2 * Copyright (C) 2011-2013 Michael Niedermayer (michaelni@gmx.at)
4 * This file is part of libswresample
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.
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.
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
21 #ifndef SWR_INTERNAL_H
22 #define SWR_INTERNAL_H
24 #include "swresample.h"
25 #include "libavutil/channel_layout.h"
28 #define SQRT3_2 1.22474487139158904909 /* sqrt(3/2) */
33 typedef int64_t integer;
38 typedef void (mix_1_1_func_type)(void *out, const void *in, void *coeffp, integer index, integer len);
39 typedef void (mix_2_1_func_type)(void *out, const void *in1, const void *in2, void *coeffp, integer index1, integer index2, integer len);
41 typedef void (mix_any_func_type)(uint8_t **out, const uint8_t **in1, void *coeffp, integer len);
43 typedef struct AudioData{
44 uint8_t *ch[SWR_CH_MAX]; ///< samples buffer per channel
45 uint8_t *data; ///< samples buffer
46 int ch_count; ///< number of channels
47 int bps; ///< bytes per sample
48 int count; ///< number of samples
49 int planar; ///< 1 if planar audio, 0 otherwise
50 enum AVSampleFormat fmt; ///< sample format
53 struct DitherContext {
54 enum SwrDitherType method;
57 float noise_scale; ///< Noise scale
58 int ns_taps; ///< Noise shaping dither taps
59 float ns_scale; ///< Noise shaping dither scale
60 float ns_scale_1; ///< Noise shaping dither scale^-1
61 int ns_pos; ///< Noise shaping dither position
62 float ns_coeffs[NS_TAPS]; ///< Noise shaping filter coefficients
63 float ns_errors[SWR_CH_MAX][2*NS_TAPS];
64 AudioData noise; ///< noise used for dithering
65 AudioData temp; ///< temporary storage when writing into the input buffer isnt possible
66 int output_sample_bits; ///< the number of used output bits, needed to scale dither correctly
70 const AVClass *av_class; ///< AVClass used for AVOption and av_log()
71 int log_level_offset; ///< logging level offset
72 void *log_ctx; ///< parent logging context
73 enum AVSampleFormat in_sample_fmt; ///< input sample format
74 enum AVSampleFormat int_sample_fmt; ///< internal sample format (AV_SAMPLE_FMT_FLTP or AV_SAMPLE_FMT_S16P)
75 enum AVSampleFormat out_sample_fmt; ///< output sample format
76 int64_t in_ch_layout; ///< input channel layout
77 int64_t out_ch_layout; ///< output channel layout
78 int in_sample_rate; ///< input sample rate
79 int out_sample_rate; ///< output sample rate
80 int flags; ///< miscellaneous flags such as SWR_FLAG_RESAMPLE
81 float slev; ///< surround mixing level
82 float clev; ///< center mixing level
83 float lfe_mix_level; ///< LFE mixing level
84 float rematrix_volume; ///< rematrixing volume coefficient
85 enum AVMatrixEncoding matrix_encoding; /**< matrixed stereo encoding */
86 const int *channel_map; ///< channel index (or -1 if muted channel) map
87 int used_ch_count; ///< number of used input channels (mapped channel count if channel_map, otherwise in.ch_count)
88 enum SwrEngine engine;
90 struct DitherContext dither;
92 int filter_size; /**< length of each FIR filter in the resampling filterbank relative to the cutoff frequency */
93 int phase_shift; /**< log2 of the number of entries in the resampling polyphase filterbank */
94 int linear_interp; /**< if 1 then the resampling FIR filter will be linearly interpolated */
95 double cutoff; /**< resampling cutoff frequency (swr: 6dB point; soxr: 0dB point). 1.0 corresponds to half the output sample rate */
96 enum SwrFilterType filter_type; /**< swr resampling filter type */
97 int kaiser_beta; /**< swr beta value for Kaiser window (only applicable if filter_type == AV_FILTER_TYPE_KAISER) */
98 double precision; /**< soxr resampling precision (in bits) */
99 int cheby; /**< soxr: if 1 then passband rolloff will be none (Chebyshev) & irrational ratio approximation precision will be higher */
101 float min_compensation; ///< swr minimum below which no compensation will happen
102 float min_hard_compensation; ///< swr minimum below which no silence inject / sample drop will happen
103 float soft_compensation_duration; ///< swr duration over which soft compensation is applied
104 float max_soft_compensation; ///< swr maximum soft compensation in seconds over soft_compensation_duration
105 float async; ///< swr simple 1 parameter async, similar to ffmpegs -async
106 int64_t firstpts_in_samples; ///< swr first pts in samples
108 int resample_first; ///< 1 if resampling must come first, 0 if rematrixing
109 int rematrix; ///< flag to indicate if rematrixing is needed (basically if input and output layouts mismatch)
110 int rematrix_custom; ///< flag to indicate that a custom matrix has been defined
