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.

235 lines
12KB

  1. /*
  2. * Common code between the AC-3 and E-AC-3 decoders
  3. * Copyright (c) 2007 Bartlomiej Wolowiec <bartek.wolowiec@gmail.com>
  4. *
  5. * This file is part of Libav.
  6. *
  7. * Libav 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. * Libav 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 Libav; if not, write to the Free Software
  19. * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
  20. */
  21. /**
  22. * @file
  23. * Common code between the AC-3 and E-AC-3 decoders.
  24. *
  25. * Summary of MDCT Coefficient Grouping:
  26. * The individual MDCT coefficient indices are often referred to in the
  27. * (E-)AC-3 specification as frequency bins. These bins are grouped together
  28. * into subbands of 12 coefficients each. The subbands are grouped together
  29. * into bands as defined in the bitstream by the band structures, which
  30. * determine the number of bands and the size of each band. The full spectrum
  31. * of 256 frequency bins is divided into 1 DC bin + 21 subbands = 253 bins.
  32. * This system of grouping coefficients is used for channel bandwidth, stereo
  33. * rematrixing, channel coupling, enhanced coupling, and spectral extension.
  34. *
  35. * +-+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+-+
  36. * |1| |12| | [12|12|12|12] | | | | | | | | | | | | |3|
  37. * +-+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+--+-+
  38. * ~~~ ~~~~ ~~~~~~~~~~~~~ ~~~
  39. * | | | |
  40. * | | | 3 unused frequency bins--+
  41. * | | |
  42. * | | +--1 band containing 4 subbands
  43. * | |
  44. * | +--1 subband of 12 frequency bins
  45. * |
  46. * +--DC frequency bin
  47. */
  48. #ifndef AVCODEC_AC3DEC_H
  49. #define AVCODEC_AC3DEC_H
  50. #include "libavutil/lfg.h"
  51. #include "ac3.h"
  52. #include "ac3dsp.h"
  53. #include "get_bits.h"
  54. #include "dsputil.h"
  55. #include "fft.h"
  56. #include "fmtconvert.h"
  57. #define AC3_OUTPUT_LFEON 8
  58. #define SPX_MAX_BANDS 17
  59. /** Large enough for maximum possible frame size when the specification limit is ignored */
  60. #define AC3_FRAME_BUFFER_SIZE 32768
  61. typedef struct {
  62. AVClass *class; ///< class for AVOptions
  63. AVCodecContext *avctx; ///< parent context
  64. AVFrame frame; ///< AVFrame for decoded output
  65. GetBitContext gbc; ///< bitstream reader
  66. ///@name Bit stream information
  67. ///@{
  68. int frame_type; ///< frame type (strmtyp)
  69. int substreamid; ///< substream identification
  70. int frame_size; ///< current frame size, in bytes
  71. int bit_rate; ///< stream bit rate, in bits-per-second
  72. int sample_rate; ///< sample frequency, in Hz
  73. int num_blocks; ///< number of audio blocks
  74. int bitstream_mode; ///< bitstream mode (bsmod)
  75. int channel_mode; ///< channel mode (acmod)
  76. int channel_layout; ///< channel layout
  77. int lfe_on; ///< lfe channel in use
  78. int channel_map; ///< custom channel map
  79. int center_mix_level; ///< Center mix level index
  80. int surround_mix_level; ///< Surround mix level index
  81. int eac3; ///< indicates if current frame is E-AC-3
  82. ///@}
  83. ///@name Frame syntax parameters
  84. int snr_offset_strategy; ///< SNR offset strategy (snroffststr)
  85. int block_switch_syntax; ///< block switch syntax enabled (blkswe)
  86. int dither_flag_syntax; ///< dither flag syntax enabled (dithflage)
  87. int bit_allocation_syntax; ///< bit allocation model syntax enabled (bamode)
