annotate src/opus-1.3/silk/stereo_LR_to_MS.c @ 83:ae30d91d2ffe

Replace these with versions built using an older toolset (so as to avoid ABI compatibilities when linking on Ubuntu 14.04 for packaging purposes)
author Chris Cannam
date Fri, 07 Feb 2020 11:51:13 +0000
parents 7aeed7906520
children
rev   line source
Chris@69 1 /***********************************************************************
Chris@69 2 Copyright (c) 2006-2011, Skype Limited. All rights reserved.
Chris@69 3 Redistribution and use in source and binary forms, with or without
Chris@69 4 modification, are permitted provided that the following conditions
Chris@69 5 are met:
Chris@69 6 - Redistributions of source code must retain the above copyright notice,
Chris@69 7 this list of conditions and the following disclaimer.
Chris@69 8 - Redistributions in binary form must reproduce the above copyright
Chris@69 9 notice, this list of conditions and the following disclaimer in the
Chris@69 10 documentation and/or other materials provided with the distribution.
Chris@69 11 - Neither the name of Internet Society, IETF or IETF Trust, nor the
Chris@69 12 names of specific contributors, may be used to endorse or promote
Chris@69 13 products derived from this software without specific prior written
Chris@69 14 permission.
Chris@69 15 THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
Chris@69 16 AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
Chris@69 17 IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
Chris@69 18 ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE
Chris@69 19 LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
Chris@69 20 CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
Chris@69 21 SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
Chris@69 22 INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
Chris@69 23 CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
Chris@69 24 ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
Chris@69 25 POSSIBILITY OF SUCH DAMAGE.
Chris@69 26 ***********************************************************************/
Chris@69 27
Chris@69 28 #ifdef HAVE_CONFIG_H
Chris@69 29 #include "config.h"
Chris@69 30 #endif
Chris@69 31
Chris@69 32 #include "main.h"
Chris@69 33 #include "stack_alloc.h"
Chris@69 34
Chris@69 35 /* Convert Left/Right stereo signal to adaptive Mid/Side representation */
Chris@69 36 void silk_stereo_LR_to_MS(
Chris@69 37 stereo_enc_state *state, /* I/O State */
Chris@69 38 opus_int16 x1[], /* I/O Left input signal, becomes mid signal */
Chris@69 39 opus_int16 x2[], /* I/O Right input signal, becomes side signal */
Chris@69 40 opus_int8 ix[ 2 ][ 3 ], /* O Quantization indices */
Chris@69 41 opus_int8 *mid_only_flag, /* O Flag: only mid signal coded */
Chris@69 42 opus_int32 mid_side_rates_bps[], /* O Bitrates for mid and side signals */
