diff options
Diffstat (limited to 'libavcodec/dca.h')
| -rw-r--r-- | libavcodec/dca.h | 387 |
1 files changed, 124 insertions, 263 deletions
diff --git a/libavcodec/dca.h b/libavcodec/dca.h index 787a9c74de..bd96bc9ee3 100644 --- a/libavcodec/dca.h +++ b/libavcodec/dca.h @@ -4,21 +4,22 @@ * Copyright (C) 2004 Benjamin Zores * Copyright (C) 2006 Benjamin Larsson * Copyright (C) 2007 Konstantin Shishkov + * Copyright (C) 2016 foo86 * - * This file is part of Libav. + * This file is part of FFmpeg. * - * Libav is free software; you can redistribute it and/or + * 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. * - * Libav is distributed in the hope that it will be useful, + * 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 Libav; if not, write to the Free Software + * License along with FFmpeg; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA */ @@ -27,279 +28,139 @@ #include <stdint.h> -#include "libavutil/float_dsp.h" +#include "libavutil/common.h" #include "libavutil/internal.h" +#include "libavutil/intreadwrite.h" + +enum DCASpeaker { + DCA_SPEAKER_C, DCA_SPEAKER_L, DCA_SPEAKER_R, DCA_SPEAKER_Ls, + DCA_SPEAKER_Rs, DCA_SPEAKER_LFE1, DCA_SPEAKER_Cs, DCA_SPEAKER_Lsr, + DCA_SPEAKER_Rsr, DCA_SPEAKER_Lss, DCA_SPEAKER_Rss, DCA_SPEAKER_Lc, + DCA_SPEAKER_Rc, DCA_SPEAKER_Lh, DCA_SPEAKER_Ch, DCA_SPEAKER_Rh, + DCA_SPEAKER_LFE2, DCA_SPEAKER_Lw, DCA_SPEAKER_Rw, DCA_SPEAKER_Oh, + DCA_SPEAKER_Lhs, DCA_SPEAKER_Rhs, DCA_SPEAKER_Chr, DCA_SPEAKER_Lhr, + DCA_SPEAKER_Rhr, DCA_SPEAKER_Cl, DCA_SPEAKER_Ll, DCA_SPEAKER_Rl, + DCA_SPEAKER_RSV1, DCA_SPEAKER_RSV2, DCA_SPEAKER_RSV3, DCA_SPEAKER_RSV4, + + DCA_SPEAKER_COUNT +}; -#include "avcodec.h" -#include "dcadsp.h" -#include "fmtconvert.h" -#include "get_bits.h" - -#define DCA_PRIM_CHANNELS_MAX (7) -#define DCA_ABITS_MAX (32) /* Should be 28 */ -#define DCA_SUBSUBFRAMES_MAX (4) -#define DCA_SUBFRAMES_MAX (16) -#define DCA_BLOCKS_MAX (16) -#define DCA_LFE_MAX (3) - -#define DCA_PRIM_CHANNELS_MAX (7) -#define DCA_ABITS_MAX (32) /* Should be 28 */ -#define DCA_SUBSUBFRAMES_MAX (4) -#define DCA_SUBFRAMES_MAX (16) -#define DCA_BLOCKS_MAX (16) -#define DCA_LFE_MAX (3) -#define DCA_XLL_FBANDS_MAX (4) -#define DCA_XLL_SEGMENTS_MAX (16) -#define DCA_XLL_CHSETS_MAX (16) -#define DCA_XLL_CHANNELS_MAX (16) -#define DCA_XLL_AORDER_MAX (15) - -/* Arbitrary limit; not sure what the maximum really is, but much larger. */ -#define DCA_XLL_DMIX_NCOEFFS_MAX (18) +enum DCASpeakerMask { + DCA_SPEAKER_MASK_C = 0x00000001, + DCA_SPEAKER_MASK_L = 0x00000002, + DCA_SPEAKER_MASK_R = 0x00000004, + DCA_SPEAKER_MASK_Ls = 0x00000008, + DCA_SPEAKER_MASK_Rs = 0x00000010, + DCA_SPEAKER_MASK_LFE1 = 0x00000020, + DCA_SPEAKER_MASK_Cs = 0x00000040, + DCA_SPEAKER_MASK_Lsr = 