swresample_internal.h 12 KB

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168169170171172173174175176177178179180181182183184185186187188189190191192193194195196197198199
  1. /*
  2. * Copyright (C) 2011-2013 Michael Niedermayer (michaelni@gmx.at)
  3. *
  4. * This file is part of libswresample
  5. *
  6. * libswresample is free software; you can redistribute it and/or
  7. * modify it under the terms of the GNU Lesser General Public
  8. * License as published by the Free Software Foundation; either
  9. * version 2.1 of the License, or (at your option) any later version.
  10. *
  11. * libswresample is distributed in the hope that it will be useful,
  12. * but WITHOUT ANY WARRANTY; without even the implied warranty of
  13. * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
  14. * Lesser General Public License for more details.
  15. *
  16. * You should have received a copy of the GNU Lesser General Public
  17. * License along with libswresample; if not, write to the Free Software
  18. * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
  19. */
  20. #ifndef SWR_INTERNAL_H
  21. #define SWR_INTERNAL_H
  22. #include "swresample.h"
  23. #include "libavutil/channel_layout.h"
  24. #include "config.h"
  25. #define SQRT3_2 1.22474487139158904909 /* sqrt(3/2) */
  26. #define NS_TAPS 20
  27. #if ARCH_X86_64
  28. typedef int64_t integer;
  29. #else
  30. typedef int integer;
  31. #endif
  32. typedef void (mix_1_1_func_type)(void *out, const void *in, void *coeffp, integer index, integer len);
  33. typedef void (mix_2_1_func_type)(void *out, const void *in1, const void *in2, void *coeffp, integer index1, integer index2, integer len);
  34. typedef void (mix_any_func_type)(uint8_t **out, const uint8_t **in1, void *coeffp, integer len);
  35. typedef struct AudioData{
  36. uint8_t *ch[SWR_CH_MAX]; ///< samples buffer per channel
  37. uint8_t *data; ///< samples buffer
  38. int ch_count; ///< number of channels
  39. int bps; ///< bytes per sample
  40. int count; ///< number of samples
  41. int planar; ///< 1 if planar audio, 0 otherwise
  42. enum AVSampleFormat fmt; ///< sample format
  43. } AudioData;
  44. struct DitherContext {
  45. enum SwrDitherType method;
  46. int noise_pos;
  47. float scale;
  48. float noise_scale; ///< Noise scale
  49. int ns_taps; ///< Noise shaping dither taps
  50. float ns_scale; ///< Noise shaping dither scale
  51. float ns_scale_1; ///< Noise shaping dither scale^-1
  52. int ns_pos; ///< Noise shaping dither position
  53. float ns_coeffs[NS_TAPS]; ///< Noise shaping filter coefficients
  54. float ns_errors[SWR_CH_MAX][2*NS_TAPS];
  55. AudioData noise; ///< noise used for dithering
  56. AudioData temp; ///< temporary storage when writing into the input buffer isnt possible
  57. int output_sample_bits; ///< the number of used output bits, needed to scale dither correctly
  58. };
  59. struct SwrContext {
  60. const AVClass *av_class; ///< AVClass used for AVOption and av_log()
  61. int log_level_offset; ///< logging level offset
  62. void *log_ctx; ///< parent logging context
  63. enum AVSampleFormat in_sample_fmt; ///< input sample format
  64. enum AVSampleFormat int_sample_fmt; ///< internal sample format (AV_SAMPLE_FMT_FLTP or AV_SAMPLE_FMT_S16P)
  65. enum AVSampleFormat out_sample_fmt; ///< output sample format
  66. int64_t in_ch_layout; ///< input channel layout
  67. int64_t out_ch_layout; ///< output channel layout
  68. int in_sample_rate; ///< input sample rate
  69. int out_sample_rate; ///< output sample rate
  70. int flags; ///< miscellaneous flags such as SWR_FLAG_RESAMPLE
  71. float slev; ///< surround mixing level