Chris@69 43 opus_int32 total_rate_bps, /* I Total bitrate */
Chris@69 44 opus_int prev_speech_act_Q8, /* I Speech activity level in previous frame */
Chris@69 45 opus_int toMono, /* I Last frame before a stereo->mono transition */
Chris@69 46 opus_int fs_kHz, /* I Sample rate (kHz) */
Chris@69 47 opus_int frame_length /* I Number of samples */
Chris@69 48 )
Chris@69 49 {
Chris@69 50 opus_int n, is10msFrame, denom_Q16, delta0_Q13, delta1_Q13;
Chris@69 51 opus_int32 sum, diff, smooth_coef_Q16, pred_Q13[ 2 ], pred0_Q13, pred1_Q13;
Chris@69 52 opus_int32 LP_ratio_Q14, HP_ratio_Q14, frac_Q16, frac_3_Q16, min_mid_rate_bps, width_Q14, w_Q24, deltaw_Q24;
Chris@69 53 VARDECL( opus_int16, side );
Chris@69 54 VARDECL( opus_int16, LP_mid );
Chris@69 55 VARDECL( opus_int16, HP_mid );
Chris@69 56 VARDECL( opus_int16, LP_side );
Chris@69 57 VARDECL( opus_int16, HP_side );
Chris@69 58 opus_int16 *mid = &x1[ -2 ];
Chris@69 59 SAVE_STACK;
Chris@69 60
Chris@69 61 ALLOC( side, frame_length + 2, opus_int16 );
Chris@69 62 /* Convert to basic mid/side signals */
Chris@69 63 for( n = 0; n < frame_length + 2; n++ ) {
Chris@69 64 sum = x1[ n - 2 ] + (opus_int32)x2[ n - 2 ];
Chris@69 65 diff = x1[ n - 2 ] - (opus_int32)x2[ n - 2 ];
Chris@69 66 mid[ n ] = (opus_int16)silk_RSHIFT_ROUND( sum, 1 );
Chris@69 67 side[ n ] = (opus_int16)silk_SAT16( silk_RSHIFT_ROUND( diff, 1 ) );
Chris@69 68 }
Chris@69 69
Chris@69 70 /* Buffering */
Chris@69 71 silk_memcpy( mid, state->sMid, 2 * sizeof( opus_int16 ) );
Chris@69 72 silk_memcpy( side, state->sSide, 2 * sizeof( opus_int16 ) );
Chris@69 73 silk_memcpy( state->sMid, &mid[ frame_length ], 2 * sizeof( opus_int16 ) );
Chris@69 74 silk_memcpy( state->sSide, &side[ frame_length ], 2 * sizeof( opus_int16 ) );
Chris@69 75
Chris@69 76 /* LP and HP filter mid signal */
Chris@69 77 ALLOC( LP_mid, frame_length, opus_int16 );
Chris@69 78 ALLOC( HP_mid, frame_length, opus_int16 );
Chris@69 79 for( n = 0; n < frame_length; n++ ) {
Chris@69 80 sum = silk_RSHIFT_ROUND( silk_ADD_LSHIFT( mid[ n ] + (opus_int32)mid[ n + 2 ], mid[ n + 1 ], 1 ), 2 );
Chris@69 81 LP_mid[ n ] = sum;
Chris@69 82 HP_mid[ n ] = mid[ n + 1 ] - sum;
Chris@69 83 }
Chris@69 84
Chris@69 85 /* LP and HP filter side signal */
Chris@69 86 ALLOC( LP_side, frame_length, opus_int16 );
Chris@69 87 ALLOC( HP_side, frame_length, opus_int16 );
Chris@69 88 for( n = 0; n < frame_length; n++ ) {
Chris@69 89 sum = silk_RSHIFT_ROUND( silk_ADD_LSHIFT( side[ n ] + (opus_int32)side[ n + 2 ], side[ n + 1 ], 1 ), 2 );
Chris@69 90 LP_side[ n ] = sum;
Chris@69 91 HP_side[ n ] = side[ n + 1 ] - sum;
Chris@69 92 }
Chris@69 93
Chris@69 94 /* Find energies and predictors */
Chris@69 95 is10msFrame = frame_length == 10 * fs_kHz;
Chris@69 96 smooth_coef_Q16 = is10msFrame ?
