LATM? */ |
74 | |
75 | if (m4ac.object_type > AOT_SBR && m4ac.object_type != AOT_ALS) { |
76 | av_log(ctx, AV_LOG_ERROR, "Muxing MPEG-4 AOT %d in LATM is not supported\n", m4ac.object_type); |
77 | return AVERROR_INVALIDDATA; |
78 | } |
79 | ctx->channel_conf = m4ac.chan_config; |
80 | ctx->object_type = m4ac.object_type; |
81 | |
82 | return 0; |
83 | } |
84 | |
85 | static int latm_write_header(AVFormatContext *s) |
86 | { |
87 | LATMContext *ctx = s->priv_data; |
88 | AVCodecParameters *par = s->streams[0]->codecpar; |
89 | |
90 | if (par->codec_id == AV_CODEC_ID_AAC_LATM) |
91 | return 0; |
92 | |
93 | if (par->extradata_size > 0 && |
94 | latm_decode_extradata(ctx, par->extradata, par->extradata_size) < 0) |
95 | return AVERROR_INVALIDDATA; |
96 | |
97 | return 0; |
98 | } |
99 | |
100 | static void latm_write_frame_header(AVFormatContext *s, PutBitContext *bs) |
101 | { |
102 | LATMContext *ctx = s->priv_data; |
103 | AVCodecParameters *par = s->streams[0]->codecpar; |
104 | int header_size; |
105 | |
106 | /* AudioMuxElement */ |
107 | put_bits(bs, 1, !!ctx->counter); |
108 | |
109 | if (!ctx->counter) { |
110 | /* StreamMuxConfig */ |
111 | put_bits(bs, 1, 0); /* audioMuxVersion */ |
112 | put_bits(bs, 1, 1); /* allStreamsSameTimeFraming */ |
113 | put_bits(bs, 6, 0); /* numSubFrames */ |
114 | put_bits(bs, 4, 0); /* numProgram */ |
115 | put_bits(bs, 3, 0); /* numLayer */ |
116 | |
117 | /* AudioSpecificConfig */ |
118 | if (ctx->object_type == AOT_ALS) { |
119 | header_size = par->extradata_size-(ctx->off >> 3); |
120 | avpriv_copy_bits(bs, &par->extradata[ctx->off >> 3], header_size); |
121 | } else { |
122 | // + 3 assumes not scalable and dependsOnCoreCoder == 0, |
123 | // see decode_ga_specific_config in libavcodec/aacdec.c |
124 | avpriv_copy_bits(bs, par->extradata, ctx->off + 3); |
125 | |
126 | if (!ctx->channel_conf) { |
127 | GetBitContext gb; |
128 | int ret = init_get_bits8(&gb, par->extradata, par->extradata_size); |
129 | av_assert0(ret >= 0); // extradata size has been checked already, so this should not fail |
130 | skip_bits_long(&gb, ctx->off + 3); |
131 | avpriv_copy_pce_data(bs, &gb); |
132 | } |
133 | } |
134 | |
135 | put_bits(bs, 3, 0); /* frameLengthType */ |
136 | put_bits(bs, 8, 0xff); /* latmBufferFullness */ |
137 | |
138 | put_bits(bs, 1, 0); /* otherDataPresent */ |
139 | put_bits(bs, 1, 0); /* crcCheckPresent */ |
140 | } |
141 | |
142 | ctx->counter++; |
143 | ctx->counter %= ctx->mod; |
144 | } |
145 | |
146 | static int latm_write_packet(AVFormatContext *s, AVPacket *pkt) |
147 | { |
148 | LATMContext *ctx = s->priv_data; |
149 | AVIOContext *pb = s->pb; |
150 | PutBitContext bs; |
151 | int i, len; |
152 | uint8_t loas_header[] = "\x56\xe0\x00"; |
153 | |
154 | if (s->streams[0]->codecpar->codec_id == AV_CODEC_ID_AAC_LATM) |
155 | return ff_raw_write_packet(s, pkt); |
156 | |
157 | if (!s->streams[0]->codecpar->extradata) { |