  72. float clev; ///< center mixing level
  73. float lfe_mix_level; ///< LFE mixing level
  74. float rematrix_volume; ///< rematrixing volume coefficient
  75. float rematrix_maxval; ///< maximum value for rematrixing output
  76. enum AVMatrixEncoding matrix_encoding; /**< matrixed stereo encoding */
  77. const int *channel_map; ///< channel index (or -1 if muted channel) map
  78. int used_ch_count; ///< number of used input channels (mapped channel count if channel_map, otherwise in.ch_count)
  79. enum SwrEngine engine;
  80. struct DitherContext dither;
  81. int filter_size; /**< length of each FIR filter in the resampling filterbank relative to the cutoff frequency */
  82. int phase_shift; /**< log2 of the number of entries in the resampling polyphase filterbank */
  83. int linear_interp; /**< if 1 then the resampling FIR filter will be linearly interpolated */
  84. double cutoff; /**< resampling cutoff frequency (swr: 6dB point; soxr: 0dB point). 1.0 corresponds to half the output sample rate */
  85. enum SwrFilterType filter_type; /**< swr resampling filter type */
  86. int kaiser_beta; /**< swr beta value for Kaiser window (only applicable if filter_type == AV_FILTER_TYPE_KAISER) */
  87. double precision; /**< soxr resampling precision (in bits) */
  88. int cheby; /**< soxr: if 1 then passband rolloff will be none (Chebyshev) & irrational ratio approximation precision will be higher */
  89. float min_compensation; ///< swr minimum below which no compensation will happen
  90. float min_hard_compensation; ///< swr minimum below which no silence inject / sample drop will happen
  91. float soft_compensation_duration; ///< swr duration over which soft compensation is applied
  92. float max_soft_compensation; ///< swr maximum soft compensation in seconds over soft_compensation_duration
  93. float async; ///< swr simple 1 parameter async, similar to ffmpegs -async
  94. int64_t firstpts_in_samples; ///< swr first pts in samples
  95. int resample_first; ///< 1 if resampling must come first, 0 if rematrixing
  96. int rematrix; ///< flag to indicate if rematrixing is needed (basically if input and output layouts mismatch)
  97. int rematrix_custom; ///< flag to indicate that a custom matrix has been defined
  98. AudioData in; ///< input audio data
  99. AudioData postin; ///< post-input audio data: used for rematrix/resample
  100. AudioData midbuf; ///< intermediate audio data (postin/preout)
  101. AudioData preout; ///< pre-output audio data: used for rematrix/resample
  102. AudioData out; ///< converted output audio data
  103. AudioData in_buffer; ///< cached audio data (convert and resample purpose)
  104. AudioData silence; ///< temporary with silence
  105. AudioData drop_temp; ///< temporary used to discard output
  106. int in_buffer_index; ///< cached buffer position
  107. int in_buffer_count; ///< cached buffer length
  108. int resample_in_constraint; ///< 1 if the input end was reach before the output end, 0 otherwise
  109. int flushed; ///< 1 if data is to be flushed and no further input is expected
  110. int64_t outpts; ///< output PTS
  111. int64_t firstpts; ///< first PTS
  112. int drop_output; ///< number of output samples to drop
  113. struct AudioConvert *in_convert; ///< input conversion context
  114. struct AudioConvert *out_convert; ///< output conversion context
  115. struct AudioConvert *full_convert; ///< full conversion context (single conversion for input and output)
  116. struct ResampleContext *resample; ///< resampling context
  117. struct Resampler const *resampler; ///< resampler virtual function table
  118. float matrix[SWR_CH_MAX][SWR_CH_MAX]; ///< floating point rematrixing coefficients
