d9e74a6f |
1 | /* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * |
2 | * Mupen64plus-rsp-hle - ucode3.cpp * |
3 | * Mupen64Plus homepage: http://code.google.com/p/mupen64plus/ * |
4 | * Copyright (C) 2009 Richard Goedeken * |
5 | * Copyright (C) 2002 Hacktarux * |
6 | * * |
7 | * This program is free software; you can redistribute it and/or modify * |
8 | * it under the terms of the GNU General Public License as published by * |
9 | * the Free Software Foundation; either version 2 of the License, or * |
10 | * (at your option) any later version. * |
11 | * * |
12 | * This program is distributed in the hope that it will be useful, * |
13 | * but WITHOUT ANY WARRANTY; without even the implied warranty of * |
14 | * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * |
15 | * GNU General Public License for more details. * |
16 | * * |
17 | * You should have received a copy of the GNU General Public License * |
18 | * along with this program; if not, write to the * |
19 | * Free Software Foundation, Inc., * |
20 | * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. * |
21 | * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * */ |
22 | |
23 | # include <string.h> |
24 | # include <stdio.h> |
25 | |
26 | extern "C" { |
27 | #include "m64p_types.h" |
28 | #include "hle.h" |
29 | #include "alist_internal.h" |
30 | } |
31 | |
32 | /* |
33 | static void SPNOOP (u32 inst1, u32 inst2) { |
34 | DebugMessage(M64MSG_ERROR, "Unknown/Unimplemented Audio Command %i in ABI 3", (int)(inst1 >> 24)); |
35 | } |
36 | */ |
37 | |
38 | extern const u16 ResampleLUT [0x200]; |
39 | |
40 | extern u32 loopval; |
41 | |
42 | extern s16 Env_Dry; |
43 | extern s16 Env_Wet; |
44 | extern s16 Vol_Left; |
45 | extern s16 Vol_Right; |
46 | extern s16 VolTrg_Left; |
47 | extern s32 VolRamp_Left; |
48 | //extern u16 VolRate_Left; |
49 | extern s16 VolTrg_Right; |
50 | extern s32 VolRamp_Right; |
51 | //extern u16 VolRate_Right; |
52 | |
53 | |
54 | extern short hleMixerWorkArea[256]; |
55 | extern u16 adpcmtable[0x88]; |
56 | |
57 | extern u8 BufferSpace[0x10000]; |
58 | |
59 | /* |
60 | static void SETVOL3 (u32 inst1, u32 inst2) { // Swapped Rate_Left and Vol |
61 | u8 Flags = (u8)(inst1 >> 0x10); |
62 | if (Flags & 0x4) { // 288 |
63 | if (Flags & 0x2) { // 290 |
64 | VolTrg_Left = *(s16*)&inst1; |
65 | VolRamp_Left = *(s32*)&inst2; |
66 | } else { |
67 | VolTrg_Right = *(s16*)&inst1; |
68 | VolRamp_Right = *(s32*)&inst2; |
69 | } |
70 | } else { |
71 | Vol_Left = *(s16*)&inst1; |
72 | Env_Dry = (s16)(*(s32*)&inst2 >> 0x10); |
73 | Env_Wet = *(s16*)&inst2; |
74 | } |
75 | } |
76 | */ |
77 | static void SETVOL3 (u32 inst1, u32 inst2) { |
78 | u8 Flags = (u8)(inst1 >> 0x10); |
79 | if (Flags & 0x4) { // 288 |
80 | if (Flags & 0x2) { // 290 |
81 | Vol_Left = (s16)inst1; // 0x50 |
82 | Env_Dry = (s16)(inst2 >> 0x10); // 0x4E |
83 | Env_Wet = (s16)inst2; // 0x4C |
84 | } else { |
85 | VolTrg_Right = (s16)inst1; // 0x46 |
