FFmpeg
All Data Structures Namespaces Files Functions Variables Typedefs Enumerations Enumerator Macros Modules Pages
fastaudio.c
Go to the documentation of this file.
1 /*
2  * MOFLEX Fast Audio decoder
3  * Copyright (c) 2015-2016 Florian Nouwt
4  * Copyright (c) 2017 Adib Surani
5  * Copyright (c) 2020 Paul B Mahol
6  *
7  * This file is part of FFmpeg.
8  *
9  * FFmpeg is free software; you can redistribute it and/or
10  * modify it under the terms of the GNU Lesser General Public
11  * License as published by the Free Software Foundation; either
12  * version 2.1 of the License, or (at your option) any later version.
13  *
14  * FFmpeg is distributed in the hope that it will be useful,
15  * but WITHOUT ANY WARRANTY; without even the implied warranty of
16  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
17  * Lesser General Public License for more details.
18  *
19  * You should have received a copy of the GNU Lesser General Public
20  * License along with FFmpeg; if not, write to the Free Software
21  * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
22  */
23 
24 #include "libavutil/intfloat.h"
25 #include "libavutil/mem.h"
26 #include "avcodec.h"
27 #include "bytestream.h"
28 #include "codec_internal.h"
29 #include "decode.h"
30 
31 typedef struct ChannelItems {
32  float f[8];
33  float last;
34 } ChannelItems;
35 
36 typedef struct FastAudioContext {
37  float table[8][64];
38 
41 
43 {
44  FastAudioContext *s = avctx->priv_data;
45 
47 
48  for (int i = 0; i < 8; i++)
49  s->table[0][i] = (i - 159.5f) / 160.f;
50  for (int i = 0; i < 11; i++)
51  s->table[0][i + 8] = (i - 37.5f) / 40.f;
52  for (int i = 0; i < 27; i++)
53  s->table[0][i + 8 + 11] = (i - 13.f) / 20.f;
54  for (int i = 0; i < 11; i++)
55  s->table[0][i + 8 + 11 + 27] = (i + 27.5f) / 40.f;
56  for (int i = 0; i < 7; i++)
57  s->table[0][i + 8 + 11 + 27 + 11] = (i + 152.5f) / 160.f;
58 
59  memcpy(s->table[1], s->table[0], sizeof(s->table[0]));
60 
61  for (int i = 0; i < 7; i++)
62  s->table[2][i] = (i - 33.5f) / 40.f;
63  for (int i = 0; i < 25; i++)
64  s->table[2][i + 7] = (i - 13.f) / 20.f;
65 
66  for (int i = 0; i < 32; i++)
67  s->table[3][i] = -s->table[2][31 - i];
68 
69  for (int i = 0; i < 16; i++)
70  s->table[4][i] = i * 0.22f / 3.f - 0.6f;
71 
72  for (int i = 0; i < 16; i++)
73  s->table[5][i] = i * 0.20f / 3.f - 0.3f;
74 
75  for (int i = 0; i < 8; i++)
76  s->table[6][i] = i * 0.36f / 3.f - 0.4f;
77 
78  for (int i = 0; i < 8; i++)
79  s->table[7][i] = i * 0.34f / 3.f - 0.2f;
80 
81  s->ch = av_calloc(avctx->ch_layout.nb_channels, sizeof(*s->ch));
82  if (!s->ch)
83  return AVERROR(ENOMEM);
84 
85  return 0;
86 }
87 
88 static int read_bits(int bits, int *ppos, unsigned *src)
89 {
90  int r, pos;
91 
92  pos = *ppos;
93  pos += bits;
94  r = src[(pos - 1) / 32] >> ((-pos) & 31);
95  *ppos = pos;
96 
97  return r & ((1 << bits) - 1);
98 }
99 
100 static const uint8_t bits[8] = { 6, 6, 5, 5, 4, 0, 3, 3, };
101 
102 static void set_sample(int i, int j, int v, float *result, int *pads, float value)
103 {
104  result[i * 64 + pads[i] + j * 3] = value * (2 * v - 7);
105 }
106 
108  int *got_frame, AVPacket *pkt)
109 {
110  FastAudioContext *s = avctx->priv_data;
111  GetByteContext gb;
112  int subframes;
113  int ret;
114 
115  subframes = pkt->size / (40 * avctx->ch_layout.nb_channels);
116  frame->nb_samples = subframes * 256;
117  if ((ret = ff_get_buffer(avctx, frame, 0)) < 0)
118  return ret;
119 
120  bytestream2_init(&gb, pkt->data, pkt->size);
121 
122  for (int subframe = 0; subframe < subframes; subframe++) {
123  for (int channel = 0; channel < avctx->ch_layout.nb_channels; channel++) {
124  ChannelItems *ch = &s->ch[channel];
125  float result[256] = { 0 };
126  unsigned src[10];
