vp3dsp.c 6.7 KB
Newer Older
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26
/*
 * Copyright (C) 2004 the ffmpeg project
 *
 * This library 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 of the License, or (at your option) any later version.
 *
 * This library 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 this library; if not, write to the Free Software
 * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
 */

/**
 * @file vp3dsp.c
 * Standard C DSP-oriented functions cribbed from the original VP3 
 * source code.
 */

#include "common.h"
#include "avcodec.h"
27
#include "dsputil.h"
28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43
#include "vp3data.h"

#define IdctAdjustBeforeShift 8
#define xC1S7 64277
#define xC2S6 60547
#define xC3S5 54491
#define xC4S4 46341
#define xC5S3 36410
#define xC6S2 25080
#define xC7S1 12785

void vp3_dsp_init_c(void)
{
    /* nop */
}

44 45
void vp3_idct_c(int16_t *input_data, int16_t *dequant_matrix,
    int coeff_count, int16_t *output_data)
46
{
47
    int32_t dequantized_data[64];
48 49 50 51 52 53 54
    int32_t *ip = dequantized_data;
    int16_t *op = output_data;

    int32_t A_, B_, C_, D_, _Ad, _Bd, _Cd, _Dd, E_, F_, G_, H_;
    int32_t _Ed, _Gd, _Add, _Bdd, _Fd, _Hd;
    int32_t t1, t2;

55 56 57 58 59 60 61
    int i, j;

    /* de-zigzag and dequantize */
    for (i = 0; i < coeff_count; i++) {
        j = dezigzag_index[i];
        dequantized_data[j] = dequant_matrix[i] * input_data[i];
    }
62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259

    /* Inverse DCT on the rows now */
    for (i = 0; i < 8; i++) {
        /* Check for non-zero values */
        if ( ip[0] | ip[1] | ip[2] | ip[3] | ip[4] | ip[5] | ip[6] | ip[7] ) {
            t1 = (int32_t)(xC1S7 * ip[1]);
            t2 = (int32_t)(xC7S1 * ip[7]);
            t1 >>= 16;
            t2 >>= 16;
            A_ = t1 + t2;

            t1 = (int32_t)(xC7S1 * ip[1]);
            t2 = (int32_t)(xC1S7 * ip[7]);
            t1 >>= 16;
            t2 >>= 16;
            B_ = t1 - t2;

            t1 = (int32_t)(xC3S5 * ip[3]);
            t2 = (int32_t)(xC5S3 * ip[5]);
            t1 >>= 16;
            t2 >>= 16;
            C_ = t1 + t2;

            t1 = (int32_t)(xC3S5 * ip[5]);
            t2 = (int32_t)(xC5S3 * ip[3]);
            t1 >>= 16;
            t2 >>= 16;
            D_ = t1 - t2;


            t1 = (int32_t)(xC4S4 * (A_ - C_));
            t1 >>= 16;
            _Ad = t1;

            t1 = (int32_t)(xC4S4 * (B_ - D_));
            t1 >>= 16;
            _Bd = t1;


            _Cd = A_ + C_;
            _Dd = B_ + D_;

            t1 = (int32_t)(xC4S4 * (ip[0] + ip[4]));
            t1 >>= 16;
            E_ = t1;

            t1 = (int32_t)(xC4S4 * (ip[0] - ip[4]));
            t1 >>= 16;
            F_ = t1;

            t1 = (int32_t)(xC2S6 * ip[2]);
            t2 = (int32_t)(xC6S2 * ip[6]);
            t1 >>= 16;
            t2 >>= 16;
            G_ = t1 + t2;

            t1 = (int32_t)(xC6S2 * ip[2]);
            t2 = (int32_t)(xC2S6 * ip[6]);
            t1 >>= 16;
            t2 >>= 16;
            H_ = t1 - t2;


            _Ed = E_ - G_;
            _Gd = E_ + G_;

            _Add = F_ + _Ad;
            _Bdd = _Bd - H_;

            _Fd = F_ - _Ad;
            _Hd = _Bd + H_;

