| 1 | /* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * |
| 2 | * Mupen64plus - TextureFilters_hq4x.cpp * |
| 3 | * Mupen64Plus homepage: http://code.google.com/p/mupen64plus/ * |
| 4 | * Copyright (C) 2003 MaxSt ( maxst@hiend3d.com ) * |
| 5 | * * |
| 6 | * This program is free software; you can redistribute it and/or modify * |
| 7 | * it under the terms of the GNU General Public License as published by * |
| 8 | * the Free Software Foundation; either version 2 of the License, or * |
| 9 | * (at your option) any later version. * |
| 10 | * * |
| 11 | * This program 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 * |
| 14 | * GNU General Public License for more details. * |
| 15 | * * |
| 16 | * You should have received a copy of the GNU General Public License * |
| 17 | * along with this program; if not, write to the * |
| 18 | * Free Software Foundation, Inc., * |
| 19 | * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. * |
| 20 | * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * */ |
| 21 | |
| 22 | #include <stdlib.h> |
| 23 | |
| 24 | #include "typedefs.h" |
| 25 | |
| 26 | static int RGBtoYUV[4096]; |
| 27 | //#define RGB32toYUV(val) (RGBtoYUV[((val&0x00FF0000)>>20)+((val&0x0000FF00)>>12)+((val&0x000000FF)>>4)]) |
| 28 | inline int RGB32toYUV(uint32 val) |
| 29 | { |
| 30 | int a,r,g,b,Y,u,v; |
| 31 | //r = (val&0x00FF0000)>>16; |
| 32 | //g = (val&0x0000FF00)>>8; |
| 33 | //b = (val&0x000000FF); |
| 34 | a = (val&0xFF000000); |
| 35 | r = (val&0x00FF0000)>>16; |
| 36 | g = (val&0x0000FF00)>>8; |
| 37 | b = (val&0x000000FF); |
| 38 | //r = (val&0x00F80000)>>16; |
| 39 | //g = (val&0x0000FC00)>>8; |
| 40 | //b = (val&0x000000F8); |
| 41 | Y = (r + g + b) >> 2; |
| 42 | u = 128 + ((r - b) >> 2); |
| 43 | v = 128 + ((-r + 2*g -b)>>3); |
| 44 | return a + (Y<<16) + (u<<8) + v; |
| 45 | } |
| 46 | #define RGB16toYUV(val) (RGBtoYUV[(val&0x0FFF)]) |
| 47 | static int YUV1, YUV2; |
| 48 | const int Amask = 0xFF000000; |
| 49 | const int Ymask = 0x00FF0000; |
| 50 | const int Umask = 0x0000FF00; |
| 51 | const int Vmask = 0x000000FF; |
| 52 | const int trA = 0x20000000; |
| 53 | const int trY = 0x00300000; |
| 54 | const int trU = 0x00000700; |
| 55 | const int trV = 0x00000006; |
| 56 | //const int trU = 0x00001800; |
| 57 | //const int trV = 0x00000018; |
| 58 | |
| 59 | #define INTERP_16_MASK_1_3(v) ((v)&0x0F0F) |
| 60 | #define INTERP_16_MASK_SHIFT_2_4(v) (((v)&0xF0F0)>>4) |
| 61 | #define INTERP_16_MASK_SHIFTBACK_2_4(v) ((INTERP_16_MASK_1_3(v))<<4) |
| 62 | |
| 63 | inline void hq4x_Interp1_16(unsigned char * pc, uint16 p1, uint16 p2) |
| 64 | { |
| 65 | *((uint16*)pc) = INTERP_16_MASK_1_3((INTERP_16_MASK_1_3(p1)*3 + INTERP_16_MASK_1_3(p2)) / 4) |
| 66 | | INTERP_16_MASK_SHIFTBACK_2_4((INTERP_16_MASK_SHIFT_2_4(p1)*3 + INTERP_16_MASK_SHIFT_2_4(p2)) / 4 ); |
| 67 | } |
| 68 | |
| 69 | inline void hq4x_Interp2_16(unsigned char * pc, uint16 p1, uint16 p2, uint16 p3) |
| 70 | { |
| 71 | *((uint16*)pc) = INTERP_16_MASK_1_3((INTERP_16_MASK_1_3(p1)*2 + INTERP_16_MASK_1_3(p2) + INTERP_16_MASK_1_3(p3)) / 4) |
| 72 | | INTERP_16_MASK_SHIFTBACK_2_4((INTERP_16_MASK_SHIFT_2_4(p1)*2 + INTERP_16_MASK_SHIFT_2_4(p2) + INTERP_16_MASK_SHIFT_2_4(p3)) / 4); |