  88. int fast_gain_syntax; ///< fast gain codes enabled (frmfgaincode)
  89. int dba_syntax; ///< delta bit allocation syntax enabled (dbaflde)
  90. int skip_syntax; ///< skip field syntax enabled (skipflde)
  91. ///@}
  92. ///@name Standard coupling
  93. int cpl_in_use[AC3_MAX_BLOCKS]; ///< coupling in use (cplinu)
  94. int cpl_strategy_exists[AC3_MAX_BLOCKS];///< coupling strategy exists (cplstre)
  95. int channel_in_cpl[AC3_MAX_CHANNELS]; ///< channel in coupling (chincpl)
  96. int phase_flags_in_use; ///< phase flags in use (phsflginu)
  97. int phase_flags[AC3_MAX_CPL_BANDS]; ///< phase flags (phsflg)
  98. int num_cpl_bands; ///< number of coupling bands (ncplbnd)
  99. uint8_t cpl_band_sizes[AC3_MAX_CPL_BANDS]; ///< number of coeffs in each coupling band
  100. int firstchincpl; ///< first channel in coupling
  101. int first_cpl_coords[AC3_MAX_CHANNELS]; ///< first coupling coordinates states (firstcplcos)
  102. int cpl_coords[AC3_MAX_CHANNELS][AC3_MAX_CPL_BANDS]; ///< coupling coordinates (cplco)
  103. ///@}
  104. ///@name Spectral extension
  105. ///@{
  106. int spx_in_use; ///< spectral extension in use (spxinu)
  107. uint8_t channel_uses_spx[AC3_MAX_CHANNELS]; ///< channel uses spectral extension (chinspx)
  108. int8_t spx_atten_code[AC3_MAX_CHANNELS]; ///< spx attenuation code (spxattencod)
  109. int spx_src_start_freq; ///< spx start frequency bin
  110. int spx_dst_end_freq; ///< spx end frequency bin
  111. int spx_dst_start_freq; ///< spx starting frequency bin for copying (copystartmant)
  112. ///< the copy region ends at the start of the spx region.
  113. int num_spx_bands; ///< number of spx bands (nspxbnds)
  114. uint8_t spx_band_sizes[SPX_MAX_BANDS]; ///< number of bins in each spx band
  115. uint8_t first_spx_coords[AC3_MAX_CHANNELS]; ///< first spx coordinates states (firstspxcos)
  116. float spx_noise_blend[AC3_MAX_CHANNELS][SPX_MAX_BANDS]; ///< spx noise blending factor (nblendfact)
  117. float spx_signal_blend[AC3_MAX_CHANNELS][SPX_MAX_BANDS];///< spx signal blending factor (sblendfact)
  118. ///@}
  119. ///@name Adaptive hybrid transform
  120. int channel_uses_aht[AC3_MAX_CHANNELS]; ///< channel AHT in use (chahtinu)
  121. int pre_mantissa[AC3_MAX_CHANNELS][AC3_MAX_COEFS][AC3_MAX_BLOCKS]; ///< pre-IDCT mantissas
  122. ///@}
  123. ///@name Channel
  124. int fbw_channels; ///< number of full-bandwidth channels
  125. int channels; ///< number of total channels
  126. int lfe_ch; ///< index of LFE channel
  127. float downmix_coeffs[AC3_MAX_CHANNELS][2]; ///< stereo downmix coefficients
  128. int downmixed; ///< indicates if coeffs are currently downmixed
  129. int output_mode; ///< output channel configuration
  130. int out_channels; ///< number of output channels
  131. ///@}
  132. ///@name Dynamic range
  133. float dynamic_range[2]; ///< dynamic range
  134. float drc_scale; ///< percentage of dynamic range compression to be applied
  135. ///@}
  136. ///@name Bandwidth
  137. int start_freq[AC3_MAX_CHANNELS]; ///< start frequency bin (strtmant)
  138. int end_freq[AC3_MAX_CHANNELS]; ///< end frequency bin (endmant)
  139. ///@}
  140. ///@name Rematrixing
  141. int num_rematrixing_bands; ///< number of rematrixing bands (nrematbnd)
  142. int rematrixing_flags[4]; ///< rematrixing flags (rematflg)
  143. ///@}
  144. ///@name Exponents
  145. int num_exp_groups[AC3_MAX_CHANNELS]; ///< Number of exponent groups (nexpgrp)
  146. int8_t dexps[AC3_MAX_CHANNELS][AC3_MAX_COEFS]; ///< decoded exponents
