You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.

242 lines
7.9 KiB

/*
* AAC decoder
* Copyright (c) 2005-2006 Oded Shimon ( ods15 ods15 dyndns org )
* Copyright (c) 2006-2007 Maxim Gavrilov ( maxim.gavrilov gmail com )
*
* This file is part of FFmpeg.
*
* FFmpeg is free software; you can redistribute it and/or
* modify it under the terms of the GNU Lesser General Public
* License as published by the Free Software Foundation; either
* version 2.1 of the License, or (at your option) any later version.
*
* FFmpeg is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
* Lesser General Public License for more details.
*
* You should have received a copy of the GNU Lesser General Public
* License along with FFmpeg; if not, write to the Free Software
* Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
*/
/**
* @file aac.c
* AAC decoder
* @author Oded Shimon ( ods15 ods15 dyndns org )
* @author Maxim Gavrilov ( maxim.gavrilov gmail com )
*/
/*
* supported tools
*
* Support? Name
* N (code in SoC repo) gain control
* Y block switching
* Y window shapes - standard
* N window shapes - Low Delay
* Y filterbank - standard
* N (code in SoC repo) filterbank - Scalable Sample Rate
* Y Temporal Noise Shaping
* N (code in SoC repo) Long Term Prediction
* Y intensity stereo
* Y channel coupling
* N frequency domain prediction
* Y Perceptual Noise Substitution
* Y Mid/Side stereo
* N Scalable Inverse AAC Quantization
* N Frequency Selective Switch
* N upsampling filter
* Y quantization & coding - AAC
* N quantization & coding - TwinVQ
* N quantization & coding - BSAC
* N AAC Error Resilience tools
* N Error Resilience payload syntax
* N Error Protection tool
* N CELP
* N Silence Compression
* N HVXC
* N HVXC 4kbits/s VR
* N Structured Audio tools
* N Structured Audio Sample Bank Format
* N MIDI
* N Harmonic and Individual Lines plus Noise
* N Text-To-Speech Interface
* N (in progress) Spectral Band Replication
* Y (not in this code) Layer-1
* Y (not in this code) Layer-2
* Y (not in this code) Layer-3
* N SinuSoidal Coding (Transient, Sinusoid, Noise)
* N (planned) Parametric Stereo
* N Direct Stream Transfer
*
* Note: - HE AAC v1 comprises LC AAC with Spectral Band Replication.
* - HE AAC v2 comprises LC AAC with Spectral Band Replication and
Parametric Stereo.
*/
#include "avcodec.h"
#include "bitstream.h"
#include "dsputil.h"
#include "aac.h"
#include "aactab.h"
#include "mpeg4audio.h"
#include <assert.h>
#include <errno.h>
#include <math.h>
#include <string.h>
#ifndef CONFIG_HARDCODED_TABLES
static float ff_aac_ivquant_tab[IVQUANT_SIZE];
#endif /* CONFIG_HARDCODED_TABLES */
static VLC vlc_scalefactors;
static VLC vlc_spectral[11];
num_front = get_bits(gb, 4);
num_side = get_bits(gb, 4);
num_back = get_bits(gb, 4);
num_lfe = get_bits(gb, 2);
num_assoc_data = get_bits(gb, 3);
num_cc = get_bits(gb, 4);
newpcs->mono_mixdown_tag = get_bits1(gb) ? get_bits(gb, 4) : -1;
newpcs->stereo_mixdown_tag = get_bits1(gb) ? get_bits(gb, 4) : -1;
if (get_bits1(gb)) {
newpcs->mixdown_coeff_index = get_bits(gb, 2);
newpcs->pseudo_surround = get_bits1(gb);
}
program_config_element_parse_tags(newpcs->che_type[ID_CPE], newpcs->che_type[ID_SCE], AAC_CHANNEL_FRONT, gb, num_front);
program_config_element_parse_tags(newpcs->che_type[ID_CPE], newpcs->che_type[ID_SCE], AAC_CHANNEL_SIDE, gb, num_side );
program_config_element_parse_tags(newpcs->che_type[ID_CPE], newpcs->che_type[ID_SCE], AAC_CHANNEL_BACK, gb, num_back );
program_config_element_parse_tags(NULL, newpcs->che_type[ID_LFE], AAC_CHANNEL_LFE, gb, num_lfe );
skip_bits_long(gb, 4 * num_assoc_data);
program_config_element_parse_tags(newpcs->che_type[ID_CCE], newpcs->che_type[ID_CCE], AAC_CHANNEL_CC, gb, num_cc );
align_get_bits(gb);
/* comment field, first byte is length */
skip_bits_long(gb, 8 * get_bits(gb, 8));
static av_cold int aac_decode_init(AVCodecContext * avccontext) {
AACContext * ac = avccontext->priv_data;
int i;
ac->avccontext = avccontext;
avccontext->sample_rate = ac->m4ac.sample_rate;
avccontext->frame_size = 1024;
AAC_INIT_VLC_STATIC( 0, 144);
AAC_INIT_VLC_STATIC( 1, 114);
AAC_INIT_VLC_STATIC( 2, 188);
AAC_INIT_VLC_STATIC( 3, 180);
AAC_INIT_VLC_STATIC( 4, 172);
AAC_INIT_VLC_STATIC( 5, 140);
AAC_INIT_VLC_STATIC( 6, 168);
AAC_INIT_VLC_STATIC( 7, 114);
AAC_INIT_VLC_STATIC( 8, 262);
AAC_INIT_VLC_STATIC( 9, 248);
AAC_INIT_VLC_STATIC(10, 384);
dsputil_init(&ac->dsp, avccontext);
// -1024 - Compensate wrong IMDCT method.