| 73 | } |
| 74 | |
| 75 | inline void hq4x_Interp3_16(unsigned char * pc, uint16 p1, uint16 p2) |
| 76 | { |
| 77 | *((uint16*)pc) = INTERP_16_MASK_1_3((INTERP_16_MASK_1_3(p1)*7 + INTERP_16_MASK_1_3(p2)) / 8) |
| 78 | | INTERP_16_MASK_SHIFTBACK_2_4((INTERP_16_MASK_SHIFT_2_4(p1)*7 + INTERP_16_MASK_SHIFT_2_4(p2)) / 8); |
| 79 | } |
| 80 | |
| 81 | inline void hq4x_Interp5_16(unsigned char * pc, uint16 p1, uint16 p2) |
| 82 | { |
| 83 | *((uint16*)pc) = INTERP_16_MASK_1_3((INTERP_16_MASK_1_3(p1) + INTERP_16_MASK_1_3(p2)) / 2) |
| 84 | | INTERP_16_MASK_SHIFTBACK_2_4((INTERP_16_MASK_SHIFT_2_4(p1) + INTERP_16_MASK_SHIFT_2_4(p2)) / 2); |
| 85 | } |
| 86 | |
| 87 | inline void hq4x_Interp6_16(unsigned char * pc, uint16 p1, uint16 p2, uint16 p3) |
| 88 | { |
| 89 | *((uint16*)pc) = INTERP_16_MASK_1_3((INTERP_16_MASK_1_3(p1)*5 + INTERP_16_MASK_1_3(p2)*2 + INTERP_16_MASK_1_3(p3)*1) / 8) |
| 90 | | INTERP_16_MASK_SHIFTBACK_2_4((INTERP_16_MASK_SHIFT_2_4(p1)*5 + INTERP_16_MASK_SHIFT_2_4(p2)*2 + INTERP_16_MASK_SHIFT_2_4(p3)*1) / 8); |
| 91 | } |
| 92 | |
| 93 | inline void hq4x_Interp7_16(unsigned char * pc, uint16 p1, uint16 p2, uint16 p3) |
| 94 | { |
| 95 | *((uint16*)pc) = INTERP_16_MASK_1_3((INTERP_16_MASK_1_3(p1)*6 + INTERP_16_MASK_1_3(p2) + INTERP_16_MASK_1_3(p3)) / 8) |
| 96 | | INTERP_16_MASK_SHIFTBACK_2_4((INTERP_16_MASK_SHIFT_2_4(p1)*6 + INTERP_16_MASK_SHIFT_2_4(p2) + INTERP_16_MASK_SHIFT_2_4(p3)) / 8); |
| 97 | } |
| 98 | |
| 99 | inline void hq4x_Interp8_16(unsigned char * pc, uint16 p1, uint16 p2) |
| 100 | { |
| 101 | //*((int*)pc) = (c1*5+c2*3)/8; |
| 102 | *((uint16*)pc) = INTERP_16_MASK_1_3((INTERP_16_MASK_1_3(p1)*5 + INTERP_16_MASK_1_3(p2)*3) / 8) |
| 103 | | INTERP_16_MASK_SHIFTBACK_2_4((INTERP_16_MASK_SHIFT_2_4(p1)*5 + INTERP_16_MASK_SHIFT_2_4(p2)*3) / 8); |
| 104 | } |
| 105 | |
| 106 | #define INTERP_32_MASK_1_3(v) ((v)&0x00FF00FF) |
| 107 | #define INTERP_32_MASK_SHIFT_2_4(v) (((v)&0xFF00FF00)>>8) |
| 108 | #define INTERP_32_MASK_SHIFTBACK_2_4(v) (((INTERP_32_MASK_1_3(v))<<8)) |
| 109 | |
| 110 | inline void hq4x_Interp1_32(unsigned char * pc, uint32 p1, uint32 p2) |
| 111 | { |
| 112 | *((uint32*)pc) = INTERP_32_MASK_1_3((INTERP_32_MASK_1_3(p1)*3 + INTERP_32_MASK_1_3(p2)) / 4) |
| 113 | | INTERP_32_MASK_SHIFTBACK_2_4((INTERP_32_MASK_SHIFT_2_4(p1)*3 + INTERP_32_MASK_SHIFT_2_4(p2)) / 4 ); |
| 114 | } |
| 115 | |
| 116 | inline void hq4x_Interp2_32(unsigned char * pc, uint32 p1, uint32 p2, uint32 p3) |
| 117 | { |
| 118 | *((uint32*)pc) = INTERP_32_MASK_1_3((INTERP_32_MASK_1_3(p1)*2 + INTERP_32_MASK_1_3(p2) + INTERP_32_MASK_1_3(p3)) / 4) |
| 119 | | INTERP_32_MASK_SHIFTBACK_2_4((INTERP_32_MASK_SHIFT_2_4(p1)*2 + INTERP_32_MASK_SHIFT_2_4(p2) + INTERP_32_MASK_SHIFT_2_4(p3)) / 4); |
| 120 | } |
| 121 | |
| 122 | inline void hq4x_Interp3_32(unsigned char * pc, uint32 p1, uint32 p2) |
| 123 | { |
| 124 | *((uint32*)pc) = INTERP_32_MASK_1_3((INTERP_32_MASK_1_3(p1)*7 + INTERP_32_MASK_1_3(p2)) / 8) |
| 125 | | INTERP_32_MASK_SHIFTBACK_2_4((INTERP_32_MASK_SHIFT_2_4(p1)*7 + INTERP_32_MASK_SHIFT_2_4(p2)) / 8); |
| 126 | } |
| 127 | |
| 128 | inline void hq4x_Interp5_32(unsigned char * pc, uint32 p1, uint32 p2) |
| 129 | { |