  119. uint8_t *native_matrix;
  120. uint8_t *native_one;
  121. uint8_t *native_simd_one;
  122. uint8_t *native_simd_matrix;
  123. int32_t matrix32[SWR_CH_MAX][SWR_CH_MAX]; ///< 17.15 fixed point rematrixing coefficients
  124. uint8_t matrix_ch[SWR_CH_MAX][SWR_CH_MAX+1]; ///< Lists of input channels per output channel that have non zero rematrixing coefficients
  125. mix_1_1_func_type *mix_1_1_f;
  126. mix_1_1_func_type *mix_1_1_simd;
  127. mix_2_1_func_type *mix_2_1_f;
  128. mix_2_1_func_type *mix_2_1_simd;
  129. mix_any_func_type *mix_any_f;
  130. /* TODO: callbacks for ASM optimizations */
  131. };
  132. typedef struct ResampleContext * (* resample_init_func)(struct ResampleContext *c, int out_rate, int in_rate, int filter_size, int phase_shift, int linear,
  133. double cutoff, enum AVSampleFormat format, enum SwrFilterType filter_type, int kaiser_beta, double precision, int cheby);
  134. typedef void (* resample_free_func)(struct ResampleContext **c);
  135. typedef int (* multiple_resample_func)(struct ResampleContext *c, AudioData *dst, int dst_size, AudioData *src, int src_size, int *consumed);
  136. typedef int (* resample_flush_func)(struct SwrContext *c);
  137. typedef int (* set_compensation_func)(struct ResampleContext *c, int sample_delta, int compensation_distance);
  138. typedef int64_t (* get_delay_func)(struct SwrContext *s, int64_t base);
  139. struct Resampler {
  140. resample_init_func init;
  141. resample_free_func free;
  142. multiple_resample_func multiple_resample;
  143. resample_flush_func flush;
  144. set_compensation_func set_compensation;
  145. get_delay_func get_delay;
  146. };
  147. extern struct Resampler const swri_resampler;
  148. int swri_realloc_audio(AudioData *a, int count);
  149. int swri_resample_int16(struct ResampleContext *c, int16_t *dst, const int16_t *src, int *consumed, int src_size, int dst_size, int update_ctx);
  150. int swri_resample_int32(struct ResampleContext *c, int32_t *dst, const int32_t *src, int *consumed, int src_size, int dst_size, int update_ctx);
  151. int swri_resample_float(struct ResampleContext *c, float *dst, const float *src, int *consumed, int src_size, int dst_size, int update_ctx);
  152. int swri_resample_double(struct ResampleContext *c,double *dst, const double *src, int *consumed, int src_size, int dst_size, int update_ctx);
  153. void swri_noise_shaping_int16 (SwrContext *s, AudioData *dsts, const AudioData *srcs, const AudioData *noises, int count);
  154. void swri_noise_shaping_int32 (SwrContext *s, AudioData *dsts, const AudioData *srcs, const AudioData *noises, int count);
  155. void swri_noise_shaping_float (SwrContext *s, AudioData *dsts, const AudioData *srcs, const AudioData *noises, int count);
  156. void swri_noise_shaping_double(SwrContext *s, AudioData *dsts, const AudioData *srcs, const AudioData *noises, int count);
  157. int swri_rematrix_init(SwrContext *s);
  158. void swri_rematrix_free(SwrContext *s);
  159. int swri_rematrix(SwrContext *s, AudioData *out, AudioData *in, int len, int mustcopy);
  160. void swri_rematrix_init_x86(struct SwrContext *s);
  161. void swri_get_dither(SwrContext *s, void *dst, int len, unsigned seed, enum AVSampleFormat noise_fmt);
  162. int swri_dither_init(SwrContext *s, enum AVSampleFormat out_fmt, enum AVSampleFormat in_fmt);
  163. void swri_audio_convert_init_arm(struct AudioConvert *ac,
  164. enum AVSampleFormat out_fmt,
  165. enum AVSampleFormat in_fmt,
  166. int channels);
  167. void swri_audio_convert_init_x86(struct AudioConvert *ac,
  168. enum AVSampleFormat out_fmt,
  169. enum AVSampleFormat in_fmt,
  170. int channels);
  171. #endif