47 #define MAX_CHANNELS 64
48 #define MAX_ELEM_ID 16
50 #define TNS_MAX_ORDER 20
51 #define MAX_LTP_LONG_SFB 40
53 #define CLIP_AVOIDANCE_FACTOR 0.95f
92 #define IS_CODEBOOK_UNSIGNED(x) (((x) - 1) & 10)
146 #define MAX_PREDICTORS 672
148 #define SCALE_DIV_512 36
149 #define SCALE_ONE_POS 140
150 #define SCALE_MAX_POS 255
151 #define SCALE_MAX_DIFF 60
152 #define SCALE_DIFF_ZERO 60
154 #define NOISE_PRE 256
155 #define NOISE_PRE_BITS 9
156 #define NOISE_OFFSET 90
int predictor_initialized
Band types following are encoded differently from others.
float pns_ener[128]
Noise energy values (used by encoder)
uint8_t use_kb_window[2]
If set, use Kaiser-Bessel window, otherwise use a sine window.
void(* subband_scale)(int *dst, int *src, int scale, int offset, int len)
This structure describes decoded (raw) audio or video data.
int coef_idx[8][4][TNS_MAX_ORDER]
float pcoeffs[1024]
coefficients for IMDCT, pristine
uint8_t window_clipping[8]
set if a certain window is near clipping
ChannelElement * che[4][MAX_ELEM_ID]
INTFLOAT * ret
PCM output.
#define DECLARE_ALIGNED(n, t, v)
int common_window
Set if channels share a common 'IndividualChannelStream' in bitstream.
void(* update_ltp)(AACContext *ac, SingleChannelElement *sce)
void(* imdct_and_windowing)(AACContext *ac, SingleChannelElement *sce)
INTFLOAT sf[120]
scalefactors
uint8_t ms_mask[128]
Set if mid/side stereo is used for each scalefactor window band.
Dynamic Range Control - decoded from the bitstream but not processed further.
Spectral data are scaled white noise not coded in the bitstream.
int band_incr
Number of DRC bands greater than 1 having DRC info.
int dmono_mode
0->not dmono, 1->use first channel, 2->use second channel
const uint16_t * swb_offset
table of offsets to the lowest spectral coefficient of a scalefactor band, sfb, for a particular wind...
uint8_t layout_map[MAX_ELEM_ID *4][3]
Output configuration under trial specified by an inband PCE.
SingleChannelElement ch[2]
CouplingPoint
The point during decoding at which channel coupling is applied.
int num_coupled
number of target elements
int exclude_mask[MAX_CHANNELS]
Channels to be excluded from DRC processing.
void(* vector_pow43)(int *coefs, int len)
SingleChannelElement * output_element[MAX_CHANNELS]
Points to each SingleChannelElement.
Scalefactor data are intensity stereo positions (in phase).
int id_select[8]
element id
ChannelElement * tag_che_map[4][MAX_ELEM_ID]
Output configuration set in a global header but not yet locked.
int dyn_rng_sgn[17]
DRC sign information; 0 - positive, 1 - negative.
float coeffs[1024]
coefficients for IMDCT, maybe processed
PredictorState predictor_state[MAX_PREDICTORS]
SpectralBandReplication sbr
enum CouplingPoint coupling_point
The point during decoding at which coupling is applied.
float is_ener[128]
Intensity stereo pos (used by encoder)
Spectral Band Replication definitions and structures.
uint8_t max_sfb
number of scalefactor bands per group
static const uint8_t offset[127][2]
int num_swb
number of scalefactor window bands
Libavcodec external API header.
int prog_ref_level
A reference level for the long-term program audio level for all channels combined.
Output configuration locked in place.
int warned_remapping_once
enum BandType band_alt[128]
alternative band type (used by encoder)
enum WindowSequence window_sequence[2]
int predictor_reset_group
int dyn_rng_ctl[17]
DRC magnitude information.
typedef void(APIENTRY *FF_PFNGLACTIVETEXTUREPROC)(GLenum texture)
void ff_aacdec_init_mips(AACContext *c)
int pce_instance_tag
Indicates with which program the DRC info is associated.
int interpolation_scheme
Indicates the interpolation scheme used in the SBR QMF domain.
int ch_select[8]
[0] shared list of gains; [1] list of gains for right channel; [2] list of gains for left channel; [3...
int force_dmono_mode
0->not dmono, 1->use first channel, 2->use second channel
int is_saved
Set if elements have stored overlap from previous frame.
int warned_num_aac_frames
main external API structure.