| 130 | *((uint32*)pc) = INTERP_32_MASK_1_3((INTERP_32_MASK_1_3(p1) + INTERP_32_MASK_1_3(p2)) / 2) |
| 131 | | INTERP_32_MASK_SHIFTBACK_2_4((INTERP_32_MASK_SHIFT_2_4(p1) + INTERP_32_MASK_SHIFT_2_4(p2)) / 2); |
| 132 | } |
| 133 | |
| 134 | inline void hq4x_Interp6_32(unsigned char * pc, uint32 p1, uint32 p2, uint32 p3) |
| 135 | { |
| 136 | *((uint32*)pc) = INTERP_32_MASK_1_3((INTERP_32_MASK_1_3(p1)*5 + INTERP_32_MASK_1_3(p2)*2 + INTERP_32_MASK_1_3(p3)*1) / 8) |
| 137 | | INTERP_32_MASK_SHIFTBACK_2_4((INTERP_32_MASK_SHIFT_2_4(p1)*5 + INTERP_32_MASK_SHIFT_2_4(p2)*2 + INTERP_32_MASK_SHIFT_2_4(p3)*1) / 8); |
| 138 | } |
| 139 | |
| 140 | inline void hq4x_Interp7_32(unsigned char * pc, uint32 p1, uint32 p2, uint32 p3) |
| 141 | { |
| 142 | *((uint32*)pc) = INTERP_32_MASK_1_3((INTERP_32_MASK_1_3(p1)*6 + INTERP_32_MASK_1_3(p2) + INTERP_32_MASK_1_3(p3)) / 8) |
| 143 | | INTERP_32_MASK_SHIFTBACK_2_4((INTERP_32_MASK_SHIFT_2_4(p1)*6 + INTERP_32_MASK_SHIFT_2_4(p2) + INTERP_32_MASK_SHIFT_2_4(p3)) / 8); |
| 144 | } |
| 145 | |
| 146 | inline void hq4x_Interp8_32(unsigned char * pc, uint32 p1, uint32 p2) |
| 147 | { |
| 148 | //*((int*)pc) = (c1*5+c2*3)/8; |
| 149 | *((uint32*)pc) = INTERP_32_MASK_1_3((INTERP_32_MASK_1_3(p1)*5 + INTERP_32_MASK_1_3(p2)*3) / 8) |
| 150 | | INTERP_32_MASK_SHIFTBACK_2_4((INTERP_32_MASK_SHIFT_2_4(p1)*5 + INTERP_32_MASK_SHIFT_2_4(p2)*3) / 8); |
| 151 | } |
| 152 | |
| 153 | #define PIXEL00_0 *((int*)(pOut)) = c[5]; |
| 154 | #define PIXEL00_11 hq4x_Interp1(pOut, c[5], c[4]); |
| 155 | #define PIXEL00_12 hq4x_Interp1(pOut, c[5], c[2]); |
| 156 | #define PIXEL00_20 hq4x_Interp2(pOut, c[5], c[2], c[4]); |
| 157 | #define PIXEL00_50 hq4x_Interp5(pOut, c[2], c[4]); |
| 158 | #define PIXEL00_80 hq4x_Interp8(pOut, c[5], c[1]); |
| 159 | #define PIXEL00_81 hq4x_Interp8(pOut, c[5], c[4]); |
| 160 | #define PIXEL00_82 hq4x_Interp8(pOut, c[5], c[2]); |
| 161 | #define PIXEL01_0 *((int*)(pOut+BPP)) = c[5]; |
| 162 | #define PIXEL01_10 hq4x_Interp1(pOut+BPP, c[5], c[1]); |
| 163 | #define PIXEL01_12 hq4x_Interp1(pOut+BPP, c[5], c[2]); |
| 164 | #define PIXEL01_14 hq4x_Interp1(pOut+BPP, c[2], c[5]); |
| 165 | #define PIXEL01_21 hq4x_Interp2(pOut+BPP, c[2], c[5], c[4]); |
| 166 | #define PIXEL01_31 hq4x_Interp3(pOut+BPP, c[5], c[4]); |
| 167 | #define PIXEL01_50 hq4x_Interp5(pOut+BPP, c[2], c[5]); |
| 168 | #define PIXEL01_60 hq4x_Interp6(pOut+BPP, c[5], c[2], c[4]); |
| 169 | #define PIXEL01_61 hq4x_Interp6(pOut+BPP, c[5], c[2], c[1]); |
| 170 | #define PIXEL01_82 hq4x_Interp8(pOut+BPP, c[5], c[2]); |
| 171 | #define PIXEL01_83 hq4x_Interp8(pOut+BPP, c[2], c[4]); |
| 172 | #define PIXEL02_0 *((int*)(pOut+BPP2)) = c[5]; |
| 173 | #define PIXEL02_10 hq4x_Interp1(pOut+BPP2, c[5], c[3]); |
| 174 | #define PIXEL02_11 hq4x_Interp1(pOut+BPP2, c[5], c[2]); |
| 175 | #define PIXEL02_13 hq4x_Interp1(pOut+BPP2, c[2], c[5]); |
| 176 | #define PIXEL02_21 hq4x_Interp2(pOut+BPP2, c[2], c[5], c[6]); |
| 177 | #define PIXEL02_32 hq4x_Interp3(pOut+BPP2, c[5], c[6]); |
| 178 | #define PIXEL02_50 hq4x_Interp5(pOut+BPP2, c[2], c[5]); |
| 179 | #define PIXEL02_60 hq4x_Interp6(pOut+BPP2, c[5], c[2], c[6]); |
| 180 | #define PIXEL02_61 hq4x_Interp6(pOut+BPP2, c[5], c[2], c[3]); |
