/src/ffmpeg/libavcodec/vmdaudio.c
Line | Count | Source |
1 | | /* |
2 | | * Sierra VMD audio decoder |
3 | | * Copyright (c) 2004 The FFmpeg Project |
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 | | * @file |
24 | | * Sierra VMD audio decoder |
25 | | * by Vladimir "VAG" Gneushev (vagsoft at mail.ru) |
26 | | * for more information on the Sierra VMD format, visit: |
27 | | * http://www.pcisys.net/~melanson/codecs/ |
28 | | * |
29 | | * The audio decoder, expects each encoded data |
30 | | * chunk to be prepended with the appropriate 16-byte frame information |
31 | | * record from the VMD file. It does not require the 0x330-byte VMD file |
32 | | * header, but it does need the audio setup parameters passed in through |
33 | | * normal libavcodec API means. |
34 | | */ |
35 | | |
36 | | #include <string.h> |
37 | | |
38 | | #include "libavutil/avassert.h" |
39 | | #include "libavutil/channel_layout.h" |
40 | | #include "libavutil/common.h" |
41 | | #include "libavutil/intreadwrite.h" |
42 | | |
43 | | #include "avcodec.h" |
44 | | #include "codec_internal.h" |
45 | | #include "decode.h" |
46 | | |
47 | 47.6k | #define BLOCK_TYPE_AUDIO 1 |
48 | 21.1k | #define BLOCK_TYPE_INITIAL 2 |
49 | 40.8k | #define BLOCK_TYPE_SILENCE 3 |
50 | | |
51 | | typedef struct VmdAudioContext { |
52 | | int out_bps; |
53 | | int chunk_size; |
54 | | } VmdAudioContext; |
55 | | |
56 | | static const uint16_t vmdaudio_table[128] = { |
57 | | 0x000, 0x008, 0x010, 0x020, 0x030, 0x040, 0x050, 0x060, 0x070, 0x080, |
58 | | 0x090, 0x0A0, 0x0B0, 0x0C0, 0x0D0, 0x0E0, 0x0F0, 0x100, 0x110, 0x120, |
59 | | 0x130, 0x140, 0x150, 0x160, 0x170, 0x180, 0x190, 0x1A0, 0x1B0, 0x1C0, |
60 | | 0x1D0, 0x1E0, 0x1F0, 0x200, 0x208, 0x210, 0x218, 0x220, 0x228, 0x230, |
61 | | 0x238, 0x240, 0x248, 0x250, 0x258, 0x260, 0x268, 0x270, 0x278, 0x280, |
62 | | 0x288, 0x290, 0x298, 0x2A0, 0x2A8, 0x2B0, 0x2B8, 0x2C0, 0x2C8, 0x2D0, |
63 | | 0x2D8, 0x2E0, 0x2E8, 0x2F0, 0x2F8, 0x300, 0x308, 0x310, 0x318, 0x320, |
64 | | 0x328, 0x330, 0x338, 0x340, 0x348, 0x350, 0x358, 0x360, 0x368, 0x370, |
65 | | 0x378, 0x380, 0x388, 0x390, 0x398, 0x3A0, 0x3A8, 0x3B0, 0x3B8, 0x3C0, |
66 | | 0x3C8, 0x3D0, 0x3D8, 0x3E0, 0x3E8, 0x3F0, 0x3F8, 0x400, 0x440, 0x480, |
67 | | 0x4C0, 0x500, 0x540, 0x580, 0x5C0, 0x600, 0x640, 0x680, 0x6C0, 0x700, |
68 | | 0x740, 0x780, 0x7C0, 0x800, 0x900, 0xA00, 0xB00, 0xC00, 0xD00, 0xE00, |
69 | | 0xF00, 0x1000, 0x1400, 0x1800, 0x1C00, 0x2000, 0x3000, 0x4000 |
70 | | }; |
71 | | |
72 | | static av_cold int vmdaudio_decode_init(AVCodecContext *avctx) |
73 | 542 | { |
74 | 542 | VmdAudioContext *s = avctx->priv_data; |
75 | 542 | int channels = avctx->ch_layout.nb_channels; |
76 | | |
77 | 542 | if (channels < 1 || channels > 2) { |
78 | 35 | av_log(avctx, AV_LOG_ERROR, "invalid number of channels\n"); |
79 | 35 | return AVERROR(EINVAL); |
80 | 35 | } |
81 | 507 | if (avctx->block_align < 1 || avctx->block_align % channels || |
82 | 504 | avctx->block_align > INT_MAX - channels) { |
83 | 4 | av_log(avctx, AV_LOG_ERROR, "invalid block align\n"); |
84 | 4 | return AVERROR(EINVAL); |