86 | //VolRamp_Right = (u16)(inst2 >> 0x10) | (s32)(s16)(inst2 << 0x10); |
87 | VolRamp_Right = (s32)inst2; // 0x48/0x4A |
88 | } |
89 | } else { |
90 | VolTrg_Left = (s16)inst1; // 0x40 |
91 | VolRamp_Left = (s32)inst2; // 0x42/0x44 |
92 | } |
93 | } |
94 | |
95 | static void ENVMIXER3 (u32 inst1, u32 inst2) { |
96 | u8 flags = (u8)((inst1 >> 16) & 0xff); |
97 | u32 addy = (inst2 & 0xFFFFFF); |
98 | |
99 | short *inp=(short *)(BufferSpace+0x4F0); |
100 | short *out=(short *)(BufferSpace+0x9D0); |
101 | short *aux1=(short *)(BufferSpace+0xB40); |
102 | short *aux2=(short *)(BufferSpace+0xCB0); |
103 | short *aux3=(short *)(BufferSpace+0xE20); |
104 | s32 MainR; |
105 | s32 MainL; |
106 | s32 AuxR; |
107 | s32 AuxL; |
108 | int i1,o1,a1,a2,a3; |
109 | //unsigned short AuxIncRate=1; |
110 | short zero[8]; |
111 | memset(zero,0,16); |
112 | |
113 | s32 LAdder, LAcc, LVol; |
114 | s32 RAdder, RAcc, RVol; |
115 | s16 RSig, LSig; // Most significant part of the Ramp Value |
116 | s16 Wet, Dry; |
117 | s16 LTrg, RTrg; |
118 | |
119 | Vol_Right = (s16)inst1; |
120 | |
121 | if (flags & A_INIT) { |
122 | LAdder = VolRamp_Left / 8; |
123 | LAcc = 0; |
124 | LVol = Vol_Left; |
125 | LSig = (s16)(VolRamp_Left >> 16); |
126 | |
127 | RAdder = VolRamp_Right / 8; |
128 | RAcc = 0; |
129 | RVol = Vol_Right; |
130 | RSig = (s16)(VolRamp_Right >> 16); |
131 | |
132 | Wet = (s16)Env_Wet; Dry = (s16)Env_Dry; // Save Wet/Dry values |
133 | LTrg = VolTrg_Left; RTrg = VolTrg_Right; // Save Current Left/Right Targets |
134 | } else { |
135 | memcpy((u8 *)hleMixerWorkArea, rsp.RDRAM+addy, 80); |
136 | Wet = *(s16 *)(hleMixerWorkArea + 0); // 0-1 |
137 | Dry = *(s16 *)(hleMixerWorkArea + 2); // 2-3 |
138 | LTrg = *(s16 *)(hleMixerWorkArea + 4); // 4-5 |
139 | RTrg = *(s16 *)(hleMixerWorkArea + 6); // 6-7 |
140 | LAdder = *(s32 *)(hleMixerWorkArea + 8); // 8-9 (hleMixerWorkArea is a 16bit pointer) |
141 | RAdder = *(s32 *)(hleMixerWorkArea + 10); // 10-11 |
142 | LAcc = *(s32 *)(hleMixerWorkArea + 12); // 12-13 |
143 | RAcc = *(s32 *)(hleMixerWorkArea + 14); // 14-15 |
144 | LVol = *(s32 *)(hleMixerWorkArea + 16); // 16-17 |
145 | RVol = *(s32 *)(hleMixerWorkArea + 18); // 18-19 |
146 | LSig = *(s16 *)(hleMixerWorkArea + 20); // 20-21 |
147 | RSig = *(s16 *)(hleMixerWorkArea + 22); // 22-23 |
148 | //u32 test = *(s32 *)(hleMixerWorkArea + 24); // 22-23 |
149 | //if (test != 0x13371337) |
150 | } |
151 | |
152 | |
153 | //if(!(flags&A_AUX)) { |
154 | // AuxIncRate=0; |
155 | // aux2=aux3=zero; |
156 | //} |
157 | |
158 | for (int y = 0; y < (0x170/2); y++) { |
159 | |
160 | // Left |
161 | LAcc += LAdder; |
162 | LVol += (LAcc >> 16); |
163 | LAcc &= 0xFFFF; |
164 | |
165 | // Right |
166 | RAcc += RAdder; |
167 | RVol += (RAcc >> 16); |
168 | RAcc &= 0xFFFF; |
169 | // **************************************************************** |
170 | // Clamp Left |
171 | if (LSig >= 0) { // VLT |
172 | if (LVol > LTrg) { |