Chris@69 97 SILK_FIX_CONST( STEREO_RATIO_SMOOTH_COEF / 2, 16 ) :
Chris@69 98 SILK_FIX_CONST( STEREO_RATIO_SMOOTH_COEF, 16 );
Chris@69 99 smooth_coef_Q16 = silk_SMULWB( silk_SMULBB( prev_speech_act_Q8, prev_speech_act_Q8 ), smooth_coef_Q16 );
Chris@69 100
Chris@69 101 pred_Q13[ 0 ] = silk_stereo_find_predictor( &LP_ratio_Q14, LP_mid, LP_side, &state->mid_side_amp_Q0[ 0 ], frame_length, smooth_coef_Q16 );
Chris@69 102 pred_Q13[ 1 ] = silk_stereo_find_predictor( &HP_ratio_Q14, HP_mid, HP_side, &state->mid_side_amp_Q0[ 2 ], frame_length, smooth_coef_Q16 );
Chris@69 103 /* Ratio of the norms of residual and mid signals */
Chris@69 104 frac_Q16 = silk_SMLABB( HP_ratio_Q14, LP_ratio_Q14, 3 );
Chris@69 105 frac_Q16 = silk_min( frac_Q16, SILK_FIX_CONST( 1, 16 ) );
Chris@69 106
Chris@69 107 /* Determine bitrate distribution between mid and side, and possibly reduce stereo width */
Chris@69 108 total_rate_bps -= is10msFrame ? 1200 : 600; /* Subtract approximate bitrate for coding stereo parameters */
Chris@69 109 if( total_rate_bps < 1 ) {
Chris@69 110 total_rate_bps = 1;
Chris@69 111 }
Chris@69 112 min_mid_rate_bps = silk_SMLABB( 2000, fs_kHz, 600 );
Chris@69 113 silk_assert( min_mid_rate_bps < 32767 );
Chris@69 114 /* Default bitrate distribution: 8 parts for Mid and (5+3*frac) parts for Side. so: mid_rate = ( 8 / ( 13 + 3 * frac ) ) * total_ rate */
Chris@69 115 frac_3_Q16 = silk_MUL( 3, frac_Q16 );
Chris@69 116 mid_side_rates_bps[ 0 ] = silk_DIV32_varQ( total_rate_bps, SILK_FIX_CONST( 8 + 5, 16 ) + frac_3_Q16, 16+3 );
Chris@69 117 /* If Mid bitrate below minimum, reduce stereo width */
Chris@69 118 if( mid_side_rates_bps[ 0 ] < min_mid_rate_bps ) {
Chris@69 119 mid_side_rates_bps[ 0 ] = min_mid_rate_bps;
Chris@69 120 mid_side_rates_bps[ 1 ] = total_rate_bps - mid_side_rates_bps[ 0 ];
Chris@69 121 /* width = 4 * ( 2 * side_rate - min_rate ) / ( ( 1 + 3 * frac ) * min_rate ) */
Chris@69 122 width_Q14 = silk_DIV32_varQ( silk_LSHIFT( mid_side_rates_bps[ 1 ], 1 ) - min_mid_rate_bps,
Chris@69 123 silk_SMULWB( SILK_FIX_CONST( 1, 16 ) + frac_3_Q16, min_mid_rate_bps ), 14+2 );
Chris@69 124 width_Q14 = silk_LIMIT( width_Q14, 0, SILK_FIX_CONST( 1, 14 ) );
Chris@69 125 } else {
Chris@69 126 mid_side_rates_bps[ 1 ] = total_rate_bps - mid_side_rates_bps[ 0 ];
Chris@69 127 width_Q14 = SILK_FIX_CONST( 1, 14 );
Chris@69 128 }
Chris@69 129
Chris@69 130 /* Smoother */
Chris@69 131 state->smth_width_Q14 = (opus_int16)silk_SMLAWB( state->smth_width_Q14, width_Q14 - state->smth_width_Q14, smooth_coef_Q16 );
Chris@69 132
Chris@69 133 /* At very low bitrates or for inputs that are nearly amplitude panned, switch to panned-mono coding */
Chris@69 134 *mid_only_flag = 0;
Chris@69 135 if( toMono ) {