| 181 | #define PIXEL02_81 hq4x_Interp8(pOut+BPP2, c[5], c[2]); |
| 182 | #define PIXEL02_83 hq4x_Interp8(pOut+BPP2, c[2], c[6]); |
| 183 | #define PIXEL03_0 *((int*)(pOut+BPP3)) = c[5]; |
| 184 | #define PIXEL03_11 hq4x_Interp1(pOut+BPP3, c[5], c[2]); |
| 185 | #define PIXEL03_12 hq4x_Interp1(pOut+BPP3, c[5], c[6]); |
| 186 | #define PIXEL03_20 hq4x_Interp2(pOut+BPP3, c[5], c[2], c[6]); |
| 187 | #define PIXEL03_50 hq4x_Interp5(pOut+BPP3, c[2], c[6]); |
| 188 | #define PIXEL03_80 hq4x_Interp8(pOut+BPP3, c[5], c[3]); |
| 189 | #define PIXEL03_81 hq4x_Interp8(pOut+BPP3, c[5], c[2]); |
| 190 | #define PIXEL03_82 hq4x_Interp8(pOut+BPP3, c[5], c[6]); |
| 191 | #define PIXEL10_0 *((int*)(pOut+BpL)) = c[5]; |
| 192 | #define PIXEL10_10 hq4x_Interp1(pOut+BpL, c[5], c[1]); |
| 193 | #define PIXEL10_11 hq4x_Interp1(pOut+BpL, c[5], c[4]); |
| 194 | #define PIXEL10_13 hq4x_Interp1(pOut+BpL, c[4], c[5]); |
| 195 | #define PIXEL10_21 hq4x_Interp2(pOut+BpL, c[4], c[5], c[2]); |
| 196 | #define PIXEL10_32 hq4x_Interp3(pOut+BpL, c[5], c[2]); |
| 197 | #define PIXEL10_50 hq4x_Interp5(pOut+BpL, c[4], c[5]); |
| 198 | #define PIXEL10_60 hq4x_Interp6(pOut+BpL, c[5], c[4], c[2]); |
| 199 | #define PIXEL10_61 hq4x_Interp6(pOut+BpL, c[5], c[4], c[1]); |
| 200 | #define PIXEL10_81 hq4x_Interp8(pOut+BpL, c[5], c[4]); |
| 201 | #define PIXEL10_83 hq4x_Interp8(pOut+BpL, c[4], c[2]); |
| 202 | #define PIXEL11_0 *((int*)(pOut+BpL+BPP)) = c[5]; |
| 203 | #define PIXEL11_30 hq4x_Interp3(pOut+BpL+BPP, c[5], c[1]); |
| 204 | #define PIXEL11_31 hq4x_Interp3(pOut+BpL+BPP, c[5], c[4]); |
| 205 | #define PIXEL11_32 hq4x_Interp3(pOut+BpL+BPP, c[5], c[2]); |
| 206 | #define PIXEL11_70 hq4x_Interp7(pOut+BpL+BPP, c[5], c[4], c[2]); |
| 207 | #define PIXEL12_0 *((int*)(pOut+BpL+BPP2)) = c[5]; |
| 208 | #define PIXEL12_30 hq4x_Interp3(pOut+BpL+BPP2, c[5], c[3]); |
| 209 | #define PIXEL12_31 hq4x_Interp3(pOut+BpL+BPP2, c[5], c[2]); |
| 210 | #define PIXEL12_32 hq4x_Interp3(pOut+BpL+BPP2, c[5], c[6]); |
| 211 | #define PIXEL12_70 hq4x_Interp7(pOut+BpL+BPP2, c[5], c[6], c[2]); |
| 212 | #define PIXEL13_0 *((int*)(pOut+BpL+BPP3)) = c[5]; |
| 213 | #define PIXEL13_10 hq4x_Interp1(pOut+BpL+BPP3, c[5], c[3]); |
| 214 | #define PIXEL13_12 hq4x_Interp1(pOut+BpL+BPP3, c[5], c[6]); |
| 215 | #define PIXEL13_14 hq4x_Interp1(pOut+BpL+BPP3, c[6], c[5]); |
| 216 | #define PIXEL13_21 hq4x_Interp2(pOut+BpL+BPP3, c[6], c[5], c[2]); |
| 217 | #define PIXEL13_31 hq4x_Interp3(pOut+BpL+BPP3, c[5], c[2]); |
| 218 | #define PIXEL13_50 hq4x_Interp5(pOut+BpL+BPP3, c[6], c[5]); |
| 219 | #define PIXEL13_60 hq4x_Interp6(pOut+BpL+BPP3, c[5], c[6], c[2]); |
| 220 | #define PIXEL13_61 hq4x_Interp6(pOut+BpL+BPP3, c[5], c[6], c[3]); |
| 221 | #define PIXEL13_82 hq4x_Interp8(pOut+BpL+BPP3, c[5], c[6]); |
| 222 | #define PIXEL13_83 hq4x_Interp8(pOut+BpL+BPP3, c[6], c[2]); |
| 223 | #define PIXEL20_0 *((int*)(pOut+BpL+BpL)) = c[5]; |
| 224 | #define PIXEL20_10 hq4x_Interp1(pOut+BpL+BpL, c[5], c[7]); |
| 225 | #define PIXEL20_12 hq4x_Interp1(pOut+BpL+BpL, c[5], c[4]); |
| 226 | #define PIXEL20_14 hq4x_Interp1(pOut+BpL+BpL, c[4], c[5]); |
| 227 | #define PIXEL20_21 hq4x_Interp2(pOut+BpL+BpL, c[4], c[5], c[8]); |