127  int inds[4], pads[4];
128  float m[8];
129  int pos = 0;
130 
131  for (int i = 0; i < 10; i++)
132  src[i] = bytestream2_get_le32(&gb);
133 
134  for (int i = 0; i < 8; i++)
135  m[7 - i] = s->table[i][read_bits(bits[i], &pos, src)];
136 
137  for (int i = 0; i < 4; i++)
138  inds[3 - i] = read_bits(6, &pos, src);
139 
140  for (int i = 0; i < 4; i++)
141  pads[3 - i] = read_bits(2, &pos, src);
142 
143  for (int i = 0, index5 = 0; i < 4; i++) {
144  float value = av_int2float((inds[i] + 1) << 20) * powf(2.f, 116.f);
145 
146  for (int j = 0, tmp = 0; j < 21; j++) {
147  set_sample(i, j, j == 20 ? tmp / 2 : read_bits(3, &pos, src), result, pads, value);
148  if (j % 10 == 9)
149  tmp = 4 * tmp + read_bits(2, &pos, src);
150  if (j == 20)
151  index5 = FFMIN(2 * index5 + tmp % 2, 63);
152  }
153 
154  m[2] = s->table[5][index5];
155  }
156 
157  for (int i = 0; i < 256; i++) {
158  float x = result[i];
159 
160  for (int j = 0; j < 8; j++) {
161  x -= m[j] * ch->f[j];
162  ch->f[j] += m[j] * x;
163  }
164 
165  memmove(&ch->f[0], &ch->f[1], sizeof(float) * 7);
166  ch->f[7] = x;
167  ch->last = x + ch->last * 0.86f;
168  result[i] = ch->last * 2.f;
169  }
170 
171  memcpy(frame->extended_data[channel] + 1024 * subframe, result, 256 * sizeof(float));
172  }
173  }
174 
175  *got_frame = 1;
176 
177  return pkt->size;
178 }
179 
181 {
182  FastAudioContext *s = avctx->priv_data;
183 
184  av_freep(&s->ch);
185 
186  return 0;
187 }
188 
190  .p.name = "fastaudio",
191  CODEC_LONG_NAME("MobiClip FastAudio"),
192  .p.type = AVMEDIA_TYPE_AUDIO,
193  .p.id = AV_CODEC_ID_FASTAUDIO,
194  .priv_data_size = sizeof(FastAudioContext),
195  .init = fastaudio_init,
197  .close = fastaudio_close,
198  .p.capabilities = AV_CODEC_CAP_DR1,
200 };
AV_SAMPLE_FMT_FLTP
@ AV_SAMPLE_FMT_FLTP
float, planar
Definition: samplefmt.h:66
r
const char * r
Definition: vf_curves.c:127
AVERROR
Filter the word “frame” indicates either a video frame or a group of audio as stored in an AVFrame structure Format for each input and each output the list of supported formats For video that means pixel format For audio that means channel sample they are references to shared objects When the negotiation mechanism computes the intersection of the formats supported at each end of a all references to both lists are replaced with a reference to the intersection And when a single format is eventually chosen for a link amongst the remaining all references to the list are updated That means that if a filter requires that its input and output have the same format amongst a supported all it has to do is use a reference to the same list of formats query_formats can leave some formats unset and return AVERROR(EAGAIN) to cause the negotiation mechanism toagain later. That can be used by filters with complex requirements to use the format negotiated on one link to set the formats supported on another. Frame references ownership and permissions
GetByteContext
Definition: bytestream.h:33
fastaudio_close
static av_cold int fastaudio_close(AVCodecContext *avctx)
Definition: fastaudio.c:180
AV_CODEC_ID_FASTAUDIO
@ AV_CODEC_ID_FASTAUDIO
Definition: codec_id.h:542
AVFrame
This structure describes decoded (raw) audio or video data.
Definition: frame.h:410
tmp
static uint8_t tmp[11]
Definition: aes_ctr.c:28
AVPacket::data
uint8_t * data
Definition: packet.h:535
FastAudioContext::ch
ChannelItems * ch
Definition: fastaudio.c:39
FFCodec
Definition: codec_internal.h:127
FastAudioContext
Definition: fastaudio.c:36
AVChannelLayout::nb_channels
int nb_channels
Number of channels in this layout.
Definition: channel_layout.h:329
intfloat.h
FFCodec::p
AVCodec p
The public AVCodec.
Definition: codec_internal.h:131
AVCodecContext::ch_layout
AVChannelLayout ch_layout
Audio channel layout.