85 | 4 | } |
86 | | |
87 | 503 | av_channel_layout_uninit(&avctx->ch_layout); |
88 | 503 | av_channel_layout_default(&avctx->ch_layout, channels); |
89 | | |
90 | 503 | if (avctx->bits_per_coded_sample == 16) |
91 | 44 | avctx->sample_fmt = AV_SAMPLE_FMT_S16; |
92 | 459 | else |
93 | 459 | avctx->sample_fmt = AV_SAMPLE_FMT_U8; |
94 | 503 | s->out_bps = av_get_bytes_per_sample(avctx->sample_fmt); |
95 | | |
96 | 503 | s->chunk_size = avctx->block_align + channels * (s->out_bps == 2); |
97 | | |
98 | 503 | av_log(avctx, AV_LOG_DEBUG, "%d channels, %d bits/sample, " |
99 | 503 | "block align = %d, sample rate = %d\n", |
100 | 503 | channels, avctx->bits_per_coded_sample, avctx->block_align, |
101 | 503 | avctx->sample_rate); |
102 | | |
103 | 503 | return 0; |
104 | 507 | } |
105 | | |
106 | | static void decode_audio_s16(int16_t *out, const uint8_t *buf, int buf_size, |
107 | | int channels) |
108 | 1.13k | { |
109 | 1.13k | int ch; |
110 | 1.13k | const uint8_t *buf_end = buf + buf_size; |
111 | 1.13k | int predictor[2]; |
112 | 1.13k | int st = channels - 1; |
113 | | |
114 | | /* decode initial raw sample */ |
115 | 3.32k | for (ch = 0; ch < channels; ch++) { |
116 | 2.18k | predictor[ch] = (int16_t)AV_RL16(buf); |
117 | 2.18k | buf += 2; |
118 | 2.18k | *out++ = predictor[ch]; |
119 | 2.18k | } |
120 | | |
121 | | /* decode DPCM samples */ |
122 | 1.13k | ch = 0; |
123 | 6.71k | while (buf < buf_end) { |
124 | 5.57k | uint8_t b = *buf++; |
125 | 5.57k | if (b & 0x80) |
126 | 1.88k | predictor[ch] -= vmdaudio_table[b & 0x7F]; |
127 | 3.69k | else |
128 | 3.69k | predictor[ch] += vmdaudio_table[b]; |
129 | 5.57k | predictor[ch] = av_clip_int16(predictor[ch]); |
130 | 5.57k | *out++ = predictor[ch]; |
131 | 5.57k | ch ^= st; |
132 | 5.57k | } |
133 | 1.13k | } |
134 | | |
135 | | static int vmdaudio_decode_frame(AVCodecContext *avctx, AVFrame *frame, |
136 | | int *got_frame_ptr, AVPacket *avpkt) |
137 | 257k | { |
138 | 257k | const uint8_t *buf = avpkt->data; |
139 | 257k | const uint8_t *buf_end; |
140 | 257k | int buf_size = avpkt->size; |
141 | 257k | VmdAudioContext *s = avctx->priv_data; |
142 | 257k | int block_type, silent_chunks, audio_chunks; |
143 | 257k | int ret; |
144 | 257k | uint8_t *output_samples_u8; |
145 | 257k | int16_t *output_samples_s16; |
146 | 257k | int channels = avctx->ch_layout.nb_channels; |
147 | | |
148 | 257k | if (buf_size < 16) { |
149 | 233k | av_log(avctx, AV_LOG_WARNING, "skipping small junk packet\n"); |
150 | 233k | *got_frame_ptr = 0; |
151 | 233k | return buf_size; |
152 | 233k | } |
153 | | |
154 | 23.8k | block_type = buf[6]; |
155 | 23.8k | if (block_type < BLOCK_TYPE_AUDIO || block_type > BLOCK_TYPE_SILENCE) { |
156 | 2.68k | av_log(avctx, AV_LOG_ERROR, "unknown block type: %d\n", block_type); |
157 | 2.68k | return AVERROR(EINVAL); |
158 | 2.68k | } |
159 | 21.1k | buf += 16; |
160 | 21.1k | buf_size -= 16; |
161 | | |
162 | | /* get number of silent chunks */ |
163 | 21.1k | silent_chunks = 0; |
164 | 21.1k | if (block_type == BLOCK_TYPE_INITIAL) { |
165 | 3.20k | uint32_t flags; |