| 228 | #define PIXEL20_31 hq4x_Interp3(pOut+BpL+BpL, c[5], c[8]); |
| 229 | #define PIXEL20_50 hq4x_Interp5(pOut+BpL+BpL, c[4], c[5]); |
| 230 | #define PIXEL20_60 hq4x_Interp6(pOut+BpL+BpL, c[5], c[4], c[8]); |
| 231 | #define PIXEL20_61 hq4x_Interp6(pOut+BpL+BpL, c[5], c[4], c[7]); |
| 232 | #define PIXEL20_82 hq4x_Interp8(pOut+BpL+BpL, c[5], c[4]); |
| 233 | #define PIXEL20_83 hq4x_Interp8(pOut+BpL+BpL, c[4], c[8]); |
| 234 | #define PIXEL21_0 *((int*)(pOut+BpL+BpL+BPP)) = c[5]; |
| 235 | #define PIXEL21_30 hq4x_Interp3(pOut+BpL+BpL+BPP, c[5], c[7]); |
| 236 | #define PIXEL21_31 hq4x_Interp3(pOut+BpL+BpL+BPP, c[5], c[8]); |
| 237 | #define PIXEL21_32 hq4x_Interp3(pOut+BpL+BpL+BPP, c[5], c[4]); |
| 238 | #define PIXEL21_70 hq4x_Interp7(pOut+BpL+BpL+BPP, c[5], c[4], c[8]); |
| 239 | #define PIXEL22_0 *((int*)(pOut+BpL+BpL+BPP2)) = c[5]; |
| 240 | #define PIXEL22_30 hq4x_Interp3(pOut+BpL+BpL+BPP2, c[5], c[9]); |
| 241 | #define PIXEL22_31 hq4x_Interp3(pOut+BpL+BpL+BPP2, c[5], c[6]); |
| 242 | #define PIXEL22_32 hq4x_Interp3(pOut+BpL+BpL+BPP2, c[5], c[8]); |
| 243 | #define PIXEL22_70 hq4x_Interp7(pOut+BpL+BpL+BPP2, c[5], c[6], c[8]); |
| 244 | #define PIXEL23_0 *((int*)(pOut+BpL+BpL+BPP3)) = c[5]; |
| 245 | #define PIXEL23_10 hq4x_Interp1(pOut+BpL+BpL+BPP3, c[5], c[9]); |
| 246 | #define PIXEL23_11 hq4x_Interp1(pOut+BpL+BpL+BPP3, c[5], c[6]); |
| 247 | #define PIXEL23_13 hq4x_Interp1(pOut+BpL+BpL+BPP3, c[6], c[5]); |
| 248 | #define PIXEL23_21 hq4x_Interp2(pOut+BpL+BpL+BPP3, c[6], c[5], c[8]); |
| 249 | #define PIXEL23_32 hq4x_Interp3(pOut+BpL+BpL+BPP3, c[5], c[8]); |
| 250 | #define PIXEL23_50 hq4x_Interp5(pOut+BpL+BpL+BPP3, c[6], c[5]); |
| 251 | #define PIXEL23_60 hq4x_Interp6(pOut+BpL+BpL+BPP3, c[5], c[6], c[8]); |
| 252 | #define PIXEL23_61 hq4x_Interp6(pOut+BpL+BpL+BPP3, c[5], c[6], c[9]); |
| 253 | #define PIXEL23_81 hq4x_Interp8(pOut+BpL+BpL+BPP3, c[5], c[6]); |
| 254 | #define PIXEL23_83 hq4x_Interp8(pOut+BpL+BpL+BPP3, c[6], c[8]); |
| 255 | #define PIXEL30_0 *((int*)(pOut+BpL+BpL+BpL)) = c[5]; |
| 256 | #define PIXEL30_11 hq4x_Interp1(pOut+BpL+BpL+BpL, c[5], c[8]); |
| 257 | #define PIXEL30_12 hq4x_Interp1(pOut+BpL+BpL+BpL, c[5], c[4]); |
| 258 | #define PIXEL30_20 hq4x_Interp2(pOut+BpL+BpL+BpL, c[5], c[8], c[4]); |
| 259 | #define PIXEL30_50 hq4x_Interp5(pOut+BpL+BpL+BpL, c[8], c[4]); |
| 260 | #define PIXEL30_80 hq4x_Interp8(pOut+BpL+BpL+BpL, c[5], c[7]); |
| 261 | #define PIXEL30_81 hq4x_Interp8(pOut+BpL+BpL+BpL, c[5], c[8]); |
| 262 | #define PIXEL30_82 hq4x_Interp8(pOut+BpL+BpL+BpL, c[5], c[4]); |
| 263 | #define PIXEL31_0 *((int*)(pOut+BpL+BpL+BpL+BPP)) = c[5]; |
| 264 | #define PIXEL31_10 hq4x_Interp1(pOut+BpL+BpL+BpL+BPP, c[5], c[7]); |
| 265 | #define PIXEL31_11 hq4x_Interp1(pOut+BpL+BpL+BpL+BPP, c[5], c[8]); |
| 266 | #define PIXEL31_13 hq4x_Interp1(pOut+BpL+BpL+BpL+BPP, c[8], c[5]); |
| 267 | #define PIXEL31_21 hq4x_Interp2(pOut+BpL+BpL+BpL+BPP, c[8], c[5], c[4]); |
| 268 | #define PIXEL31_32 hq4x_Interp3(pOut+BpL+BpL+BpL+BPP, c[5], c[4]); |
| 269 | #define PIXEL31_50 hq4x_Interp5(pOut+BpL+BpL+BpL+BPP, c[8], c[5]); |
| 270 | #define PIXEL31_60 hq4x_Interp6(pOut+BpL+BpL+BpL+BPP, c[5], c[8], c[4]); |
| 271 | #define PIXEL31_61 hq4x_Interp6(pOut+BpL+BpL+BpL+BPP, c[5], c[8], c[7]); |