Chris@69 136 /* Last frame before stereo->mono transition; collapse stereo width */
Chris@69 137 width_Q14 = 0;
Chris@69 138 pred_Q13[ 0 ] = 0;
Chris@69 139 pred_Q13[ 1 ] = 0;
Chris@69 140 silk_stereo_quant_pred( pred_Q13, ix );
Chris@69 141 } else if( state->width_prev_Q14 == 0 &&
Chris@69 142 ( 8 * total_rate_bps < 13 * min_mid_rate_bps || silk_SMULWB( frac_Q16, state->smth_width_Q14 ) < SILK_FIX_CONST( 0.05, 14 ) ) )
Chris@69 143 {
Chris@69 144 /* Code as panned-mono; previous frame already had zero width */
Chris@69 145 /* Scale down and quantize predictors */
Chris@69 146 pred_Q13[ 0 ] = silk_RSHIFT( silk_SMULBB( state->smth_width_Q14, pred_Q13[ 0 ] ), 14 );
Chris@69 147 pred_Q13[ 1 ] = silk_RSHIFT( silk_SMULBB( state->smth_width_Q14, pred_Q13[ 1 ] ), 14 );
Chris@69 148 silk_stereo_quant_pred( pred_Q13, ix );
Chris@69 149 /* Collapse stereo width */
Chris@69 150 width_Q14 = 0;
Chris@69 151 pred_Q13[ 0 ] = 0;
Chris@69 152 pred_Q13[ 1 ] = 0;
Chris@69 153 mid_side_rates_bps[ 0 ] = total_rate_bps;
Chris@69 154 mid_side_rates_bps[ 1 ] = 0;
Chris@69 155 *mid_only_flag = 1;
Chris@69 156 } else if( state->width_prev_Q14 != 0 &&
Chris@69 157 ( 8 * total_rate_bps < 11 * min_mid_rate_bps || silk_SMULWB( frac_Q16, state->smth_width_Q14 ) < SILK_FIX_CONST( 0.02, 14 ) ) )
Chris@69 158 {
Chris@69 159 /* Transition to zero-width stereo */
Chris@69 160 /* Scale down and quantize predictors */
Chris@69 161 pred_Q13[ 0 ] = silk_RSHIFT( silk_SMULBB( state->smth_width_Q14, pred_Q13[ 0 ] ), 14 );
Chris@69 162 pred_Q13[ 1 ] = silk_RSHIFT( silk_SMULBB( state->smth_width_Q14, pred_Q13[ 1 ] ), 14 );
Chris@69 163 silk_stereo_quant_pred( pred_Q13, ix );
Chris@69 164 /* Collapse stereo width */
Chris@69 165 width_Q14 = 0;
Chris@69 166 pred_Q13[ 0 ] = 0;
Chris@69 167 pred_Q13[ 1 ] = 0;
Chris@69 168 } else if( state->smth_width_Q14 > SILK_FIX_CONST( 0.95, 14 ) ) {
Chris@69 169 /* Full-width stereo coding */
Chris@69 170 silk_stereo_quant_pred( pred_Q13, ix );
Chris@69 171 width_Q14 = SILK_FIX_CONST( 1, 14 );
Chris@69 172 } else {
Chris@69 173 /* Reduced-width stereo coding; scale down and quantize predictors */
Chris@69 174 pred_Q13[ 0 ] = silk_RSHIFT( silk_SMULBB( state->smth_width_Q14, pred_Q13[ 0 ] ), 14 );
Chris@69 175 pred_Q13[ 1 ] = silk_RSHIFT( silk_SMULBB( state->smth_width_Q14, pred_Q13[ 1 ] ), 14 );
Chris@69 176 silk_stereo_quant_pred( pred_Q13, ix );
Chris@69 177 width_Q14 = state->smth_width_Q14;
Chris@69 178 }
Chris@69 179
Chris@69 180 /* Make sure to keep on encoding until the tapered output has been transmitted */
Chris@69 181 if( *mid_only_flag == 1 ) {
Chris@69 182 state->silent_side_len += frame_length - STEREO_INTERP_LEN_MS * fs_kHz;