  147. int exp_strategy[AC3_MAX_BLOCKS][AC3_MAX_CHANNELS]; ///< exponent strategies (expstr)
  148. ///@}
  149. ///@name Bit allocation
  150. AC3BitAllocParameters bit_alloc_params; ///< bit allocation parameters
  151. int first_cpl_leak; ///< first coupling leak state (firstcplleak)
  152. int snr_offset[AC3_MAX_CHANNELS]; ///< signal-to-noise ratio offsets (snroffst)
  153. int fast_gain[AC3_MAX_CHANNELS]; ///< fast gain values/SMR's (fgain)
  154. uint8_t bap[AC3_MAX_CHANNELS][AC3_MAX_COEFS]; ///< bit allocation pointers
  155. int16_t psd[AC3_MAX_CHANNELS][AC3_MAX_COEFS]; ///< scaled exponents
  156. int16_t band_psd[AC3_MAX_CHANNELS][AC3_CRITICAL_BANDS]; ///< interpolated exponents
  157. int16_t mask[AC3_MAX_CHANNELS][AC3_CRITICAL_BANDS]; ///< masking curve values
  158. int dba_mode[AC3_MAX_CHANNELS]; ///< delta bit allocation mode
  159. int dba_nsegs[AC3_MAX_CHANNELS]; ///< number of delta segments
  160. uint8_t dba_offsets[AC3_MAX_CHANNELS][8]; ///< delta segment offsets
  161. uint8_t dba_lengths[AC3_MAX_CHANNELS][8]; ///< delta segment lengths
  162. uint8_t dba_values[AC3_MAX_CHANNELS][8]; ///< delta values for each segment
  163. ///@}
  164. ///@name Zero-mantissa dithering
  165. int dither_flag[AC3_MAX_CHANNELS]; ///< dither flags (dithflg)
  166. AVLFG dith_state; ///< for dither generation
  167. ///@}
  168. ///@name IMDCT
  169. int block_switch[AC3_MAX_CHANNELS]; ///< block switch flags (blksw)
  170. FFTContext imdct_512; ///< for 512 sample IMDCT
  171. FFTContext imdct_256; ///< for 256 sample IMDCT
  172. ///@}
  173. ///@name Optimization
  174. DSPContext dsp; ///< for optimization
  175. AC3DSPContext ac3dsp;
  176. FmtConvertContext fmt_conv; ///< optimized conversion functions
  177. float mul_bias; ///< scaling for float_to_int16 conversion
  178. ///@}
  179. ///@name Aligned arrays
  180. DECLARE_ALIGNED(16, int, fixed_coeffs)[AC3_MAX_CHANNELS][AC3_MAX_COEFS]; ///< fixed-point transform coefficients
  181. DECLARE_ALIGNED(32, float, transform_coeffs)[AC3_MAX_CHANNELS][AC3_MAX_COEFS]; ///< transform coefficients
  182. DECLARE_ALIGNED(32, float, delay)[AC3_MAX_CHANNELS][AC3_BLOCK_SIZE]; ///< delay - added to the next block
  183. DECLARE_ALIGNED(32, float, window)[AC3_BLOCK_SIZE]; ///< window coefficients
  184. DECLARE_ALIGNED(32, float, tmp_output)[AC3_BLOCK_SIZE]; ///< temporary storage for output before windowing
  185. DECLARE_ALIGNED(32, float, output)[AC3_MAX_CHANNELS][AC3_BLOCK_SIZE]; ///< output after imdct transform and windowing
  186. DECLARE_ALIGNED(32, uint8_t, input_buffer)[AC3_FRAME_BUFFER_SIZE + FF_INPUT_BUFFER_PADDING_SIZE]; ///< temp buffer to prevent overread
  187. ///@}
  188. } AC3DecodeContext;
  189. /**
  190. * Parse the E-AC-3 frame header.
  191. * This parses both the bit stream info and audio frame header.
  192. */
  193. int ff_eac3_parse_header(AC3DecodeContext *s);
  194. /**
  195. * Decode mantissas in a single channel for the entire frame.
  196. * This is used when AHT mode is enabled.
  197. */
  198. void ff_eac3_decode_transform_coeffs_aht_ch(AC3DecodeContext *s, int ch);
  199. void ff_ac3_downmix_c(float (*samples)[256], float (*matrix)[2],
  200. int out_ch, int in_ch, int len);
  201. /**
  202. * Apply spectral extension to each channel by copying lower frequency
  203. * coefficients to higher frequency bins and applying side information to
  204. * approximate the original high frequency signal.
  205. */
  206. void ff_eac3_apply_spectral_extension(AC3DecodeContext *s);
  207. #endif /* AVCODEC_AC3DEC_H */