0x00000080, + DCA_SPEAKER_MASK_Rsr = 0x00000100, + DCA_SPEAKER_MASK_Lss = 0x00000200, + DCA_SPEAKER_MASK_Rss = 0x00000400, + DCA_SPEAKER_MASK_Lc = 0x00000800, + DCA_SPEAKER_MASK_Rc = 0x00001000, + DCA_SPEAKER_MASK_Lh = 0x00002000, + DCA_SPEAKER_MASK_Ch = 0x00004000, + DCA_SPEAKER_MASK_Rh = 0x00008000, + DCA_SPEAKER_MASK_LFE2 = 0x00010000, + DCA_SPEAKER_MASK_Lw = 0x00020000, + DCA_SPEAKER_MASK_Rw = 0x00040000, + DCA_SPEAKER_MASK_Oh = 0x00080000, + DCA_SPEAKER_MASK_Lhs = 0x00100000, + DCA_SPEAKER_MASK_Rhs = 0x00200000, + DCA_SPEAKER_MASK_Chr = 0x00400000, + DCA_SPEAKER_MASK_Lhr = 0x00800000, + DCA_SPEAKER_MASK_Rhr = 0x01000000, + DCA_SPEAKER_MASK_Cl = 0x02000000, + DCA_SPEAKER_MASK_Ll = 0x04000000, + DCA_SPEAKER_MASK_Rl = 0x08000000, +}; -#define DCA_MAX_FRAME_SIZE 16384 -#define DCA_MAX_EXSS_HEADER_SIZE 4096 +#define DCA_SPEAKER_LAYOUT_MONO (DCA_SPEAKER_MASK_C) +#define DCA_SPEAKER_LAYOUT_STEREO (DCA_SPEAKER_MASK_L | DCA_SPEAKER_MASK_R) +#define DCA_SPEAKER_LAYOUT_2POINT1 (DCA_SPEAKER_LAYOUT_STEREO | DCA_SPEAKER_MASK_LFE1) +#define DCA_SPEAKER_LAYOUT_3_0 (DCA_SPEAKER_LAYOUT_STEREO | DCA_SPEAKER_MASK_C) +#define DCA_SPEAKER_LAYOUT_2_1 (DCA_SPEAKER_LAYOUT_STEREO | DCA_SPEAKER_MASK_Cs) +#define DCA_SPEAKER_LAYOUT_3_1 (DCA_SPEAKER_LAYOUT_3_0 | DCA_SPEAKER_MASK_Cs) +#define DCA_SPEAKER_LAYOUT_2_2 (DCA_SPEAKER_LAYOUT_STEREO | DCA_SPEAKER_MASK_Ls | DCA_SPEAKER_MASK_Rs) +#define DCA_SPEAKER_LAYOUT_5POINT0 (DCA_SPEAKER_LAYOUT_3_0 | DCA_SPEAKER_MASK_Ls | DCA_SPEAKER_MASK_Rs) +#define DCA_SPEAKER_LAYOUT_5POINT1 (DCA_SPEAKER_LAYOUT_5POINT0 | DCA_SPEAKER_MASK_LFE1) +#define DCA_SPEAKER_LAYOUT_7POINT0_WIDE (DCA_SPEAKER_LAYOUT_5POINT0 | DCA_SPEAKER_MASK_Lw | DCA_SPEAKER_MASK_Rw) +#define DCA_SPEAKER_LAYOUT_7POINT1_WIDE (DCA_SPEAKER_LAYOUT_7POINT0_WIDE | DCA_SPEAKER_MASK_LFE1) + +#define DCA_HAS_STEREO(mask) \ + ((mask & DCA_SPEAKER_LAYOUT_STEREO) == DCA_SPEAKER_LAYOUT_STEREO) + +enum DCASpeakerPair { + DCA_SPEAKER_PAIR_C = 0x0001, + DCA_SPEAKER_PAIR_LR = 0x0002, + DCA_SPEAKER_PAIR_LsRs = 0x0004, + DCA_SPEAKER_PAIR_LFE1 = 0x0008, + DCA_SPEAKER_PAIR_Cs = 0x0010, + DCA_SPEAKER_PAIR_LhRh = 0x0020, + DCA_SPEAKER_PAIR_LsrRsr = 0x0040, + DCA_SPEAKER_PAIR_Ch = 0x0080, + DCA_SPEAKER_PAIR_Oh = 0x0100, + DCA_SPEAKER_PAIR_LcRc = 0x0200, + DCA_SPEAKER_PAIR_LwRw = 0x0400, + DCA_SPEAKER_PAIR_LssRss = 0x0800, + DCA_SPEAKER_PAIR_LFE2 = 0x1000, + DCA_SPEAKER_PAIR_LhsRhs = 0x2000, + DCA_SPEAKER_PAIR_Chr = 0x4000, + DCA_SPEAKER_PAIR_LhrRhr = 0x8000 +}; -#define DCA_BUFFER_PADDING_SIZE 1024 +/** + * Return number of individual channels in DCASpeakerPair mask + */ +static inline int ff_dca_count_chs_for_mask(unsigned int mask) +{ + return av_popcount((mask & 0xffff) | ((mask & 0xae66) << 16)); +} + +enum DCARepresentationType { + DCA_REPR_TYPE_LtRt = 2, + DCA_REPR_TYPE_LhRh = 3 +}; enum DCAExtensionMask { - DCA_EXT_CORE = 0x001, ///< core in core substream - DCA_EXT_XXCH = 0x002, ///< XXCh channels extension in core substream - DCA_EXT_X96 = 0x004, ///< 96/24 extension in core substream - DCA_EXT_XCH = 0x008, ///< XCh channel extension in core substream - DCA_EXT_EXSS_CORE = 0x010, ///< core in ExSS (extension substream) - DCA_EXT_EXSS_XBR = 0x020, ///< extended bitrate extension in ExSS - DCA_EXT_EXSS_XXCH = 0x040, ///< XXCh channels extension in ExSS - DCA_EXT_EXSS_X96 = 0x080, ///< 96/24 extension in ExSS - DCA_EXT_EXSS_LBR = 0x100, ///< low bitrate component in ExSS - DCA_EXT_EXSS_XLL = 0x200, ///< lossless extension in ExSS + DCA_CSS_CORE = 0x001, + DCA_CSS_XXCH = 0x002, + DCA_CSS_X96 = 0x004, + DCA_CSS_XCH = 0x008, + DCA_CSS_MASK = 0x00f, + DCA_EXSS_CORE = 0x010, + DCA_EXSS_XBR = 0x020, + DCA_EXSS_XXCH = 0x040, + DCA_EXSS_X96 = 0x080, + DCA_EXSS_LBR = 0x100, + DCA_EXSS_XLL = 0x200, + DCA_EXSS_RSV1 = 0x400, + DCA_EXSS_RSV2 = 0x800, + DCA_EXSS_MASK = 0xff0, }; -typedef struct XllChSetSubHeader { - int channels; ///< number of channels in channel set, at most 16 - int residual_encode; ///< residual channel encoding - int bit_resolution; ///< input sample bit-width - int bit_width; ///< original input sample bit-width - int sampling_frequency; ///< sampling frequency - int samp_freq_interp; ///< sampling frequency interpolation multiplier - int replacement_set; ///< replacement channel set group - int active_replace_set; ///< current channel set is active channel set - int primary_ch_set; - int downmix_coeff_code_embedded; - int downmix_embedded; - int downmix_type; - int hier_chset; ///< hierarchical channel set - int downmix_ncoeffs; - int downmix_coeffs[DCA_XLL_DMIX_NCOEFFS_MAX]; - int ch_mask_enabled; - int ch_mask; - int mapping_coeffs_present; - int num_freq_bands; - - /* m_nOrigChanOrder */ - uint8_t orig_chan_order[DCA_XLL_FBANDS_MAX][DCA_XLL_CHANNELS_MAX]; - uint8_t orig_chan_order_inv[DCA_XLL_FBANDS_MAX][DCA_XLL_CHANNELS_MAX]; - /* Coefficients for channel pairs (at most 8), m_anPWChPairsCoeffs */ - int8_t pw_ch_pairs_coeffs[DCA_XLL_FBANDS_MAX][DCA_XLL_CHANNELS_MAX/2]; - /* m_nCurrHighestLPCOrder */ - uint8_t adapt_order_max[DCA_XLL_FBANDS_MAX]; - /* m_pnAdaptPredOrder */ - uint8_t adapt_order[DCA_XLL_FBANDS_MAX][DCA_XLL_CHANNELS_MAX]; - /* m_pnFixedPredOrder */ - uint8_t fixed_order[DCA_XLL_FBANDS_MAX][DCA_XLL_CHANNELS_MAX]; - /* m_pnLPCReflCoeffsQInd, unsigned version */ - uint8_t lpc_refl_coeffs_q_ind[DCA_XLL_FBANDS_MAX] - [DCA_XLL_CHANNELS_MAX][DCA_XLL_AORDER_MAX]; - - int lsb_fsize[DCA_XLL_FBANDS_MAX]; - int8_t scalable_lsbs[DCA_XLL_FBANDS_MAX][DCA_XLL_CHANNELS_MAX]; - int8_t