            /*  Final sequence of operations over-write original inputs. */
            ip[0] = (int16_t)((_Gd + _Cd )   >> 0);
            ip[7] = (int16_t)((_Gd - _Cd )   >> 0);

            ip[1] = (int16_t)((_Add + _Hd )  >> 0);
            ip[2] = (int16_t)((_Add - _Hd )  >> 0);

            ip[3] = (int16_t)((_Ed + _Dd )   >> 0);
            ip[4] = (int16_t)((_Ed - _Dd )   >> 0);

            ip[5] = (int16_t)((_Fd + _Bdd )  >> 0);
            ip[6] = (int16_t)((_Fd - _Bdd )  >> 0);

        }

        ip += 8;            /* next row */
    }

    ip = dequantized_data;

    for ( i = 0; i < 8; i++) {
        /* Check for non-zero values (bitwise or faster than ||) */
        if ( ip[0 * 8] | ip[1 * 8] | ip[2 * 8] | ip[3 * 8] |
             ip[4 * 8] | ip[5 * 8] | ip[6 * 8] | ip[7 * 8] ) {

            t1 = (int32_t)(xC1S7 * ip[1*8]);
            t2 = (int32_t)(xC7S1 * ip[7*8]);
            t1 >>= 16;
            t2 >>= 16;
            A_ = t1 + t2;

            t1 = (int32_t)(xC7S1 * ip[1*8]);
            t2 = (int32_t)(xC1S7 * ip[7*8]);
            t1 >>= 16;
            t2 >>= 16;
            B_ = t1 - t2;

            t1 = (int32_t)(xC3S5 * ip[3*8]);
            t2 = (int32_t)(xC5S3 * ip[5*8]);
            t1 >>= 16;
            t2 >>= 16;
            C_ = t1 + t2;

            t1 = (int32_t)(xC3S5 * ip[5*8]);
            t2 = (int32_t)(xC5S3 * ip[3*8]);
            t1 >>= 16;
            t2 >>= 16;
            D_ = t1 - t2;


            t1 = (int32_t)(xC4S4 * (A_ - C_));
            t1 >>= 16;
            _Ad = t1;

            t1 = (int32_t)(xC4S4 * (B_ - D_));
            t1 >>= 16;
            _Bd = t1;


            _Cd = A_ + C_;
            _Dd = B_ + D_;

            t1 = (int32_t)(xC4S4 * (ip[0*8] + ip[4*8]));
            t1 >>= 16;
            E_ = t1;

            t1 = (int32_t)(xC4S4 * (ip[0*8] - ip[4*8]));
            t1 >>= 16;
            F_ = t1;

            t1 = (int32_t)(xC2S6 * ip[2*8]);
            t2 = (int32_t)(xC6S2 * ip[6*8]);
            t1 >>= 16;
            t2 >>= 16;
            G_ = t1 + t2;

            t1 = (int32_t)(xC6S2 * ip[2*8]);
            t2 = (int32_t)(xC2S6 * ip[6*8]);
            t1 >>= 16;
            t2 >>= 16;
            H_ = t1 - t2;


            _Ed = E_ - G_;
            _Gd = E_ + G_;

            _Add = F_ + _Ad;
            _Bdd = _Bd - H_;

            _Fd = F_ - _Ad;
            _Hd = _Bd + H_;

            _Gd += IdctAdjustBeforeShift;
            _Add += IdctAdjustBeforeShift;
            _Ed += IdctAdjustBeforeShift;
            _Fd += IdctAdjustBeforeShift;

            /* Final sequence of operations over-write original inputs. */
            op[0*8] = (int16_t)((_Gd + _Cd )   >> 4);
            op[7*8] = (int16_t)((_Gd - _Cd )   >> 4);

            op[1*8] = (int16_t)((_Add + _Hd )  >> 4);
            op[2*8] = (int16_t)((_Add - _Hd )  >> 4);

            op[3*8] = (int16_t)((_Ed + _Dd )   >> 4);
            op[4*8] = (int16_t)((_Ed - _Dd )   >> 4);

            op[5*8] = (int16_t)((_Fd + _Bdd )  >> 4);
            op[6*8] = (int16_t)((_Fd - _Bdd )  >> 4);

        } else {

            op[0*8] = 0;
            op[7*8] = 0;
            op[1*8] = 0;
            op[2*8] = 0;
            op[3*8] = 0;
            op[4*8] = 0;
            op[5*8] = 0;
            op[6*8] = 0;
        }

        ip++;            /* next column */
        op++;
    }
}