166 | 3.20k | if (buf_size < 4) { |
167 | 1.84k | av_log(avctx, AV_LOG_ERROR, "packet is too small\n"); |
168 | 1.84k | return AVERROR(EINVAL); |
169 | 1.84k | } |
170 | 1.35k | flags = AV_RB32(buf); |
171 | 1.35k | silent_chunks = av_popcount(flags); |
172 | 1.35k | buf += 4; |
173 | 1.35k | buf_size -= 4; |
174 | 17.9k | } else if (block_type == BLOCK_TYPE_SILENCE) { |
175 | 11.0k | silent_chunks = 1; |
176 | 11.0k | buf_size = 0; // should already be zero but set it just to be sure |
177 | 11.0k | } |
178 | | |
179 | | /* ensure output buffer is large enough */ |
180 | 19.3k | audio_chunks = buf_size / s->chunk_size; |
181 | | |
182 | | /* drop incomplete chunks */ |
183 | 19.3k | buf_size = audio_chunks * s->chunk_size; |
184 | | |
185 | 19.3k | if (silent_chunks + audio_chunks >= INT_MAX / avctx->block_align) |
186 | 813 | return AVERROR_INVALIDDATA; |
187 | | |
188 | | /* get output buffer */ |
189 | 18.4k | frame->nb_samples = ((silent_chunks + audio_chunks) * avctx->block_align) / |
190 | 18.4k | avctx->ch_layout.nb_channels; |
191 | 18.4k | if ((ret = ff_get_buffer(avctx, frame, 0)) < 0) |
192 | 13.2k | return ret; |
193 | 5.27k | output_samples_u8 = frame->data[0]; |
194 | 5.27k | output_samples_s16 = (int16_t *)frame->data[0]; |
195 | | |
196 | | /* decode silent chunks */ |
197 | 5.27k | if (silent_chunks > 0) { |
198 | 3.70k | int silent_size = avctx->block_align * silent_chunks; |
199 | 3.70k | av_assert0(avctx->block_align * silent_chunks <= frame->nb_samples * avctx->ch_layout.nb_channels); |
200 | | |
201 | 3.70k | if (s->out_bps == 2) { |
202 | 336 | memset(output_samples_s16, 0x00, silent_size * 2); |
203 | 336 | output_samples_s16 += silent_size; |
204 | 3.37k | } else { |
205 | 3.37k | memset(output_samples_u8, 0x80, silent_size); |
206 | 3.37k | output_samples_u8 += silent_size; |
207 | 3.37k | } |
208 | 3.70k | } |
209 | | |
210 | | /* decode audio chunks */ |
211 | 5.27k | if (audio_chunks > 0) { |
212 | 1.75k | buf_end = buf + buf_size; |
213 | 1.75k | av_assert0((buf_size & (avctx->ch_layout.nb_channels > 1)) == 0); |
214 | 10.7M | while (buf_end - buf >= s->chunk_size) { |
215 | 10.7M | if (s->out_bps == 2) { |
216 | 1.13k | decode_audio_s16(output_samples_s16, buf, s->chunk_size, channels); |
217 | 1.13k | output_samples_s16 += avctx->block_align; |
218 | 10.7M | } else { |
219 | 10.7M | memcpy(output_samples_u8, buf, s->chunk_size); |
220 | 10.7M | output_samples_u8 += avctx->block_align; |
221 | 10.7M | } |
222 | 10.7M | buf += s->chunk_size; |
223 | 10.7M | } |
224 | 1.75k | } |
225 | | |
226 | 5.27k | *got_frame_ptr = 1; |
227 | | |
228 | 5.27k | return avpkt->size; |
229 | 5.27k | } |
230 | | |
231 | | const FFCodec ff_vmdaudio_decoder = { |
232 | | .p.name = "vmdaudio", |
233 | | CODEC_LONG_NAME("Sierra VMD audio"), |
234 | | .p.type = AVMEDIA_TYPE_AUDIO, |
235 | | .p.id = AV_CODEC_ID_VMDAUDIO, |
236 | | .priv_data_size = sizeof(VmdAudioContext), |
237 | | .init = vmdaudio_decode_init, |
238 | | FF_CODEC_DECODE_CB(vmdaudio_decode_frame), |
239 | | .p.capabilities = AV_CODEC_CAP_DR1, |
240 | | }; |