158 | if(pkt->size > 2 && pkt->data[0] == 0x56 && (pkt->data[1] >> 4) == 0xe && |
159 | (AV_RB16(pkt->data + 1) & 0x1FFF) + 3 == pkt->size) |
160 | return ff_raw_write_packet(s, pkt); |
161 | else |
162 | return AVERROR_INVALIDDATA; |
163 | } |
164 | |
165 | if (pkt->size > 0x1fff) |
166 | goto too_large; |
167 | |
168 | init_put_bits(&bs, ctx->buffer, pkt->size+1024+MAX_EXTRADATA_SIZE); |
169 | |
170 | latm_write_frame_header(s, &bs); |
171 | |
172 | /* PayloadLengthInfo() */ |
173 | for (i = 0; i <= pkt->size-255; i+=255) |
174 | put_bits(&bs, 8, 255); |
175 | |
176 | put_bits(&bs, 8, pkt->size-i); |
177 | |
178 | /* The LATM payload is written unaligned */ |
179 | |
180 | /* PayloadMux() */ |
181 | if (pkt->size && (pkt->data[0] & 0xe1) == 0x81) { |
182 | // Convert byte-aligned DSE to non-aligned. |
183 | // Due to the input format encoding we know that |
184 | // it is naturally byte-aligned in the input stream, |
185 | // so there are no padding bits to account for. |
186 | // To avoid having to add padding bits and rearrange |
187 | // the whole stream we just remove the byte-align flag. |
188 | // This allows us to remux our FATE AAC samples into latm |
189 | // files that are still playable with minimal effort. |
190 | put_bits(&bs, 8, pkt->data[0] & 0xfe); |
191 | avpriv_copy_bits(&bs, pkt->data + 1, 8*pkt->size - 8); |
192 | } else |
193 | avpriv_copy_bits(&bs, pkt->data, 8*pkt->size); |
194 | |
195 | avpriv_align_put_bits(&bs); |
196 | flush_put_bits(&bs); |
197 | |
198 | len = put_bits_count(&bs) >> 3; |
199 | |
200 | if (len > 0x1fff) |
201 | goto too_large; |
202 | |
203 | loas_header[1] |= (len >> 8) & 0x1f; |
204 | loas_header[2] |= len & 0xff; |
205 | |
206 | avio_write(pb, loas_header, 3); |
207 | avio_write(pb, ctx->buffer, len); |
208 | |
209 | return 0; |
210 | |
211 | too_large: |
212 | av_log(s, AV_LOG_ERROR, "LATM packet size larger than maximum size 0x1fff\n"); |
213 | return AVERROR_INVALIDDATA; |
214 | } |
215 | |
216 | static int latm_check_bitstream(struct AVFormatContext *s, const AVPacket *pkt) |
217 | { |
218 | int ret = 1; |
219 | AVStream *st = s->streams[pkt->stream_index]; |
220 | |
221 | if (st->codecpar->codec_id == AV_CODEC_ID_AAC) { |
222 | if (pkt->size > 2 && (AV_RB16(pkt->data) & 0xfff0) == 0xfff0) |
223 | ret = ff_stream_add_bitstream_filter(st, "aac_adtstoasc", NULL); |
224 | } |
225 | |
226 | return ret; |
227 | } |
228 | |
229 | AVOutputFormat ff_latm_muxer = { |
230 | .name = "latm", |
231 | .long_name = NULL_IF_CONFIG_SMALL("LOAS/LATM"), |
232 | .mime_type = "audio/MP4A-LATM", |
233 | .extensions = "latm,loas", |
234 | .priv_data_size = sizeof(LATMContext), |
235 | .audio_codec = AV_CODEC_ID_AAC, |
236 | .video_codec = AV_CODEC_ID_NONE, |
237 | .write_header = latm_write_header, |
238 | .write_packet = latm_write_packet, |
239 | .priv_class = &latm_muxer_class, |
240 | .check_bitstream= latm_check_bitstream, |
241 | .flags = AVFMT_NOTIMESTAMPS, |
242 | }; |
243 |