michael@0: /***********************************************************************
michael@0: Copyright (c) 2006-2011, Skype Limited. All rights reserved.
michael@0: Redistribution and use in source and binary forms, with or without
michael@0: modification, are permitted provided that the following conditions
michael@0: are met:
michael@0: - Redistributions of source code must retain the above copyright notice,
michael@0: this list of conditions and the following disclaimer.
michael@0: - Redistributions in binary form must reproduce the above copyright
michael@0: notice, this list of conditions and the following disclaimer in the
michael@0: documentation and/or other materials provided with the distribution.
michael@0: - Neither the name of Internet Society, IETF or IETF Trust, nor the
michael@0: names of specific contributors, may be used to endorse or promote
michael@0: products derived from this software without specific prior written
michael@0: permission.
michael@0: THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
michael@0: AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
michael@0: IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
michael@0: ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE
michael@0: LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
michael@0: CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
michael@0: SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
michael@0: INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
michael@0: CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
michael@0: ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
michael@0: POSSIBILITY OF SUCH DAMAGE.
michael@0: ***********************************************************************/
michael@0: 
michael@0: #ifdef HAVE_CONFIG_H
michael@0: #include "config.h"
michael@0: #endif
michael@0: 
michael@0: #include "main.h"
michael@0: #include "stack_alloc.h"
michael@0: 
michael@0: /* Convert Left/Right stereo signal to adaptive Mid/Side representation */
michael@0: void silk_stereo_LR_to_MS(
michael@0:     stereo_enc_state            *state,                         /* I/O  State                                       */
michael@0:     opus_int16                  x1[],                           /* I/O  Left input signal, becomes mid signal       */
michael@0:     opus_int16                  x2[],                           /* I/O  Right input signal, becomes side signal     */
michael@0:     opus_int8                   ix[ 2 ][ 3 ],                   /* O    Quantization indices                        */
michael@0:     opus_int8                   *mid_only_flag,                 /* O    Flag: only mid signal coded                 */
michael@0:     opus_int32                  mid_side_rates_bps[],           /* O    Bitrates for mid and side signals           */
michael@0:     opus_int32                  total_rate_bps,                 /* I    Total bitrate                               */
michael@0:     opus_int                    prev_speech_act_Q8,             /* I    Speech activity level in previous frame     */
michael@0:     opus_int                    toMono,                         /* I    Last frame before a stereo->mono transition */
michael@0:     opus_int                    fs_kHz,                         /* I    Sample rate (kHz)                           */
michael@0:     opus_int                    frame_length                    /* I    Number of samples                           */
michael@0: )
michael@0: {
michael@0:     opus_int   n, is10msFrame, denom_Q16, delta0_Q13, delta1_Q13;
michael@0:     opus_int32 sum, diff, smooth_coef_Q16, pred_Q13[ 2 ], pred0_Q13, pred1_Q13;