173 | LVol = LTrg; |
174 | } |
175 | } else { // VGE |
176 | if (LVol < LTrg) { |
177 | LVol = LTrg; |
178 | } |
179 | } |
180 | |
181 | // Clamp Right |
182 | if (RSig >= 0) { // VLT |
183 | if (RVol > RTrg) { |
184 | RVol = RTrg; |
185 | } |
186 | } else { // VGE |
187 | if (RVol < RTrg) { |
188 | RVol = RTrg; |
189 | } |
190 | } |
191 | // **************************************************************** |
192 | MainL = ((Dry * LVol) + 0x4000) >> 15; |
193 | MainR = ((Dry * RVol) + 0x4000) >> 15; |
194 | |
195 | o1 = out [y^S]; |
196 | a1 = aux1[y^S]; |
197 | i1 = inp [y^S]; |
198 | |
199 | o1+=((i1*MainL)+0x4000)>>15; |
200 | a1+=((i1*MainR)+0x4000)>>15; |
201 | |
202 | // **************************************************************** |
203 | |
204 | if(o1>32767) o1=32767; |
205 | else if(o1<-32768) o1=-32768; |
206 | |
207 | if(a1>32767) a1=32767; |
208 | else if(a1<-32768) a1=-32768; |
209 | |
210 | // **************************************************************** |
211 | |
212 | out[y^S]=o1; |
213 | aux1[y^S]=a1; |
214 | |
215 | // **************************************************************** |
216 | //if (!(flags&A_AUX)) { |
217 | a2 = aux2[y^S]; |
218 | a3 = aux3[y^S]; |
219 | |
220 | AuxL = ((Wet * LVol) + 0x4000) >> 15; |
221 | AuxR = ((Wet * RVol) + 0x4000) >> 15; |
222 | |
223 | a2+=((i1*AuxL)+0x4000)>>15; |
224 | a3+=((i1*AuxR)+0x4000)>>15; |
225 | |
226 | if(a2>32767) a2=32767; |
227 | else if(a2<-32768) a2=-32768; |
228 | |
229 | if(a3>32767) a3=32767; |
230 | else if(a3<-32768) a3=-32768; |
231 | |
232 | aux2[y^S]=a2; |
233 | aux3[y^S]=a3; |
234 | } |
235 | //} |
236 | |
237 | *(s16 *)(hleMixerWorkArea + 0) = Wet; // 0-1 |
238 | *(s16 *)(hleMixerWorkArea + 2) = Dry; // 2-3 |
239 | *(s16 *)(hleMixerWorkArea + 4) = LTrg; // 4-5 |
240 | *(s16 *)(hleMixerWorkArea + 6) = RTrg; // 6-7 |
241 | *(s32 *)(hleMixerWorkArea + 8) = LAdder; // 8-9 (hleMixerWorkArea is a 16bit pointer) |
242 | *(s32 *)(hleMixerWorkArea + 10) = RAdder; // 10-11 |
243 | *(s32 *)(hleMixerWorkArea + 12) = LAcc; // 12-13 |
244 | *(s32 *)(hleMixerWorkArea + 14) = RAcc; // 14-15 |
245 | *(s32 *)(hleMixerWorkArea + 16) = LVol; // 16-17 |
246 | *(s32 *)(hleMixerWorkArea + 18) = RVol; // 18-19 |
247 | *(s16 *)(hleMixerWorkArea + 20) = LSig; // 20-21 |
248 | *(s16 *)(hleMixerWorkArea + 22) = RSig; // 22-23 |
249 | //*(u32 *)(hleMixerWorkArea + 24) = 0x13371337; // 22-23 |
250 | memcpy(rsp.RDRAM+addy, (u8 *)hleMixerWorkArea,80); |
251 | } |
252 | |
253 | static void CLEARBUFF3 (u32 inst1, u32 inst2) { |
254 | u16 addr = (u16)(inst1 & 0xffff); |
255 | u16 count = (u16)(inst2 & 0xffff); |
256 | memset(BufferSpace+addr+0x4f0, 0, count); |
257 | } |
258 | |
259 | static void MIXER3 (u32 inst1, u32 inst2) { // Needs accuracy verification... |
260 | u16 dmemin = (u16)(inst2 >> 0x10) + 0x4f0; |
261 | u16 dmemout = (u16)(inst2 & 0xFFFF) + 0x4f0; |
262 | //u8 flags = (u8)((inst1 >> 16) & 0xff); |
263 | s32 gain = (s16)(inst1 & 0xFFFF); |