// 32768 - Required to scale values to the correct range for the bias method
// for float to int16 conversion.
if(ac->dsp.float_to_int16 == ff_float_to_int16_c) {
ac->add_bias = 385.0f;
ac->sf_scale = 1. / (-1024. * 32768.);
ac->sf_offset = 0;
} else {
ac->add_bias = 0.0f;
ac->sf_scale = 1. / -1024.;
ac->sf_offset = 60;
}
#ifndef CONFIG_HARDCODED_TABLES
for (i = 1 - IVQUANT_SIZE/2; i < IVQUANT_SIZE/2; i++)
ff_aac_ivquant_tab[i + IVQUANT_SIZE/2 - 1] = cbrt(fabs(i)) * i;
#endif /* CONFIG_HARDCODED_TABLES */
INIT_VLC_STATIC(&vlc_scalefactors, 7, sizeof(ff_aac_scalefactor_code)/sizeof(ff_aac_scalefactor_code[0]),
ff_aac_scalefactor_bits, sizeof(ff_aac_scalefactor_bits[0]), sizeof(ff_aac_scalefactor_bits[0]),
ff_aac_scalefactor_code, sizeof(ff_aac_scalefactor_code[0]), sizeof(ff_aac_scalefactor_code[0]),
352);
ff_mdct_init(&ac->mdct, 11, 1);
ff_mdct_init(&ac->mdct_small, 8, 1);
return 0;
}
int byte_align = get_bits1(gb);
int count = get_bits(gb, 8);
if (count == 255)
count += get_bits(gb, 8);
if (byte_align)
align_get_bits(gb);
skip_bits_long(gb, 8 * count);
}
/**
* inverse quantization
*
* @param a quantized value to be dequantized
* @return Returns dequantized value.
*/
static inline float ivquant(int a) {
if (a + (unsigned int)IVQUANT_SIZE/2 - 1 < (unsigned int)IVQUANT_SIZE - 1)
return ff_aac_ivquant_tab[a + IVQUANT_SIZE/2 - 1];
else
return cbrtf(fabsf(a)) * a;
}
* @param pulse pointer to pulse data struct
* @param icoef array of quantized spectral data
*/
static void add_pulses(int icoef[1024], const Pulse * pulse, const IndividualChannelStream * ics) {
int i, off = ics->swb_offset[pulse->start];
for (i = 0; i < pulse->num_pulse; i++) {
int ic;
off += pulse->offset[i];
ic = (icoef[off] - 1)>>31;
icoef[off] += (pulse->amp[i]^ic) - ic;
}
}
static av_cold int aac_decode_close(AVCodecContext * avccontext) {
AACContext * ac = avccontext->priv_data;
int i, j;
for (i = 0; i < MAX_TAGID; i++) {
for(j = 0; j < 4; j++)
av_freep(&ac->che[j][i]);
}
ff_mdct_end(&ac->mdct);
ff_mdct_end(&ac->mdct_small);
av_freep(&ac->interleaved_output);
return 0 ;
}
AVCodec aac_decoder = {
"aac",
CODEC_TYPE_AUDIO,
CODEC_ID_AAC,
sizeof(AACContext),
aac_decode_init,
NULL,
aac_decode_close,
aac_decode_frame,
.long_name = NULL_IF_CONFIG_SMALL("Advanced Audio Coding"),
};