| 272 | #define PIXEL31_81 hq4x_Interp8(pOut+BpL+BpL+BpL+BPP, c[5], c[8]); |
| 273 | #define PIXEL31_83 hq4x_Interp8(pOut+BpL+BpL+BpL+BPP, c[8], c[4]); |
| 274 | #define PIXEL32_0 *((int*)(pOut+BpL+BpL+BpL+BPP2)) = c[5]; |
| 275 | #define PIXEL32_10 hq4x_Interp1(pOut+BpL+BpL+BpL+BPP2, c[5], c[9]); |
| 276 | #define PIXEL32_12 hq4x_Interp1(pOut+BpL+BpL+BpL+BPP2, c[5], c[8]); |
| 277 | #define PIXEL32_14 hq4x_Interp1(pOut+BpL+BpL+BpL+BPP2, c[8], c[5]); |
| 278 | #define PIXEL32_21 hq4x_Interp2(pOut+BpL+BpL+BpL+BPP2, c[8], c[5], c[6]); |
| 279 | #define PIXEL32_31 hq4x_Interp3(pOut+BpL+BpL+BpL+BPP2, c[5], c[6]); |
| 280 | #define PIXEL32_50 hq4x_Interp5(pOut+BpL+BpL+BpL+BPP2, c[8], c[5]); |
| 281 | #define PIXEL32_60 hq4x_Interp6(pOut+BpL+BpL+BpL+BPP2, c[5], c[8], c[6]); |
| 282 | #define PIXEL32_61 hq4x_Interp6(pOut+BpL+BpL+BpL+BPP2, c[5], c[8], c[9]); |
| 283 | #define PIXEL32_82 hq4x_Interp8(pOut+BpL+BpL+BpL+BPP2, c[5], c[8]); |
| 284 | #define PIXEL32_83 hq4x_Interp8(pOut+BpL+BpL+BpL+BPP2, c[8], c[6]); |
| 285 | #define PIXEL33_0 *((int*)(pOut+BpL+BpL+BpL+BPP3)) = c[5]; |
| 286 | #define PIXEL33_11 hq4x_Interp1(pOut+BpL+BpL+BpL+BPP3, c[5], c[6]); |
| 287 | #define PIXEL33_12 hq4x_Interp1(pOut+BpL+BpL+BpL+BPP3, c[5], c[8]); |
| 288 | #define PIXEL33_20 hq4x_Interp2(pOut+BpL+BpL+BpL+BPP3, c[5], c[8], c[6]); |
| 289 | #define PIXEL33_50 hq4x_Interp5(pOut+BpL+BpL+BpL+BPP3, c[8], c[6]); |
| 290 | #define PIXEL33_80 hq4x_Interp8(pOut+BpL+BpL+BpL+BPP3, c[5], c[9]); |
| 291 | #define PIXEL33_81 hq4x_Interp8(pOut+BpL+BpL+BpL+BPP3, c[5], c[6]); |
| 292 | #define PIXEL33_82 hq4x_Interp8(pOut+BpL+BpL+BpL+BPP3, c[5], c[8]); |
| 293 | |
| 294 | |
| 295 | |
| 296 | inline bool Diff_16(uint16 w1, uint16 w2) |
| 297 | { |
| 298 | YUV1 = RGB16toYUV(w1); |
| 299 | YUV2 = RGB16toYUV(w2); |
| 300 | return ( ( abs((YUV1 & Amask) - (YUV2 & Amask)) > trA ) || |
| 301 | ( abs((YUV1 & Ymask) - (YUV2 & Ymask)) > trY ) || |
| 302 | ( abs((YUV1 & Umask) - (YUV2 & Umask)) > trU ) || |
| 303 | ( abs((YUV1 & Vmask) - (YUV2 & Vmask)) > trV ) ); |
| 304 | } |
| 305 | inline bool Diff_32(uint32 w1, uint32 w2) |
| 306 | { |
| 307 | YUV1 = RGB32toYUV(w1); |
| 308 | YUV2 = RGB32toYUV(w2); |
| 309 | return ( ( abs((YUV1 & Amask) - (YUV2 & Amask)) > trA ) || |
| 310 | ( abs((YUV1 & Ymask) - (YUV2 & Ymask)) > trY ) || |
| 311 | ( abs((YUV1 & Umask) - (YUV2 & Umask)) > trU ) || |
| 312 | ( abs((YUV1 & Vmask) - (YUV2 & Vmask)) > trV ) ); |
| 313 | } |
| 314 | |
| 315 | void hq4x_16( unsigned char * pIn, unsigned char * pOut, int Xres, int Yres, int SrcPPL, int BpL ) |
| 316 | { |
| 317 | #define hq4x_Interp1 hq4x_Interp1_16 |
| 318 | #define hq4x_Interp2 hq4x_Interp2_16 |
| 319 | #define hq4x_Interp3 hq4x_Interp3_16 |
| 320 | #define hq4x_Interp4 hq4x_Interp4_16 |
| 321 | #define hq4x_Interp5 hq4x_Interp5_16 |
| 322 | #define hq4x_Interp6 hq4x_Interp6_16 |
| 323 | #define hq4x_Interp7 hq4x_Interp7_16 |
| 324 | #define hq4x_Interp8 hq4x_Interp8_16 |
| 325 | #define Diff Diff_16 |
| 326 | #define BPP 2 |
| 327 | #define BPP2 4 |
| 328 | #define BPP3 6 |
| 329 | |
| 330 | int i, j, k; |
| 331 | int prevline, nextline; |
| 332 | uint16 w[10]; |
| 333 | uint16 c[10]; |