264 | s32 temp; |
265 | |
266 | for (int x=0; x < 0x170; x+=2) { // I think I can do this a lot easier |
267 | temp = (*(s16 *)(BufferSpace+dmemin+x) * gain) >> 15; |
268 | temp += *(s16 *)(BufferSpace+dmemout+x); |
269 | |
270 | if ((s32)temp > 32767) |
271 | temp = 32767; |
272 | if ((s32)temp < -32768) |
273 | temp = -32768; |
274 | |
275 | *(u16 *)(BufferSpace+dmemout+x) = (u16)(temp & 0xFFFF); |
276 | } |
277 | } |
278 | |
279 | static void LOADBUFF3 (u32 inst1, u32 inst2) { |
280 | u32 v0; |
281 | u32 cnt = (((inst1 >> 0xC)+3)&0xFFC); |
282 | v0 = (inst2 & 0xfffffc); |
283 | u32 src = (inst1&0xffc)+0x4f0; |
284 | memcpy (BufferSpace+src, rsp.RDRAM+v0, cnt); |
285 | } |
286 | |
287 | static void SAVEBUFF3 (u32 inst1, u32 inst2) { |
288 | u32 v0; |
289 | u32 cnt = (((inst1 >> 0xC)+3)&0xFFC); |
290 | v0 = (inst2 & 0xfffffc); |
291 | u32 src = (inst1&0xffc)+0x4f0; |
292 | memcpy (rsp.RDRAM+v0, BufferSpace+src, cnt); |
293 | } |
294 | |
295 | static void LOADADPCM3 (u32 inst1, u32 inst2) { // Loads an ADPCM table - Works 100% Now 03-13-01 |
296 | u32 v0; |
297 | v0 = (inst2 & 0xffffff); |
298 | //memcpy (dmem+0x3f0, rsp.RDRAM+v0, inst1&0xffff); |
299 | //assert ((inst1&0xffff) <= 0x80); |
300 | u16 *table = (u16 *)(rsp.RDRAM+v0); |
301 | for (u32 x = 0; x < ((inst1&0xffff)>>0x4); x++) { |
302 | adpcmtable[(0x0+(x<<3))^S] = table[0]; |
303 | adpcmtable[(0x1+(x<<3))^S] = table[1]; |
304 | |
305 | adpcmtable[(0x2+(x<<3))^S] = table[2]; |
306 | adpcmtable[(0x3+(x<<3))^S] = table[3]; |
307 | |
308 | adpcmtable[(0x4+(x<<3))^S] = table[4]; |
309 | adpcmtable[(0x5+(x<<3))^S] = table[5]; |
310 | |
311 | adpcmtable[(0x6+(x<<3))^S] = table[6]; |
312 | adpcmtable[(0x7+(x<<3))^S] = table[7]; |
313 | table += 8; |
314 | } |
315 | } |
316 | |
317 | static void DMEMMOVE3 (u32 inst1, u32 inst2) { // Needs accuracy verification... |
318 | u32 v0, v1; |
319 | u32 cnt; |
320 | v0 = (inst1 & 0xFFFF) + 0x4f0; |
321 | v1 = (inst2 >> 0x10) + 0x4f0; |
322 | u32 count = ((inst2+3) & 0xfffc); |
323 | |
324 | //memcpy (dmem+v1, dmem+v0, count-1); |
325 | for (cnt = 0; cnt < count; cnt++) { |
326 | *(u8 *)(BufferSpace+((cnt+v1)^S8)) = *(u8 *)(BufferSpace+((cnt+v0)^S8)); |
327 | } |
328 | } |
329 | |
330 | static void SETLOOP3 (u32 inst1, u32 inst2) { |
331 | loopval = (inst2 & 0xffffff); |
332 | } |
333 | |
334 | static void ADPCM3 (u32 inst1, u32 inst2) { // Verified to be 100% Accurate... |
335 | unsigned char Flags=(u8)(inst2>>0x1c)&0xff; |
336 | //unsigned short Gain=(u16)(inst1&0xffff); |
337 | unsigned int Address=(inst1 & 0xffffff);// + SEGMENTS[(inst2>>24)&0xf]; |
338 | unsigned short inPtr=(inst2>>12)&0xf; |
339 | //short *out=(s16 *)(testbuff+(AudioOutBuffer>>2)); |
340 | short *out=(short *)(BufferSpace+(inst2&0xfff)+0x4f0); |
341 | //unsigned char *in=(unsigned char *)(BufferSpace+((inst2>>12)&0xf)+0x4f0); |
342 | short count=(short)((inst2 >> 16)&0xfff); |
343 | unsigned char icode; |