Definition: avcodec.h:1042
av_int2float
static av_always_inline float av_int2float(uint32_t i)
Reinterpret a 32-bit integer as a float.
Definition: intfloat.h:40
fastaudio_init
static av_cold int fastaudio_init(AVCodecContext *avctx)
Definition: fastaudio.c:42
pkt
AVPacket * pkt
Definition: movenc.c:60
av_cold
#define av_cold
Definition: attributes.h:90
set_sample
static void set_sample(int i, int j, int v, float *result, int *pads, float value)
Definition: fastaudio.c:102
FF_CODEC_DECODE_CB
#define FF_CODEC_DECODE_CB(func)
Definition: codec_internal.h:341
s
#define s(width, name)
Definition: cbs_vp9.c:198
AVMEDIA_TYPE_AUDIO
@ AVMEDIA_TYPE_AUDIO
Definition: avutil.h:202
decode.h
CODEC_LONG_NAME
#define CODEC_LONG_NAME(str)
Definition: codec_internal.h:326
ff_fastaudio_decoder
const FFCodec ff_fastaudio_decoder
Definition: fastaudio.c:189
ChannelItems::last
float last
Definition: fastaudio.c:33
result
and forward the result(frame or status change) to the corresponding input. If nothing is possible
FastAudioContext::table
float table[8][64]
Definition: fastaudio.c:37
f
f
Definition: af_crystalizer.c:122
ff_get_buffer
int ff_get_buffer(AVCodecContext *avctx, AVFrame *frame, int flags)
Get a buffer for a frame.
Definition: decode.c:1608
init
int(* init)(AVBSFContext *ctx)
Definition: dts2pts.c:368
AV_CODEC_CAP_DR1
#define AV_CODEC_CAP_DR1
Codec uses get_buffer() or get_encode_buffer() for allocating buffers and supports custom allocators.
Definition: codec.h:52
AVPacket::size
int size
Definition: packet.h:536
ChannelItems::f
float f[8]
Definition: fastaudio.c:32
powf
#define powf(x, y)
Definition: libm.h:52
codec_internal.h
for
for(k=2;k<=8;++k)
Definition: h264pred_template.c:424
AVCodecContext::sample_fmt
enum AVSampleFormat sample_fmt
audio sample format
Definition: avcodec.h:1034
read_bits
static int read_bits(int bits, int *ppos, unsigned *src)
Definition: fastaudio.c:88
CODEC_SAMPLEFMTS
#define CODEC_SAMPLEFMTS(...)
Definition: codec_internal.h:380
i
#define i(width, name, range_min, range_max)
Definition: cbs_h2645.c:256
value
it s the only field you need to keep assuming you have a context There is some magic you don t need to care about around this just let it vf default value
Definition: writing_filters.txt:86
FFMIN
#define FFMIN(a, b)
Definition: macros.h:49
AVCodec::name
const char * name
Name of the codec implementation.
Definition: codec.h:179
av_calloc
void * av_calloc(size_t nmemb, size_t size)
Definition: mem.c:264
avcodec.h
ret
ret
Definition: filter_design.txt:187
frame
these buffered frames must be flushed immediately if a new input produces new the filter must not call request_frame to get more It must just process the frame or queue it The task of requesting more frames is left to the filter s request_frame method or the application If a filter has several the filter must be ready for frames arriving randomly on any input any filter with several inputs will most likely require some kind of queuing mechanism It is perfectly acceptable to have a limited queue and to drop frames when the inputs are too unbalanced request_frame For filters that do not use the this method is called when a frame is wanted on an output For a it should directly call filter_frame on the corresponding output For a if there are queued frames already one of these frames should be pushed If the filter should request a frame on one of its repeatedly until at least one frame has been pushed Return or at least make progress towards producing a frame
Definition: filter_design.txt:264
pos
unsigned int pos
Definition: spdifenc.c:414
ChannelItems
Definition: fastaudio.c:31
AVCodecContext
main external API structure.
Definition: avcodec.h:431
mem.h
AVPacket
This structure stores compressed data.
Definition: packet.h:512
AVCodecContext::priv_data
void * priv_data
Definition: avcodec.h:458
av_freep
#define av_freep(p)
Definition: tableprint_vlc.h:35
bytestream.h
bytestream2_init
static av_always_inline void bytestream2_init(GetByteContext *g, const uint8_t *buf, int buf_size)
Definition: bytestream.h:137
fastaudio_decode
static int fastaudio_decode(AVCodecContext *avctx, AVFrame *frame, int *got_frame, AVPacket *pkt)
Definition: fastaudio.c:107
bits
static const uint8_t bits[8]
Definition: fastaudio.c:100
src
#define src
Definition: vp8dsp.c:248
channel
channel
Definition: ebur128.h:39