annotate sipr16k.c @ 12454:f4355cd85faa libavcodec

Port latest x264 deblock asm (before they moved to using NV12 as internal format), LGPL'ed with permission from Jason and Loren. This includes mmx2 code, so remove inline asm from h264dsp_mmx.c accordingly.

/*
 * SIPR decoder for the 16k mode
 *
 * Copyright (c) 2008 Vladimir Voroshilov
 * Copyright (c) 2009 Vitor Sessak
 *
 * This file is part of FFmpeg.
 *
 * FFmpeg is free software; you can redistribute it and/or
 * modify it under the terms of the GNU Lesser General Public
 * License as published by the Free Software Foundation; either
 * version 2.1 of the License, or (at your option) any later version.
 *
 * FFmpeg is distributed in the hope that it will be useful,
 * but WITHOUT ANY WARRANTY; without even the implied warranty of
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
 * Lesser General Public License for more details.
 *
 * You should have received a copy of the GNU Lesser General Public
 * License along with FFmpeg; if not, write to the Free Software
 * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
 */

#include <math.h>

#include "sipr.h"
#include "libavutil/mathematics.h"
#include "lsp.h"
#include "celp_math.h"
#include "acelp_vectors.h"
#include "acelp_pitch_delay.h"
#include "acelp_filters.h"
#include "celp_filters.h"

#include "sipr16kdata.h"

/**
 * Convert an lsf vector into an lsp vector.
 *
 * @param lsf               input lsf vector
 * @param lsp               output lsp vector
 */
static void lsf2lsp(const float *lsf, double *lsp)
{
    int i;

    for (i = 0; i < LP_FILTER_ORDER_16k; i++)
        lsp[i] = cosf(lsf[i]);
}

static void dequant(float *out, const int *idx, const float *cbs[])
{
    int i;

    for (i = 0; i < 4; i++)
        memcpy(out + 3*i, cbs[i] + 3*idx[i], 3*sizeof(float));

    memcpy(out + 12, cbs[4] + 4*idx[4], 4*sizeof(float));
}

static void lsf_decode_fp_16k(float* lsf_history, float* isp_new,
                              const int* parm, int ma_pred)
{
    int i;
    float isp_q[LP_FILTER_ORDER_16k];

    dequant(isp_q, parm, lsf_codebooks_16k);

    for (i = 0; i < LP_FILTER_ORDER_16k; i++) {
        isp_new[i] = (1 - qu[ma_pred]) * isp_q[i]
                    +     qu[ma_pred]  * lsf_history[i]
                    + mean_lsf_16k[i];
    }

    memcpy(lsf_history, isp_q, LP_FILTER_ORDER_16k * sizeof(float));
}

static int dec_delay3_1st(int index)
{
    if (index < 390) {
        return index + 88;
    } else
        return 3 * index - 690;
}

static int dec_delay3_2nd(int index, int pit_min, int pit_max,
                          int pitch_lag_prev)
{
    if (index < 62) {
        int pitch_delay_min = av_clip(pitch_lag_prev - 10,
                                      pit_min, pit_max - 19);
        return 3 * pitch_delay_min + index - 2;
    } else
        return 3 * pitch_lag_prev;
}

static void postfilter(float *out_data, float* synth, float* iir_mem,
                       float* filt_mem[2], float* mem_preemph)
{
    float buf[30 + LP_FILTER_ORDER_16k];
    float *tmpbuf = buf + LP_FILTER_ORDER_16k;
    float s;
    int i;

    for (i = 0; i < LP_FILTER_ORDER_16k; i++)
        filt_mem[0][i] = iir_mem[i] * ff_pow_0_5[i];

    memcpy(tmpbuf - LP_FILTER_ORDER_16k, mem_preemph,
           LP_FILTER_ORDER_16k*sizeof(*buf));

    ff_celp_lp_synthesis_filterf(tmpbuf, filt_mem[1], synth, 30,
                                 LP_FILTER_ORDER_16k);