michael@0:     opus_int32 LP_ratio_Q14, HP_ratio_Q14, frac_Q16, frac_3_Q16, min_mid_rate_bps, width_Q14, w_Q24, deltaw_Q24;
michael@0:     VARDECL( opus_int16, side );
michael@0:     VARDECL( opus_int16, LP_mid );
michael@0:     VARDECL( opus_int16, HP_mid );
michael@0:     VARDECL( opus_int16, LP_side );
michael@0:     VARDECL( opus_int16, HP_side );
michael@0:     opus_int16 *mid = &x1[ -2 ];
michael@0:     SAVE_STACK;
michael@0: 
michael@0:     ALLOC( side, frame_length + 2, opus_int16 );
michael@0:     /* Convert to basic mid/side signals */
michael@0:     for( n = 0; n < frame_length + 2; n++ ) {
michael@0:         sum  = x1[ n - 2 ] + (opus_int32)x2[ n - 2 ];
michael@0:         diff = x1[ n - 2 ] - (opus_int32)x2[ n - 2 ];
michael@0:         mid[  n ] = (opus_int16)silk_RSHIFT_ROUND( sum, 1 );
michael@0:         side[ n ] = (opus_int16)silk_SAT16( silk_RSHIFT_ROUND( diff, 1 ) );
michael@0:     }
michael@0: 
michael@0:     /* Buffering */
michael@0:     silk_memcpy( mid,  state->sMid,  2 * sizeof( opus_int16 ) );
michael@0:     silk_memcpy( side, state->sSide, 2 * sizeof( opus_int16 ) );
michael@0:     silk_memcpy( state->sMid,  &mid[  frame_length ], 2 * sizeof( opus_int16 ) );
michael@0:     silk_memcpy( state->sSide, &side[ frame_length ], 2 * sizeof( opus_int16 ) );
michael@0: 
michael@0:     /* LP and HP filter mid signal */
michael@0:     ALLOC( LP_mid, frame_length, opus_int16 );
michael@0:     ALLOC( HP_mid, frame_length, opus_int16 );
michael@0:     for( n = 0; n < frame_length; n++ ) {
michael@0:         sum = silk_RSHIFT_ROUND( silk_ADD_LSHIFT( mid[ n ] + mid[ n + 2 ], mid[ n + 1 ], 1 ), 2 );
michael@0:         LP_mid[ n ] = sum;
michael@0:         HP_mid[ n ] = mid[ n + 1 ] - sum;
michael@0:     }
michael@0: 
michael@0:     /* LP and HP filter side signal */
michael@0:     ALLOC( LP_side, frame_length, opus_int16 );
michael@0:     ALLOC( HP_side, frame_length, opus_int16 );
michael@0:     for( n = 0; n < frame_length; n++ ) {
michael@0:         sum = silk_RSHIFT_ROUND( silk_ADD_LSHIFT( side[ n ] + side[ n + 2 ], side[ n + 1 ], 1 ), 2 );
michael@0:         LP_side[ n ] = sum;
michael@0:         HP_side[ n ] = side[ n + 1 ] - sum;
michael@0:     }
michael@0: 
michael@0:     /* Find energies and predictors */
michael@0:     is10msFrame = frame_length == 10 * fs_kHz;
michael@0:     smooth_coef_Q16 = is10msFrame ?
michael@0:         SILK_FIX_CONST( STEREO_RATIO_SMOOTH_COEF / 2, 16 ) :
michael@0:         SILK_FIX_CONST( STEREO_RATIO_SMOOTH_COEF,     16 );
michael@0:     smooth_coef_Q16 = silk_SMULWB( silk_SMULBB( prev_speech_act_Q8, prev_speech_act_Q8 ), smooth_coef_Q16 );
michael@0: 
michael@0:     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 );
michael@0:     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 );
michael@0:     /* Ratio of the norms of residual and mid signals */
michael@0:     frac_Q16 = silk_SMLABB( HP_ratio_Q14, LP_ratio_Q14, 3 );
michael@0:     frac_Q16 = silk_min( frac_Q16, SILK_FIX_CONST( 1, 16 ) );
michael@0: 
michael@0:     /* Determine bitrate distribution between mid and side, and possibly reduce stereo width */
michael@0:     total_rate_bps -= is10msFrame ? 1200 : 600;      /* Subtract approximate bitrate for coding stereo parameters */
michael@0:     if( total_rate_bps < 1 ) {
michael@0:         total_rate_bps = 1;
michael@0:     }
michael@0:     min_mid_rate_bps = silk_SMLABB( 2000, fs_kHz, 900 );
michael@0:     silk_assert( min_mid_rate_bps < 32767 );
michael@0:     /* Default bitrate distribution: 8 parts for Mid and (5+3*frac) parts for Side. so: mid_rate = ( 8 / ( 13 + 3 * frac ) ) * total_ rate */