IndividualChannelStream ics
uint8_t pi<< 24) CONV_FUNC_GROUP(AV_SAMPLE_FMT_FLT, float, AV_SAMPLE_FMT_U8, uint8_t,(*(constuint8_t *) pi-0x80)*(1.0f/(1<< 7))) CONV_FUNC_GROUP(AV_SAMPLE_FMT_DBL, double, AV_SAMPLE_FMT_U8, uint8_t,(*(constuint8_t *) pi-0x80)*(1.0/(1<< 7))) CONV_FUNC_GROUP(AV_SAMPLE_FMT_U8, uint8_t, AV_SAMPLE_FMT_S16, int16_t,(*(constint16_t *) pi >>8)+0x80) CONV_FUNC_GROUP(AV_SAMPLE_FMT_FLT, float, AV_SAMPLE_FMT_S16, int16_t,*(constint16_t *) pi *(1.0f/(1<< 15))) CONV_FUNC_GROUP(AV_SAMPLE_FMT_DBL, double, AV_SAMPLE_FMT_S16, int16_t,*(constint16_t *) pi *(1.0/(1<< 15))) CONV_FUNC_GROUP(AV_SAMPLE_FMT_U8, uint8_t, AV_SAMPLE_FMT_S32, int32_t,(*(constint32_t *) pi >>24)+0x80) CONV_FUNC_GROUP(AV_SAMPLE_FMT_FLT, float, AV_SAMPLE_FMT_S32, int32_t,*(constint32_t *) pi *(1.0f/(1U<< 31))) CONV_FUNC_GROUP(AV_SAMPLE_FMT_DBL, double, AV_SAMPLE_FMT_S32, int32_t,*(constint32_t *) pi *(1.0/(1U<< 31))) CONV_FUNC_GROUP(AV_SAMPLE_FMT_U8, uint8_t, AV_SAMPLE_FMT_FLT, float, av_clip_uint8(lrintf(*(constfloat *) pi *(1<< 7))+0x80)) CONV_FUNC_GROUP(AV_SAMPLE_FMT_S16, int16_t, AV_SAMPLE_FMT_FLT, float, av_clip_int16(lrintf(*(constfloat *) pi *(1<< 15)))) CONV_FUNC_GROUP(AV_SAMPLE_FMT_S32, int32_t, AV_SAMPLE_FMT_FLT, float, av_clipl_int32(llrintf(*(constfloat *) pi *(1U<< 31)))) CONV_FUNC_GROUP(AV_SAMPLE_FMT_U8, uint8_t, AV_SAMPLE_FMT_DBL, double, av_clip_uint8(lrint(*(constdouble *) pi *(1<< 7))+0x80)) CONV_FUNC_GROUP(AV_SAMPLE_FMT_S16, int16_t, AV_SAMPLE_FMT_DBL, double, av_clip_int16(lrint(*(constdouble *) pi *(1<< 15)))) CONV_FUNC_GROUP(AV_SAMPLE_FMT_S32, int32_t, AV_SAMPLE_FMT_DBL, double, av_clipl_int32(llrint(*(constdouble *) pi *(1U<< 31))))#defineSET_CONV_FUNC_GROUP(ofmt, ifmt) staticvoidset_generic_function(AudioConvert *ac){}voidff_audio_convert_free(AudioConvert **ac){if(!*ac) return;ff_dither_free(&(*ac) ->dc);av_freep(ac);}AudioConvert *ff_audio_convert_alloc(AVAudioResampleContext *avr, enumAVSampleFormatout_fmt, enumAVSampleFormatin_fmt, intchannels, intsample_rate, intapply_map){AudioConvert *ac;intin_planar, out_planar;ac=av_mallocz(sizeof(*ac));if(!ac) returnNULL;ac->avr=avr;ac->out_fmt=out_fmt;ac->in_fmt=in_fmt;ac->channels=channels;ac->apply_map=apply_map;if(avr->dither_method!=AV_RESAMPLE_DITHER_NONE &&av_get_packed_sample_fmt(out_fmt)==AV_SAMPLE_FMT_S16 &&av_get_bytes_per_sample(in_fmt)>2){ac->dc=ff_dither_alloc(avr, out_fmt, in_fmt, channels, sample_rate, apply_map);if(!ac->dc){av_free(ac);returnNULL;}returnac;}in_planar=ff_sample_fmt_is_planar(in_fmt, channels);out_planar=ff_sample_fmt_is_planar(out_fmt, channels);if(in_planar==out_planar){ac->func_type=CONV_FUNC_TYPE_FLAT;ac->planes=in_planar?ac->channels:1;}elseif(in_planar) ac->func_type=CONV_FUNC_TYPE_INTERLEAVE;elseac->func_type=CONV_FUNC_TYPE_DEINTERLEAVE;set_generic_function(ac);if(ARCH_AARCH64) ff_audio_convert_init_aarch64(ac);if(ARCH_ARM) ff_audio_convert_init_arm(ac);if(ARCH_X86) ff_audio_convert_init_x86(ac);returnac;}intff_audio_convert(AudioConvert *ac, AudioData *out, AudioData *in){intuse_generic=1;intlen=in->nb_samples;intp;if(ac->dc){av_log(ac->avr, AV_LOG_TRACE,"%dsamples-audio_convert:%sto%s(dithered)\n", len, av_get_sample_fmt_name(ac->in_fmt), av_get_sample_fmt_name(ac->out_fmt));returnff_convert_dither(ac-> in
void(* apply_tns)(INTFLOAT coef[1024], TemporalNoiseShaping *tns, IndividualChannelStream *ics, int decode)
Describe the class of an AVClass context structure.
Spectral data are coded with an escape sequence.
const uint8_t * swb_sizes
table of scalefactor band sizes for a particular window
OCStatus
Output configuration status.
int ms_mode
Signals mid/side stereo flags coding mode (used by encoder)
Output configuration under trial specified by a frame header.
float ltp_state[3072]
time signal for LTP
uint8_t zeroes[128]
band is not coded (used by encoder)
int band_type_run_end[120]
band type run end points
int sf_idx[128]
scalefactor indices (used by encoder)
uint8_t is_mode
Set if any bands have been encoded using intensity stereo (used by encoder)
int band_top[17]
Indicates the top of the i-th DRC band in units of 4 spectral lines.
float prcoeffs[1024]
Main prediction coefs (used by encoder)
int predictor_reset_count[31]
used by encoder to count prediction resets
Scalefactor data are intensity stereo positions (out of phase).
static int decode(AVCodecContext *avctx, void *data, int *got_sub, AVPacket *avpkt)
INTFLOAT coef[8][4][TNS_MAX_ORDER]
DynamicRangeControl che_drc
OutputConfiguration oc[2]
void(* apply_ltp)(AACContext *ac, SingleChannelElement *sce)
uint8_t prediction_used[41]
Single Channel Element - used for both SCE and LFE elements.
void(* windowing_and_mdct_ltp)(AACContext *ac, INTFLOAT *out, INTFLOAT *in, IndividualChannelStream *ics)
Individual Channel Stream.
float clip_avoidance_factor
set if any window is near clipping to the necessary atennuation factor to avoid it ...
channel element - generic struct for SCE/CPE/CCE/LFE
Scalefactors and spectral data are all zero.
enum BandType band_type[128]
band types
uint8_t pi<< 24) CONV_FUNC_GROUP(AV_SAMPLE_FMT_FLT, float, AV_SAMPLE_FMT_U8, uint8_t,(*(constuint8_t *) pi-0x80)*(1.0f/(1<< 7))) CONV_FUNC_GROUP(AV_SAMPLE_FMT_DBL, double, AV_SAMPLE_FMT_U8, uint8_t,(*(constuint8_t *) pi-0x80)*(1.0/(1<< 7))) CONV_FUNC_GROUP(AV_SAMPLE_FMT_U8, uint8_t, AV_SAMPLE_FMT_S16, int16_t,(*(constint16_t *) pi >>8)+0x80) CONV_FUNC_GROUP(AV_SAMPLE_FMT_FLT, float, AV_SAMPLE_FMT_S16, int16_t,*(constint16_t *) pi *(1.0f/(1<< 15))) CONV_FUNC_GROUP(AV_SAMPLE_FMT_DBL, double, AV_SAMPLE_FMT_S16, int16_t,*(constint16_t *) pi *(1.0/(1<< 15))) CONV_FUNC_GROUP(AV_SAMPLE_FMT_U8, uint8_t, AV_SAMPLE_FMT_S32, int32_t,(*(constint32_t *) pi >>24)+0x80) CONV_FUNC_GROUP(AV_SAMPLE_FMT_FLT, float, AV_SAMPLE_FMT_S32, int32_t,*(constint32_t *) pi *(1.0f/(1U<< 31))) CONV_FUNC_GROUP(AV_SAMPLE_FMT_DBL, double, AV_SAMPLE_FMT_S32, int32_t,*(constint32_t *) pi *(1.0/(1U<< 31))) CONV_FUNC_GROUP(AV_SAMPLE_FMT_U8, uint8_t, AV_SAMPLE_FMT_FLT, float, av_clip_uint8(lrintf(*(constfloat *) pi *(1<< 7))+0x80)) CONV_FUNC_GROUP(AV_SAMPLE_FMT_S16, int16_t, AV_SAMPLE_FMT_FLT, float, av_clip_int16(lrintf(*(constfloat *) pi *(1<< 15)))) CONV_FUNC_GROUP(AV_SAMPLE_FMT_S32, int32_t, AV_SAMPLE_FMT_FLT, float, av_clipl_int32(llrintf(*(constfloat *) pi *(1U<< 31)))) CONV_FUNC_GROUP(AV_SAMPLE_FMT_U8, uint8_t, AV_SAMPLE_FMT_DBL, double, av_clip_uint8(lrint(*(constdouble *) pi *(1<< 7))+0x80)) CONV_FUNC_GROUP(AV_SAMPLE_FMT_S16, int16_t, AV_SAMPLE_FMT_DBL, double, av_clip_int16(lrint(*(constdouble *) pi *(1<< 15)))) CONV_FUNC_GROUP(AV_SAMPLE_FMT_S32, int32_t, AV_SAMPLE_FMT_DBL, double, av_clipl_int32(llrint(*(constdouble *) pi *(1U<< 31))))#defineSET_CONV_FUNC_GROUP(ofmt, ifmt) staticvoidset_generic_function(AudioConvert *ac){}voidff_audio_convert_free(AudioConvert **ac){if(!*ac) return;ff_dither_free(&(*ac) ->dc);av_freep(ac);}AudioConvert *ff_audio_convert_alloc(AVAudioResampleContext *avr, enumAVSampleFormatout_fmt, enumAVSampleFormatin_fmt, intchannels, intsample_rate, intapply_map){AudioConvert *ac;intin_planar, out_planar;ac=av_mallocz(sizeof(*ac));if(!ac) returnNULL;ac->avr=avr;ac->out_fmt=out_fmt;ac->in_fmt=in_fmt;ac->channels=channels;ac->apply_map=apply_map;if(avr->dither_method!=AV_RESAMPLE_DITHER_NONE &&av_get_packed_sample_fmt(out_fmt)==AV_SAMPLE_FMT_S16 &&av_get_bytes_per_sample(in_fmt)>2){ac->dc=ff_dither_alloc(avr, out_fmt, in_fmt, channels, sample_rate, apply_map);if(!ac->dc){av_free(ac);returnNULL;}returnac;}in_planar=ff_sample_fmt_is_planar(in_fmt, channels);out_planar=ff_sample_fmt_is_planar(out_fmt, channels);if(in_planar==out_planar){ac->func_type=CONV_FUNC_TYPE_FLAT;ac->planes=in_planar?ac->channels:1;}elseif(in_planar) ac->func_type=CONV_FUNC_TYPE_INTERLEAVE;elseac->func_type=CONV_FUNC_TYPE_DEINTERLEAVE;set_generic_function(ac);if(ARCH_AARCH64) ff_audio_convert_init_aarch64(ac);if(ARCH_ARM) ff_audio_convert_init_arm(ac);if(ARCH_X86) ff_audio_convert_init_x86(ac);returnac;}intff_audio_convert(AudioConvert *ac, AudioData *out, AudioData *in){intuse_generic=1;intlen=in->nb_samples;intp;if(ac->dc){av_log(ac->avr, AV_LOG_TRACE,"%dsamples-audio_convert:%sto%s(dithered)\n", len, av_get_sample_fmt_name(ac->in_fmt), av_get_sample_fmt_name(ac->out_fmt));returnff_convert_dither(ac-> out
float ret_buf[2048]
PCM output buffer.
int8_t used[MAX_LTP_LONG_SFB]
uint8_t is_mask[128]
Set if intensity stereo is used (used by encoder)
Spectral Band Replication.
This and later band types encode two values (rather than four) with one code word.