| 334 | |
| 335 | // +----+----+----+ |
| 336 | // | | | | |
| 337 | // | w1 | w2 | w3 | |
| 338 | // +----+----+----+ |
| 339 | // | | | | |
| 340 | // | w4 | w5 | w6 | |
| 341 | // +----+----+----+ |
| 342 | // | | | | |
| 343 | // | w7 | w8 | w9 | |
| 344 | // +----+----+----+ |
| 345 | |
| 346 | for (j=0; j<Yres; j++) |
| 347 | { |
| 348 | if (j>0) prevline = -SrcPPL*2; else prevline = 0; |
| 349 | if (j<Yres-1) nextline = SrcPPL*2; else nextline = 0; |
| 350 | |
| 351 | for (i=0; i<Xres; i++) |
| 352 | { |
| 353 | w[2] = *((uint16*)(pIn + prevline)); |
| 354 | w[5] = *((uint16*)pIn); |
| 355 | w[8] = *((uint16*)(pIn + nextline)); |
| 356 | |
| 357 | if (i>0) |
| 358 | { |
| 359 | w[1] = *((uint16*)(pIn + prevline - 2)); |
| 360 | w[4] = *((uint16*)(pIn - 2)); |
| 361 | w[7] = *((uint16*)(pIn + nextline - 2)); |
| 362 | } |
| 363 | else |
| 364 | { |
| 365 | w[1] = w[2]; |
| 366 | w[4] = w[5]; |
| 367 | w[7] = w[8]; |
| 368 | } |
| 369 | |
| 370 | if (i<Xres-1) |
| 371 | { |
| 372 | w[3] = *((uint16*)(pIn + prevline + 2)); |
| 373 | w[6] = *((uint16*)(pIn + 2)); |
| 374 | w[9] = *((uint16*)(pIn + nextline + 2)); |
| 375 | } |
| 376 | else |
| 377 | { |
| 378 | w[3] = w[2]; |
| 379 | w[6] = w[5]; |
| 380 | w[9] = w[8]; |
| 381 | } |
| 382 | |
| 383 | int pattern = 0; |
| 384 | int flag = 1; |
| 385 | |
| 386 | YUV1 = RGB16toYUV(w[5]); |
| 387 | |
| 388 | for (k=1; k<=9; k++) |
| 389 | { |
| 390 | if (k==5) continue; |
| 391 | |
| 392 | if ( w[k] != w[5] ) |
| 393 | { |
| 394 | YUV2 = RGB16toYUV(w[k]); |
| 395 | if ( ( abs((YUV1 & Amask) - (YUV2 & Amask)) > trA ) || |
| 396 | ( abs((YUV1 & Ymask) - (YUV2 & Ymask)) > trY ) || |
| 397 | ( abs((YUV1 & Umask) - (YUV2 & Umask)) > trU ) || |
| 398 | ( abs((YUV1 & Vmask) - (YUV2 & Vmask)) > trV ) ) |
| 399 | pattern |= flag; |
| 400 | } |
| 401 | flag <<= 1; |
| 402 | } |
| 403 | |
| 404 | for (k=1; k<=9; k++) |
| 405 | c[k] = w[k]; |
| 406 | |
| 407 | #include "TextureFilters_hq4x.h" |
| 408 | |
| 409 | pIn+=2; |
| 410 | pOut+=8; |
| 411 | } |
| 412 | pIn += 2*(SrcPPL-Xres); |
| 413 | pOut+= 8*(SrcPPL-Xres); |
| 414 | pOut+=BpL; |
| 415 | pOut+=BpL; |
| 416 | pOut+=BpL; |
| 417 | } |
| 418 | |
| 419 | #undef BPP |
| 420 | #undef BPP2 |
| 421 | #undef BPP3 |
| 422 | #undef Diff |
| 423 | #undef hq4x_Interp1 |
| 424 | #undef hq4x_Interp2 |
| 425 | #undef hq4x_Interp3 |
| 426 | #undef hq4x_Interp4 |
| 427 | #undef hq4x_Interp5 |
| 428 | #undef hq4x_Interp6 |
| 429 | #undef hq4x_Interp7 |
| 430 | #undef hq4x_Interp8 |
| 431 | } |
| 432 | |
| 433 | void hq4x_32( unsigned char * pIn, unsigned char * pOut, int Xres, int Yres, int SrcPPL, int BpL ) |
| 434 | { |
| 435 | #define hq4x_Interp1 hq4x_Interp1_32 |
| 436 | #define hq4x_Interp2 hq4x_Interp2_32 |
| 437 | #define hq4x_Interp3 hq4x_Interp3_32 |
| 438 | #define hq4x_Interp4 hq4x_Interp4_32 |
| 439 | #define hq4x_Interp5 hq4x_Interp5_32 |
| 440 | #define hq4x_Interp6 hq4x_Interp6_32 |
| 441 | #define hq4x_Interp7 hq4x_Interp7_32 |
| 442 | #define hq4x_Interp8 hq4x_Interp8_32 |
| 443 | #define Diff Diff_32 |
| 444 | #define BPP 4 |
| 445 | #define BPP2 8 |
| 446 | #define BPP3 12 |
| 447 | |
| 448 | int i, j, k; |
| 449 | int prevline, nextline; |