344 | unsigned char code; |
345 | int vscale; |
346 | unsigned short index; |
347 | unsigned short j; |
348 | int a[8]; |
349 | short *book1,*book2; |
350 | |
351 | memset(out,0,32); |
352 | |
353 | if(!(Flags&0x1)) |
354 | { |
355 | if(Flags&0x2) |
356 | {/* |
357 | for(int i=0;i<16;i++) |
358 | { |
359 | out[i]=*(short *)&rsp.RDRAM[(loopval+i*2)^2]; |
360 | }*/ |
361 | memcpy(out,&rsp.RDRAM[loopval],32); |
362 | } |
363 | else |
364 | {/* |
365 | for(int i=0;i<16;i++) |
366 | { |
367 | out[i]=*(short *)&rsp.RDRAM[(Address+i*2)^2]; |
368 | }*/ |
369 | memcpy(out,&rsp.RDRAM[Address],32); |
370 | } |
371 | } |
372 | |
373 | int l1=out[14^S]; |
374 | int l2=out[15^S]; |
375 | int inp1[8]; |
376 | int inp2[8]; |
377 | out+=16; |
378 | while(count>0) |
379 | { |
380 | // the first interation through, these values are |
381 | // either 0 in the case of A_INIT, from a special |
382 | // area of memory in the case of A_LOOP or just |
383 | // the values we calculated the last time |
384 | |
385 | code=BufferSpace[(0x4f0+inPtr)^S8]; |
386 | index=code&0xf; |
387 | index<<=4; // index into the adpcm code table |
388 | book1=(short *)&adpcmtable[index]; |
389 | book2=book1+8; |
390 | code>>=4; // upper nibble is scale |
391 | vscale=(0x8000>>((12-code)-1)); // very strange. 0x8000 would be .5 in 16:16 format |
392 | // so this appears to be a fractional scale based |
393 | // on the 12 based inverse of the scale value. note |
394 | // that this could be negative, in which case we do |
395 | // not use the calculated vscale value... see the |
396 | // if(code>12) check below |
397 | |
398 | inPtr++; // coded adpcm data lies next |
399 | j=0; |
400 | while(j<8) // loop of 8, for 8 coded nibbles from 4 bytes |
401 | // which yields 8 short pcm values |
402 | { |
403 | icode=BufferSpace[(0x4f0+inPtr)^S8]; |
404 | inPtr++; |
405 | |
406 | inp1[j]=(s16)((icode&0xf0)<<8); // this will in effect be signed |
407 | if(code<12) |
408 | inp1[j]=((int)((int)inp1[j]*(int)vscale)>>16); |
409 | /*else |
410 | int catchme=1;*/ |
411 | j++; |
412 | |
413 | inp1[j]=(s16)((icode&0xf)<<12); |
414 | if(code<12) |
415 | inp1[j]=((int)((int)inp1[j]*(int)vscale)>>16); |
416 | /*else |
417 | int catchme=1;*/ |
418 | j++; |
419 | } |
420 | j=0; |
421 | while(j<8) |
422 | { |
423 | icode=BufferSpace[(0x4f0+inPtr)^S8]; |
424 | inPtr++; |
425 | |
426 | inp2[j]=(short)((icode&0xf0)<<8); // this will in effect be signed |
427 | if(code<12) |
428 | inp2[j]=((int)((int)inp2[j]*(int)vscale)>>16); |
429 | /*else |
430 | int catchme=1;*/ |
431 | j++; |
432 | |
433 | inp2[j]=(short)((icode&0xf)<<12); |
434 | if(code<12) |
435 | inp2[j]=((int)((int)inp2[j]*(int)vscale)>>16); |
436 | /*else |
437 | int catchme=1;*/ |
438 | j++; |
439 | } |
440 | |
441 | a[0]= (int)book1[0]*(int)l1; |
442 | a[0]+=(int)book2[0]*(int)l2; |
443 | a[0]+=(int)inp1[0]*(int)2048; |
444 | |
445 | a[1] =(int)book1[1]*(int)l1; |
446 | a[1]+=(int)book2[1]*(int)l2; |