112 AudioData in; ///< input audio data
113 AudioData postin; ///< post-input audio data: used for rematrix/resample
114 AudioData midbuf; ///< intermediate audio data (postin/preout)
115 AudioData preout; ///< pre-output audio data: used for rematrix/resample
116 AudioData out; ///< converted output audio data
117 AudioData in_buffer; ///< cached audio data (convert and resample purpose)
118 AudioData silence; ///< temporary with silence
119 AudioData drop_temp; ///< temporary used to discard output
120 int in_buffer_index; ///< cached buffer position
121 int in_buffer_count; ///< cached buffer length
122 int resample_in_constraint; ///< 1 if the input end was reach before the output end, 0 otherwise
123 int flushed; ///< 1 if data is to be flushed and no further input is expected
124 int64_t outpts; ///< output PTS
125 int64_t firstpts; ///< first PTS
126 int drop_output; ///< number of output samples to drop
128 struct AudioConvert *in_convert; ///< input conversion context
129 struct AudioConvert *out_convert; ///< output conversion context
130 struct AudioConvert *full_convert; ///< full conversion context (single conversion for input and output)
131 struct ResampleContext *resample; ///< resampling context
132 struct Resampler const *resampler; ///< resampler virtual function table
134 float matrix[SWR_CH_MAX][SWR_CH_MAX]; ///< floating point rematrixing coefficients
135 uint8_t *native_matrix;
137 uint8_t *native_simd_matrix;
138 int32_t matrix32[SWR_CH_MAX][SWR_CH_MAX]; ///< 17.15 fixed point rematrixing coefficients
139 uint8_t matrix_ch[SWR_CH_MAX][SWR_CH_MAX+1]; ///< Lists of input channels per output channel that have non zero rematrixing coefficients
140 mix_1_1_func_type *mix_1_1_f;
141 mix_1_1_func_type *mix_1_1_simd;
143 mix_2_1_func_type *mix_2_1_f;
144 mix_2_1_func_type *mix_2_1_simd;
146 mix_any_func_type *mix_any_f;
148 /* TODO: callbacks for ASM optimizations */
151 typedef struct ResampleContext * (* resample_init_func)(struct ResampleContext *c, int out_rate, int in_rate, int filter_size, int phase_shift, int linear,
152 double cutoff, enum AVSampleFormat format, enum SwrFilterType filter_type, int kaiser_beta, double precision, int cheby);
153 typedef void (* resample_free_func)(struct ResampleContext **c);
154 typedef int (* multiple_resample_func)(struct ResampleContext *c, AudioData *dst, int dst_size, AudioData *src, int src_size, int *consumed);
155 typedef int (* resample_flush_func)(struct SwrContext *c);
156 typedef int (* set_compensation_func)(struct ResampleContext *c, int sample_delta, int compensation_distance);
157 typedef int64_t (* get_delay_func)(struct SwrContext *s, int64_t base);
160 resample_init_func init;
161 resample_free_func free;
162 multiple_resample_func multiple_resample;
163 resample_flush_func flush;
164 set_compensation_func set_compensation;
165 get_delay_func get_delay;
168 extern struct Resampler const swri_resampler;
170 int swri_realloc_audio(AudioData *a, int count);
171 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);
172 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);
173 int swri_resample_float(struct ResampleContext *c, float *dst, const float *src, int *consumed, int src_size, int dst_size, int update_ctx);
174 int swri_resample_double(struct ResampleContext *c,double *dst, const double *src, int *consumed, int src_size, int dst_size, int update_ctx);
176 void swri_noise_shaping_int16 (SwrContext *s, AudioData *dsts, const AudioData *srcs, const AudioData *noises, int count);
177 void swri_noise_shaping_int32 (SwrContext *s, AudioData *dsts, const AudioData *srcs, const AudioData *noises, int count);
178 void swri_noise_shaping_float (SwrContext *s, AudioData *dsts, const AudioData *srcs, const AudioData *noises, int count);
179 void swri_noise_shaping_double(SwrContext *s, AudioData *dsts, const AudioData *srcs, const AudioData *noises, int count);
181 int swri_rematrix_init(SwrContext *s);
182 void swri_rematrix_free(SwrContext *s);
183 int swri_rematrix(SwrContext *s, AudioData *out, AudioData *in, int len, int mustcopy);
184 void swri_rematrix_init_x86(struct SwrContext *s);
186 void swri_get_dither(SwrContext *s, void *dst, int len, unsigned seed, enum AVSampleFormat noise_fmt);
187 int swri_dither_init(SwrContext *s, enum AVSampleFormat out_fmt, enum AVSampleFormat in_fmt);
189 void swri_audio_convert_init_arm(struct AudioConvert *ac,
190 enum AVSampleFormat out_fmt,
191 enum AVSampleFormat in_fmt,
193 void swri_audio_convert_init_x86(struct AudioConvert *ac,
194 enum AVSampleFormat out_fmt,
195 enum AVSampleFormat in_fmt,