bit_width_adj_per_ch[DCA_XLL_FBANDS_MAX][DCA_XLL_CHANNELS_MAX]; -} XllChSetSubHeader; - -typedef struct XllNavi { - GetBitContext gb; // Context for parsing the data segments - unsigned band_size[DCA_XLL_FBANDS_MAX]; - unsigned segment_size[DCA_XLL_FBANDS_MAX][DCA_XLL_SEGMENTS_MAX]; - unsigned chset_size[DCA_XLL_FBANDS_MAX][DCA_XLL_SEGMENTS_MAX][DCA_XLL_CHSETS_MAX]; -} XllNavi; - -typedef struct QMF64_table { - float dct4_coeff[32][32]; - float dct2_coeff[32][32]; - float rcos[32]; - float rsin[32]; -} QMF64_table; - -/* Primary audio coding header */ -typedef struct DCAAudioHeader { - int subband_activity[DCA_PRIM_CHANNELS_MAX]; ///< subband activity count - int vq_start_subband[DCA_PRIM_CHANNELS_MAX]; ///< high frequency vq start subband - int joint_intensity[DCA_PRIM_CHANNELS_MAX]; ///< joint intensity coding index - int transient_huffman[DCA_PRIM_CHANNELS_MAX]; ///< transient mode code book - int scalefactor_huffman[DCA_PRIM_CHANNELS_MAX]; ///< scale factor code book - int bitalloc_huffman[DCA_PRIM_CHANNELS_MAX]; ///< bit allocation quantizer select - int quant_index_huffman[DCA_PRIM_CHANNELS_MAX][DCA_ABITS_MAX]; ///< quantization index codebook select - uint32_t scalefactor_adj[DCA_PRIM_CHANNELS_MAX][DCA_ABITS_MAX]; ///< scale factor adjustment - - int subframes; ///< number of subframes - int total_channels; ///< number of channels including extensions - int prim_channels; ///< number of primary audio channels -} DCAAudioHeader; - -typedef struct DCAChan { - DECLARE_ALIGNED(32, int32_t, subband_samples)[DCA_BLOCKS_MAX][DCA_SUBBANDS][SAMPLES_PER_SUBBAND]; - - /* Subband samples history (for ADPCM) */ - DECLARE_ALIGNED(32, int32_t, subband_samples_hist)[DCA_SUBBANDS][4]; - int hist_index; - - /* Half size is sufficient for core decoding, but for 96 kHz data - * we need QMF with 64 subbands and 1024 samples. */ - DECLARE_ALIGNED(32, float, subband_fir_hist)[1024]; - DECLARE_ALIGNED(32, float, subband_fir_noidea)[64]; - - /* Primary audio coding side information */ - int prediction_mode[DCA_SUBBANDS]; ///< prediction mode (ADPCM used or not) - int prediction_vq[DCA_SUBBANDS]; ///< prediction VQ coefs - int bitalloc[DCA_SUBBANDS]; ///< bit allocation index - int transition_mode[DCA_SUBBANDS]; ///< transition mode (transients) - int32_t scale_factor[DCA_SUBBANDS][2];///< scale factors (2 if transient) - int joint_huff; ///< joint subband scale factors codebook - int joint_scale_factor[DCA_SUBBANDS]; ///< joint subband scale factors - - int32_t high_freq_vq[DCA_SUBBANDS]; ///< VQ encoded high frequency subbands -} DCAChan; - - -typedef struct DCAContext { - AVClass *class; ///< class for AVOptions - AVCodecContext *avctx; - /* Frame header */ - int frame_type; ///< type of the current frame - int samples_deficit; ///< deficit sample count - int crc_present; ///< crc is present in the bitstream - int sample_blocks; ///< number of PCM sample blocks - int frame_size; ///< primary frame byte size - int amode; ///< audio channels arrangement - int sample_rate; ///< audio sampling rate - int bit_rate; ///< transmission bit rate - int bit_rate_index; ///< transmission bit rate index +enum DCADownMixType { + DCA_DMIX_TYPE_1_0, + DCA_DMIX_TYPE_LoRo, + DCA_DMIX_TYPE_LtRt, + DCA_DMIX_TYPE_3_0, + DCA_DMIX_TYPE_2_1, + DCA_DMIX_TYPE_2_2, + DCA_DMIX_TYPE_3_1, - int dynrange; ///< embedded dynamic range flag - int timestamp; ///< embedded time stamp flag - int aux_data; ///< auxiliary data flag - int hdcd; ///< source material is mastered in HDCD - int ext_descr; ///< extension audio descriptor flag - int ext_coding; ///< extended coding flag - int aspf; ///< audio sync word insertion flag - int lfe; ///< low frequency effects flag - int predictor_history; ///< predictor history flag - int header_crc; ///< header crc check bytes - int multirate_inter; ///< multirate interpolator switch - int version; ///< encoder software revision - int copy_history; ///< copy history - int source_pcm_res; ///< source pcm resolution - int front_sum; ///< front sum/difference flag - int surround_sum; ///< surround sum/difference flag - int dialog_norm; ///< dialog normalisation parameter - - /* Primary audio coding header */ - DCAAudioHeader audio_header; - - /* Primary audio coding side information */ - int subsubframes[DCA_SUBFRAMES_MAX]; ///< number of subsubframes - int partial_samples[DCA_SUBFRAMES_MAX]; ///< partial subsubframe samples count - float downmix_coef[DCA_PRIM_CHANNELS_MAX + 1][2]; ///< stereo downmix coefficients - int dynrange_coef; ///< dynamic range coefficient - - /* Core substream's embedded downmix coefficients (cf. ETSI TS 102 114 V1.4.1) - * Input: primary audio channels (incl. LFE if present) - * Output: downmix audio channels (up to 4, no LFE) */ - uint8_t core_downmix; ///< embedded downmix coefficients available - uint8_t core_downmix_amode; ///< audio channel arrangement of embedded downmix - uint16_t core_downmix_codes[DCA_PRIM_CHANNELS_MAX + 1][4]; ///< embedded downmix coefficients (9-bit codes) - - - float lfe_data[2 * DCA_LFE_MAX * (DCA_BLOCKS_MAX + 4)]; ///< Low frequency effect data - int lfe_scale_factor; - - /* Subband samples history (for ADPCM) */ - DECLARE_ALIGNED(32, float, raXin)[32]; - - DCAChan dca_chan[DCA_PRIM_CHANNELS_MAX]; - - int output; ///< type of output - - float *samples_chanptr[DCA_PRIM_CHANNELS_MAX + 1]; - float *extra_channels[DCA_PRIM_CHANNELS_MAX + 1]; - uint8_t *extra_channels_buffer; - unsigned int extra_channels_buffer_size; - - uint8_t dca_buffer[DCA_MAX_FRAME_SIZE + DCA_MAX_EXSS_HEADER_SIZE + DCA_BUFFER_PADDING_SIZE]; - int dca_buffer_size; ///< how much data is in the dca_buffer - - const int8_t *channel_order_tab; ///< channel reordering table, lfe and non lfe - GetBitContext gb; - /* Current position in DCA frame */ - int current_subframe; - int current_subsubframe; - - int core_ext_mask; ///< present extensions in the core substream - int exss_ext_mask; ///< Non-core extensions - - /* XCh extension information */ - int xch_present; ///< XCh extension present and valid - int xch_base_channel; ///< index of first (only) channel containing XCH data - int xch_disable; ///< whether the XCh extension should be decoded or not - - /* XLL extension information */ - int xll_disable; - int xll_nch_sets; ///< number of channel sets per frame - int xll_channels; ///< total number of channels (in all channel sets) - int xll_residual_channels; ///< number of residual channels - int xll_segments; ///< number of segments per frame - int xll_log_smpl_in_seg; ///< supposedly this is "nBits4SamplLoci" - int xll_smpl_in_seg; ///< samples in segment per one frequency band for the first channel set - int xll_bits4seg_size; ///< number of bits used to read segment size - int xll_banddata_crc; ///< presence of CRC16 within each frequency band - int xll_scalable_lsb; - int xll_bits4ch_mask; ///< channel position mask - int xll_fixed_lsb_width; - XllChSetSubHeader xll_chsets[DCA_XLL_CHSETS_MAX]; - XllNavi xll_navi; - int *xll_sample_buf; - unsigned int xll_sample_buf_size; - - /* ExSS header parser */ - int static_fields; ///< static fields present - int mix_metadata; ///< mixing metadata present - int num_mix_configs; ///< number of mix out configurations - int mix_config_num_ch[4]; ///< number of channels in each mix out configuration - - int profile; - int one2one_map_chtospkr; - - int debug_flag; ///< used for suppressing repeated error messages output - AVFloatDSPContext fdsp; - FFTContext imdct; - SynthFilterContext synth; - DCADSPContext dcadsp; - QMF64_table *qmf64_table; - FmtConvertContext fmt_conv; -} DCAContext; + DCA_DMIX_TYPE_COUNT +}; extern av_export const uint32_t avpriv_dca_sample_rates[16]; +extern const uint32_t ff_dca_sampling_freqs[16]; +extern const uint8_t ff_dca_freq_ranges[16]; + /** * Convert bitstream to one representation based on sync marker */ -int ff_dca_convert_bitstream(const uint8_t *src, int src_size, uint8_t *dst, - int max_size); - -void ff_dca_exss_parse_header(DCAContext *s); - -int ff_dca_xll_decode_header(DCAContext *s); -int ff_dca_xll_decode_navi(DCAContext *s, int asset_end); -int ff_dca_xll_decode_audio(DCAContext *s, AVFrame *frame); +int avpriv_dca_convert_bitstream(const uint8_t *src, int src_size, uint8_t *dst, + int max_size); #endif /* AVCODEC_DCA_H */ |