    memcpy(synth - LP_FILTER_ORDER_16k, mem_preemph,
           LP_FILTER_ORDER_16k * sizeof(*synth));

    ff_celp_lp_synthesis_filterf(synth, filt_mem[0], synth, 30,
                                 LP_FILTER_ORDER_16k);

    memcpy(out_data + 30 - LP_FILTER_ORDER_16k,
           synth    + 30 - LP_FILTER_ORDER_16k,
           LP_FILTER_ORDER_16k * sizeof(*synth));

    ff_celp_lp_synthesis_filterf(out_data + 30, filt_mem[0],
                                 synth + 30, 2 * L_SUBFR_16k - 30,
                                 LP_FILTER_ORDER_16k);


    memcpy(mem_preemph, out_data + 2*L_SUBFR_16k - LP_FILTER_ORDER_16k,
           LP_FILTER_ORDER_16k * sizeof(*synth));

    FFSWAP(float *, filt_mem[0], filt_mem[1]);
    for (i = 0, s = 0; i < 30; i++, s += 1.0/30)
        out_data[i] = tmpbuf[i] + s * (synth[i] - tmpbuf[i]);
}

/**
 * Floating point version of ff_acelp_lp_decode().
 */
static void acelp_lp_decodef(float *lp_1st, float *lp_2nd,
                             const double *lsp_2nd, const double *lsp_prev)
{
    double lsp_1st[LP_FILTER_ORDER_16k];
    int i;

    /* LSP values for first subframe (3.2.5 of G.729, Equation 24) */
    for (i = 0; i < LP_FILTER_ORDER_16k; i++)
        lsp_1st[i] = (lsp_2nd[i] + lsp_prev[i]) * 0.5;

    ff_acelp_lspd2lpc(lsp_1st, lp_1st, LP_FILTER_ORDER_16k >> 1);

    /* LSP values for second subframe (3.2.5 of G.729) */
    ff_acelp_lspd2lpc(lsp_2nd, lp_2nd, LP_FILTER_ORDER_16k >> 1);
}

/**
 * Floating point version of ff_acelp_decode_gain_code().
 */
static float acelp_decode_gain_codef(float gain_corr_factor, const float *fc_v,
                                     float mr_energy, const float *quant_energy,
                                     const float *ma_prediction_coeff,
                                     int subframe_size, int ma_pred_order)
{
    mr_energy +=
        ff_dot_productf(quant_energy, ma_prediction_coeff, ma_pred_order);

    mr_energy = gain_corr_factor * exp(M_LN10 / 20. * mr_energy) /
        sqrt((0.01 + ff_dot_productf(fc_v, fc_v, subframe_size)));
    return mr_energy;
}

#define DIVIDE_BY_3(x) ((x) * 10923 >> 15)

void ff_sipr_decode_frame_16k(SiprContext *ctx, SiprParameters *params,
                              float *out_data)
{
    int frame_size = SUBFRAME_COUNT_16k * L_SUBFR_16k;
    float *synth = ctx->synth_buf + LP_FILTER_ORDER_16k;
    float lsf_new[LP_FILTER_ORDER_16k];
    double lsp_new[LP_FILTER_ORDER_16k];
    float Az[2][LP_FILTER_ORDER_16k];
    float fixed_vector[L_SUBFR_16k];
    float pitch_fac, gain_code;

    int i;
    int pitch_delay_3x;

    float *excitation = ctx->excitation + 292;

    lsf_decode_fp_16k(ctx->lsf_history, lsf_new, params->vq_indexes,
                      params->ma_pred_switch);

    ff_set_min_dist_lsf(lsf_new, LSFQ_DIFF_MIN / 2, LP_FILTER_ORDER_16k);

    lsf2lsp(lsf_new, lsp_new);

    acelp_lp_decodef(Az[0], Az[1], lsp_new, ctx->lsp_history_16k);