michael@0:     frac_3_Q16 = silk_MUL( 3, frac_Q16 );
michael@0:     mid_side_rates_bps[ 0 ] = silk_DIV32_varQ( total_rate_bps, SILK_FIX_CONST( 8 + 5, 16 ) + frac_3_Q16, 16+3 );
michael@0:     /* If Mid bitrate below minimum, reduce stereo width */
michael@0:     if( mid_side_rates_bps[ 0 ] < min_mid_rate_bps ) {
michael@0:         mid_side_rates_bps[ 0 ] = min_mid_rate_bps;
michael@0:         mid_side_rates_bps[ 1 ] = total_rate_bps - mid_side_rates_bps[ 0 ];
michael@0:         /* width = 4 * ( 2 * side_rate - min_rate ) / ( ( 1 + 3 * frac ) * min_rate ) */
michael@0:         width_Q14 = silk_DIV32_varQ( silk_LSHIFT( mid_side_rates_bps[ 1 ], 1 ) - min_mid_rate_bps,
michael@0:             silk_SMULWB( SILK_FIX_CONST( 1, 16 ) + frac_3_Q16, min_mid_rate_bps ), 14+2 );
michael@0:         width_Q14 = silk_LIMIT( width_Q14, 0, SILK_FIX_CONST( 1, 14 ) );
michael@0:     } else {
michael@0:         mid_side_rates_bps[ 1 ] = total_rate_bps - mid_side_rates_bps[ 0 ];
michael@0:         width_Q14 = SILK_FIX_CONST( 1, 14 );
michael@0:     }
michael@0: 
michael@0:     /* Smoother */
michael@0:     state->smth_width_Q14 = (opus_int16)silk_SMLAWB( state->smth_width_Q14, width_Q14 - state->smth_width_Q14, smooth_coef_Q16 );
michael@0: 
michael@0:     /* At very low bitrates or for inputs that are nearly amplitude panned, switch to panned-mono coding */
michael@0:     *mid_only_flag = 0;
michael@0:     if( toMono ) {
michael@0:         /* Last frame before stereo->mono transition; collapse stereo width */
michael@0:         width_Q14 = 0;
michael@0:         pred_Q13[ 0 ] = 0;
michael@0:         pred_Q13[ 1 ] = 0;
michael@0:         silk_stereo_quant_pred( pred_Q13, ix );
michael@0:     } else if( state->width_prev_Q14 == 0 &&
michael@0:         ( 8 * total_rate_bps < 13 * min_mid_rate_bps || silk_SMULWB( frac_Q16, state->smth_width_Q14 ) < SILK_FIX_CONST( 0.05, 14 ) ) )
michael@0:     {
michael@0:         /* Code as panned-mono; previous frame already had zero width */
michael@0:         /* Scale down and quantize predictors */
michael@0:         pred_Q13[ 0 ] = silk_RSHIFT( silk_SMULBB( state->smth_width_Q14, pred_Q13[ 0 ] ), 14 );
michael@0:         pred_Q13[ 1 ] = silk_RSHIFT( silk_SMULBB( state->smth_width_Q14, pred_Q13[ 1 ] ), 14 );
michael@0:         silk_stereo_quant_pred( pred_Q13, ix );
michael@0:         /* Collapse stereo width */
michael@0:         width_Q14 = 0;
michael@0:         pred_Q13[ 0 ] = 0;
michael@0:         pred_Q13[ 1 ] = 0;
michael@0:         mid_side_rates_bps[ 0 ] = total_rate_bps;
michael@0:         mid_side_rates_bps[ 1 ] = 0;
michael@0:         *mid_only_flag = 1;
michael@0:     } else if( state->width_prev_Q14 != 0 &&
michael@0:         ( 8 * total_rate_bps < 11 * min_mid_rate_bps || silk_SMULWB( frac_Q16, state->smth_width_Q14 ) < SILK_FIX_CONST( 0.02, 14 ) ) )
michael@0:     {
michael@0:         /* Transition to zero-width stereo */
michael@0:         /* Scale down and quantize predictors */
michael@0:         pred_Q13[ 0 ] = silk_RSHIFT( silk_SMULBB( state->smth_width_Q14, pred_Q13[ 0 ] ), 14 );
michael@0:         pred_Q13[ 1 ] = silk_RSHIFT( silk_SMULBB( state->smth_width_Q14, pred_Q13[ 1 ] ), 14 );
michael@0:         silk_stereo_quant_pred( pred_Q13, ix );
michael@0:         /* Collapse stereo width */
michael@0:         width_Q14 = 0;
michael@0:         pred_Q13[ 0 ] = 0;
michael@0:         pred_Q13[ 1 ] = 0;
michael@0:     } else if( state->smth_width_Q14 > SILK_FIX_CONST( 0.95, 14 ) ) {
michael@0:         /* Full-width stereo coding */
michael@0:         silk_stereo_quant_pred( pred_Q13, ix );
michael@0:         width_Q14 = SILK_FIX_CONST( 1, 14 );