Chris@69 183 if( state->silent_side_len < LA_SHAPE_MS * fs_kHz ) {
Chris@69 184 *mid_only_flag = 0;
Chris@69 185 } else {
Chris@69 186 /* Limit to avoid wrapping around */
Chris@69 187 state->silent_side_len = 10000;
Chris@69 188 }
Chris@69 189 } else {
Chris@69 190 state->silent_side_len = 0;
Chris@69 191 }
Chris@69 192
Chris@69 193 if( *mid_only_flag == 0 && mid_side_rates_bps[ 1 ] < 1 ) {
Chris@69 194 mid_side_rates_bps[ 1 ] = 1;
Chris@69 195 mid_side_rates_bps[ 0 ] = silk_max_int( 1, total_rate_bps - mid_side_rates_bps[ 1 ]);
Chris@69 196 }
Chris@69 197
Chris@69 198 /* Interpolate predictors and subtract prediction from side channel */
Chris@69 199 pred0_Q13 = -state->pred_prev_Q13[ 0 ];
Chris@69 200 pred1_Q13 = -state->pred_prev_Q13[ 1 ];
Chris@69 201 w_Q24 = silk_LSHIFT( state->width_prev_Q14, 10 );
Chris@69 202 denom_Q16 = silk_DIV32_16( (opus_int32)1 << 16, STEREO_INTERP_LEN_MS * fs_kHz );
Chris@69 203 delta0_Q13 = -silk_RSHIFT_ROUND( silk_SMULBB( pred_Q13[ 0 ] - state->pred_prev_Q13[ 0 ], denom_Q16 ), 16 );
Chris@69 204 delta1_Q13 = -silk_RSHIFT_ROUND( silk_SMULBB( pred_Q13[ 1 ] - state->pred_prev_Q13[ 1 ], denom_Q16 ), 16 );
Chris@69 205 deltaw_Q24 = silk_LSHIFT( silk_SMULWB( width_Q14 - state->width_prev_Q14, denom_Q16 ), 10 );
Chris@69 206 for( n = 0; n < STEREO_INTERP_LEN_MS * fs_kHz; n++ ) {
Chris@69 207 pred0_Q13 += delta0_Q13;
Chris@69 208 pred1_Q13 += delta1_Q13;
Chris@69 209 w_Q24 += deltaw_Q24;
Chris@69 210 sum = silk_LSHIFT( silk_ADD_LSHIFT( mid[ n ] + (opus_int32)mid[ n + 2 ], mid[ n + 1 ], 1 ), 9 ); /* Q11 */
Chris@69 211 sum = silk_SMLAWB( silk_SMULWB( w_Q24, side[ n + 1 ] ), sum, pred0_Q13 ); /* Q8 */
Chris@69 212 sum = silk_SMLAWB( sum, silk_LSHIFT( (opus_int32)mid[ n + 1 ], 11 ), pred1_Q13 ); /* Q8 */
Chris@69 213 x2[ n - 1 ] = (opus_int16)silk_SAT16( silk_RSHIFT_ROUND( sum, 8 ) );
Chris@69 214 }
Chris@69 215
Chris@69 216 pred0_Q13 = -pred_Q13[ 0 ];
Chris@69 217 pred1_Q13 = -pred_Q13[ 1 ];
Chris@69 218 w_Q24 = silk_LSHIFT( width_Q14, 10 );
Chris@69 219 for( n = STEREO_INTERP_LEN_MS * fs_kHz; n < frame_length; n++ ) {
Chris@69 220 sum = silk_LSHIFT( silk_ADD_LSHIFT( mid[ n ] + (opus_int32)mid[ n + 2 ], mid[ n + 1 ], 1 ), 9 ); /* Q11 */
Chris@69 221 sum = silk_SMLAWB( silk_SMULWB( w_Q24, side[ n + 1 ] ), sum, pred0_Q13 ); /* Q8 */
Chris@69 222 sum = silk_SMLAWB( sum, silk_LSHIFT( (opus_int32)mid[ n + 1 ], 11 ), pred1_Q13 ); /* Q8 */
Chris@69 223 x2[ n - 1 ] = (opus_int16)silk_SAT16( silk_RSHIFT_ROUND( sum, 8 ) );
Chris@69 224 }
Chris@69 225 state->pred_prev_Q13[ 0 ] = (opus_int16)pred_Q13[ 0 ];
Chris@69 226 state->pred_prev_Q13[ 1 ] = (opus_int16)pred_Q13[ 1 ];
Chris@69 227 state->width_prev_Q14 = (opus_int16)width_Q14;
Chris@69 228 RESTORE_STACK;
Chris@69 229 }