    memcpy(ctx->lsp_history_16k, lsp_new, LP_FILTER_ORDER_16k * sizeof(double));

    memcpy(synth - LP_FILTER_ORDER_16k, ctx->synth,
           LP_FILTER_ORDER_16k * sizeof(*synth));

    for (i = 0; i < SUBFRAME_COUNT_16k; i++) {
        int i_subfr = i * L_SUBFR_16k;
        AMRFixed f;
        float gain_corr_factor;
        int pitch_delay_int;
        int pitch_delay_frac;

        if (!i) {
            pitch_delay_3x = dec_delay3_1st(params->pitch_delay[i]);
        } else
            pitch_delay_3x = dec_delay3_2nd(params->pitch_delay[i],
                                            PITCH_MIN, PITCH_MAX,
                                            ctx->pitch_lag_prev);

        pitch_fac = gain_pitch_cb_16k[params->gp_index[i]];
        f.pitch_fac = FFMIN(pitch_fac, 1.0);
        f.pitch_lag = DIVIDE_BY_3(pitch_delay_3x+1);
        ctx->pitch_lag_prev = f.pitch_lag;

        pitch_delay_int  = DIVIDE_BY_3(pitch_delay_3x + 2);
        pitch_delay_frac = pitch_delay_3x + 2 - 3*pitch_delay_int;

        ff_acelp_interpolatef(&excitation[i_subfr],
                              &excitation[i_subfr] - pitch_delay_int + 1,
                              sinc_win, 3, pitch_delay_frac + 1,
                              LP_FILTER_ORDER, L_SUBFR_16k);


        memset(fixed_vector, 0, sizeof(fixed_vector));

        ff_decode_10_pulses_35bits(params->fc_indexes[i], &f,
                                   ff_fc_4pulses_8bits_tracks_13, 5, 4);

        ff_set_fixed_vector(fixed_vector, &f, 1.0, L_SUBFR_16k);

        gain_corr_factor = gain_cb_16k[params->gc_index[i]];
        gain_code = gain_corr_factor *
            acelp_decode_gain_codef(sqrt(L_SUBFR_16k), fixed_vector,
                                    19.0 - 15.0/(0.05*M_LN10/M_LN2),
                                    pred_16k, ctx->energy_history,
                                    L_SUBFR_16k, 2);

        ctx->energy_history[1] = ctx->energy_history[0];
        ctx->energy_history[0] = 20.0 * log10f(gain_corr_factor);

        ff_weighted_vector_sumf(&excitation[i_subfr], &excitation[i_subfr],
                                fixed_vector, pitch_fac,
                                gain_code, L_SUBFR_16k);

        ff_celp_lp_synthesis_filterf(synth + i_subfr, Az[i],
                                     &excitation[i_subfr], L_SUBFR_16k,
                                     LP_FILTER_ORDER_16k);

    }
    memcpy(ctx->synth, synth + frame_size - LP_FILTER_ORDER_16k,
           LP_FILTER_ORDER_16k * sizeof(*synth));

    memmove(ctx->excitation, ctx->excitation + 2 * L_SUBFR_16k,
            (L_INTERPOL+PITCH_MAX) * sizeof(float));

    postfilter(out_data, synth, ctx->iir_mem, ctx->filt_mem, ctx->mem_preemph);

    memcpy(ctx->iir_mem, Az[1], LP_FILTER_ORDER_16k * sizeof(float));
}

void ff_sipr_init_16k(SiprContext *ctx)
{
    int i;

    for (i = 0; i < LP_FILTER_ORDER_16k; i++)
        ctx->lsp_history_16k[i] = cos((i + 1) * M_PI/(LP_FILTER_ORDER_16k + 1));

    ctx->filt_mem[0] = ctx->filt_buf[0];
    ctx->filt_mem[1] = ctx->filt_buf[1];

    ctx->pitch_lag_prev = 180;
}