michael@0:     } else {
michael@0:         /* Reduced-width stereo coding; scale down and quantize predictors */
michael@0:         pred_Q13[ 0 ] = silk_RSHIFT( silk_SMULBB( state->smth_width_Q14, pred_Q13[ 0 ] ), 14 );
michael@0:         pred_Q13[ 1 ] = silk_RSHIFT( silk_SMULBB( state->smth_width_Q14, pred_Q13[ 1 ] ), 14 );
michael@0:         silk_stereo_quant_pred( pred_Q13, ix );
michael@0:         width_Q14 = state->smth_width_Q14;
michael@0:     }
michael@0: 
michael@0:     /* Make sure to keep on encoding until the tapered output has been transmitted */
michael@0:     if( *mid_only_flag == 1 ) {
michael@0:         state->silent_side_len += frame_length - STEREO_INTERP_LEN_MS * fs_kHz;
michael@0:         if( state->silent_side_len < LA_SHAPE_MS * fs_kHz ) {
michael@0:             *mid_only_flag = 0;
michael@0:         } else {
michael@0:             /* Limit to avoid wrapping around */
michael@0:             state->silent_side_len = 10000;
michael@0:         }
michael@0:     } else {
michael@0:         state->silent_side_len = 0;
michael@0:     }
michael@0: 
michael@0:     if( *mid_only_flag == 0 && mid_side_rates_bps[ 1 ] < 1 ) {
michael@0:         mid_side_rates_bps[ 1 ] = 1;
michael@0:         mid_side_rates_bps[ 0 ] = silk_max_int( 1, total_rate_bps - mid_side_rates_bps[ 1 ]);
michael@0:     }
michael@0: 
michael@0:     /* Interpolate predictors and subtract prediction from side channel */
michael@0:     pred0_Q13  = -state->pred_prev_Q13[ 0 ];
michael@0:     pred1_Q13  = -state->pred_prev_Q13[ 1 ];
michael@0:     w_Q24      =  silk_LSHIFT( state->width_prev_Q14, 10 );
michael@0:     denom_Q16  = silk_DIV32_16( (opus_int32)1 << 16, STEREO_INTERP_LEN_MS * fs_kHz );
michael@0:     delta0_Q13 = -silk_RSHIFT_ROUND( silk_SMULBB( pred_Q13[ 0 ] - state->pred_prev_Q13[ 0 ], denom_Q16 ), 16 );
michael@0:     delta1_Q13 = -silk_RSHIFT_ROUND( silk_SMULBB( pred_Q13[ 1 ] - state->pred_prev_Q13[ 1 ], denom_Q16 ), 16 );
michael@0:     deltaw_Q24 =  silk_LSHIFT( silk_SMULWB( width_Q14 - state->width_prev_Q14, denom_Q16 ), 10 );
michael@0:     for( n = 0; n < STEREO_INTERP_LEN_MS * fs_kHz; n++ ) {
michael@0:         pred0_Q13 += delta0_Q13;
michael@0:         pred1_Q13 += delta1_Q13;
michael@0:         w_Q24   += deltaw_Q24;
michael@0:         sum = silk_LSHIFT( silk_ADD_LSHIFT( mid[ n ] + mid[ n + 2 ], mid[ n + 1 ], 1 ), 9 );    /* Q11 */
michael@0:         sum = silk_SMLAWB( silk_SMULWB( w_Q24, side[ n + 1 ] ), sum, pred0_Q13 );               /* Q8  */
michael@0:         sum = silk_SMLAWB( sum, silk_LSHIFT( (opus_int32)mid[ n + 1 ], 11 ), pred1_Q13 );       /* Q8  */
michael@0:         x2[ n - 1 ] = (opus_int16)silk_SAT16( silk_RSHIFT_ROUND( sum, 8 ) );
michael@0:     }
michael@0: 
michael@0:     pred0_Q13 = -pred_Q13[ 0 ];
michael@0:     pred1_Q13 = -pred_Q13[ 1 ];
michael@0:     w_Q24     =  silk_LSHIFT( width_Q14, 10 );
michael@0:     for( n = STEREO_INTERP_LEN_MS * fs_kHz; n < frame_length; n++ ) {
michael@0:         sum = silk_LSHIFT( silk_ADD_LSHIFT( mid[ n ] + mid[ n + 2 ], mid[ n + 1 ], 1 ), 9 );    /* Q11 */
michael@0:         sum = silk_SMLAWB( silk_SMULWB( w_Q24, side[ n + 1 ] ), sum, pred0_Q13 );               /* Q8  */
michael@0:         sum = silk_SMLAWB( sum, silk_LSHIFT( (opus_int32)mid[ n + 1 ], 11 ), pred1_Q13 );       /* Q8  */
michael@0:         x2[ n - 1 ] = (opus_int16)silk_SAT16( silk_RSHIFT_ROUND( sum, 8 ) );
michael@0:     }
michael@0:     state->pred_prev_Q13[ 0 ] = (opus_int16)pred_Q13[ 0 ];
michael@0:     state->pred_prev_Q13[ 1 ] = (opus_int16)pred_Q13[ 1 ];
michael@0:     state->width_prev_Q14     = (opus_int16)width_Q14;
michael@0:     RESTORE_STACK;
michael@0: }