| 450 | uint32 w[10]; |
| 451 | uint32 c[10]; |
| 452 | |
| 453 | // +----+----+----+ |
| 454 | // | | | | |
| 455 | // | w1 | w2 | w3 | |
| 456 | // +----+----+----+ |
| 457 | // | | | | |
| 458 | // | w4 | w5 | w6 | |
| 459 | // +----+----+----+ |
| 460 | // | | | | |
| 461 | // | w7 | w8 | w9 | |
| 462 | // +----+----+----+ |
| 463 | |
| 464 | for (j=0; j<Yres; j++) |
| 465 | { |
| 466 | if (j>0) prevline = -SrcPPL*4; else prevline = 0; |
| 467 | if (j<Yres-1) nextline = SrcPPL*4; else nextline = 0; |
| 468 | |
| 469 | for (i=0; i<Xres; i++) |
| 470 | { |
| 471 | w[2] = *((uint32*)(pIn + prevline)); |
| 472 | w[5] = *((uint32*)pIn); |
| 473 | w[8] = *((uint32*)(pIn + nextline)); |
| 474 | |
| 475 | if (i>0) |
| 476 | { |
| 477 | w[1] = *((uint32*)(pIn + prevline - 4)); |
| 478 | w[4] = *((uint32*)(pIn - 4)); |
| 479 | w[7] = *((uint32*)(pIn + nextline - 4)); |
| 480 | } |
| 481 | else |
| 482 | { |
| 483 | w[1] = w[2]; |
| 484 | w[4] = w[5]; |
| 485 | w[7] = w[8]; |
| 486 | } |
| 487 | |
| 488 | if (i<Xres-1) |
| 489 | { |
| 490 | w[3] = *((uint32*)(pIn + prevline + 4)); |
| 491 | w[6] = *((uint32*)(pIn + 4)); |
| 492 | w[9] = *((uint32*)(pIn + nextline + 4)); |
| 493 | } |
| 494 | else |
| 495 | { |
| 496 | w[3] = w[2]; |
| 497 | w[6] = w[5]; |
| 498 | w[9] = w[8]; |
| 499 | } |
| 500 | |
| 501 | int pattern = 0; |
| 502 | int flag = 1; |
| 503 | |
| 504 | YUV1 = RGB32toYUV(w[5]); |
| 505 | |
| 506 | for (k=1; k<=9; k++) |
| 507 | { |
| 508 | if (k==5) continue; |
| 509 | |
| 510 | if ( w[k] != w[5] ) |
| 511 | { |
| 512 | YUV2 = RGB32toYUV(w[k]); |
| 513 | if ( ( abs((YUV1 & Amask) - (YUV2 & Amask)) > trA ) || |
| 514 | ( abs((YUV1 & Ymask) - (YUV2 & Ymask)) > trY ) || |
| 515 | ( abs((YUV1 & Umask) - (YUV2 & Umask)) > trU ) || |
| 516 | ( abs((YUV1 & Vmask) - (YUV2 & Vmask)) > trV ) ) |
| 517 | pattern |= flag; |
| 518 | } |
| 519 | flag <<= 1; |
| 520 | } |
| 521 | |
| 522 | for (k=1; k<=9; k++) |
| 523 | c[k] = w[k]; |
| 524 | |
| 525 | #include "TextureFilters_hq4x.h" |
| 526 | |
| 527 | pIn+=4; |
| 528 | pOut+=16; |
| 529 | } |
| 530 | |
| 531 | pIn += 4*(SrcPPL-Xres); |
| 532 | pOut+= 16*(SrcPPL-Xres); |
| 533 | pOut+=BpL; |
| 534 | pOut+=BpL; |
| 535 | pOut+=BpL; |
| 536 | } |
| 537 | |
| 538 | #undef BPP |
| 539 | #undef BPP2 |
| 540 | #undef BPP3 |
| 541 | #undef Diff |
| 542 | #undef hq4x_Interp1 |
| 543 | #undef hq4x_Interp2 |
| 544 | #undef hq4x_Interp3 |
| 545 | #undef hq4x_Interp4 |
| 546 | #undef hq4x_Interp5 |
| 547 | #undef hq4x_Interp6 |
| 548 | #undef hq4x_Interp7 |
| 549 | #undef hq4x_Interp8 |
| 550 | } |
| 551 | |
| 552 | void hq4x_InitLUTs(void) |
| 553 | { |
| 554 | static bool done = false; |
| 555 | int i, j, k, r, g, b, Y, u, v; |
| 556 | |
| 557 | if( !done ) |
| 558 | { |
| 559 | for (i=0; i<16; i++) |
| 560 | { |
| 561 | for (j=0; j<16; j++) |
| 562 | { |
| 563 | for (k=0; k<16; k++) |
| 564 | { |
| 565 | r = i << 4; |
| 566 | g = j << 4; |
| 567 | b = k << 4; |
| 568 | Y = (r + g + b) >> 2; |
| 569 | u = 128 + ((r - b) >> 2); |
| 570 | v = 128 + ((-r + 2*g -b)>>3); |
| 571 | RGBtoYUV[ (i << 8) + (j << 4) + k ] = (Y<<16) + (u<<8) + v; |
| 572 | } |
| 573 | } |
| 574 | } |
| 575 | done = true; |
| 576 | } |
| 577 | } |
| 578 | |