447 | a[1]+=(int)book2[0]*inp1[0]; |
448 | a[1]+=(int)inp1[1]*(int)2048; |
449 | |
450 | a[2] =(int)book1[2]*(int)l1; |
451 | a[2]+=(int)book2[2]*(int)l2; |
452 | a[2]+=(int)book2[1]*inp1[0]; |
453 | a[2]+=(int)book2[0]*inp1[1]; |
454 | a[2]+=(int)inp1[2]*(int)2048; |
455 | |
456 | a[3] =(int)book1[3]*(int)l1; |
457 | a[3]+=(int)book2[3]*(int)l2; |
458 | a[3]+=(int)book2[2]*inp1[0]; |
459 | a[3]+=(int)book2[1]*inp1[1]; |
460 | a[3]+=(int)book2[0]*inp1[2]; |
461 | a[3]+=(int)inp1[3]*(int)2048; |
462 | |
463 | a[4] =(int)book1[4]*(int)l1; |
464 | a[4]+=(int)book2[4]*(int)l2; |
465 | a[4]+=(int)book2[3]*inp1[0]; |
466 | a[4]+=(int)book2[2]*inp1[1]; |
467 | a[4]+=(int)book2[1]*inp1[2]; |
468 | a[4]+=(int)book2[0]*inp1[3]; |
469 | a[4]+=(int)inp1[4]*(int)2048; |
470 | |
471 | a[5] =(int)book1[5]*(int)l1; |
472 | a[5]+=(int)book2[5]*(int)l2; |
473 | a[5]+=(int)book2[4]*inp1[0]; |
474 | a[5]+=(int)book2[3]*inp1[1]; |
475 | a[5]+=(int)book2[2]*inp1[2]; |
476 | a[5]+=(int)book2[1]*inp1[3]; |
477 | a[5]+=(int)book2[0]*inp1[4]; |
478 | a[5]+=(int)inp1[5]*(int)2048; |
479 | |
480 | a[6] =(int)book1[6]*(int)l1; |
481 | a[6]+=(int)book2[6]*(int)l2; |
482 | a[6]+=(int)book2[5]*inp1[0]; |
483 | a[6]+=(int)book2[4]*inp1[1]; |
484 | a[6]+=(int)book2[3]*inp1[2]; |
485 | a[6]+=(int)book2[2]*inp1[3]; |
486 | a[6]+=(int)book2[1]*inp1[4]; |
487 | a[6]+=(int)book2[0]*inp1[5]; |
488 | a[6]+=(int)inp1[6]*(int)2048; |
489 | |
490 | a[7] =(int)book1[7]*(int)l1; |
491 | a[7]+=(int)book2[7]*(int)l2; |
492 | a[7]+=(int)book2[6]*inp1[0]; |
493 | a[7]+=(int)book2[5]*inp1[1]; |
494 | a[7]+=(int)book2[4]*inp1[2]; |
495 | a[7]+=(int)book2[3]*inp1[3]; |
496 | a[7]+=(int)book2[2]*inp1[4]; |
497 | a[7]+=(int)book2[1]*inp1[5]; |
498 | a[7]+=(int)book2[0]*inp1[6]; |
499 | a[7]+=(int)inp1[7]*(int)2048; |
500 | |
501 | for(j=0;j<8;j++) |
502 | { |
503 | a[j^S]>>=11; |
504 | if(a[j^S]>32767) a[j^S]=32767; |
505 | else if(a[j^S]<-32768) a[j^S]=-32768; |
506 | *(out++)=a[j^S]; |
507 | //*(out+j)=a[j^S]; |
508 | } |
509 | //out += 0x10; |
510 | l1=a[6]; |
511 | l2=a[7]; |
512 | |
513 | a[0]= (int)book1[0]*(int)l1; |
514 | a[0]+=(int)book2[0]*(int)l2; |
515 | a[0]+=(int)inp2[0]*(int)2048; |
516 | |
517 | a[1] =(int)book1[1]*(int)l1; |
518 | a[1]+=(int)book2[1]*(int)l2; |
519 | a[1]+=(int)book2[0]*inp2[0]; |
520 | a[1]+=(int)inp2[1]*(int)2048; |
521 | |
522 | a[2] =(int)book1[2]*(int)l1; |
523 | a[2]+=(int)book2[2]*(int)l2; |
524 | a[2]+=(int)book2[1]*inp2[0]; |
525 | a[2]+=(int)book2[0]*inp2[1]; |
526 | a[2]+=(int)inp2[2]*(int)2048; |
527 | |
528 | a[3] =(int)book1[3]*(int)l1; |
529 | a[3]+=(int)book2[3]*(int)l2; |
530 | a[3]+=(int)book2[2]*inp2[0]; |
531 | a[3]+=(int)book2[1]*inp2[1]; |
532 | a[3]+=(int)book2[0]*inp2[2]; |
533 | a[3]+=(int)inp2[3]*(int)2048; |
534 | |
535 | a[4] =(int)book1[4]*(int)l1; |
536 | a[4]+=(int)book2[4]*(int)l2; |
537 | a[4]+=(int)book2[3]*inp2[0]; |
538 | a[4]+=(int)book2[2]*inp2[1]; |
539 | a[4]+=(int)book2[1]*inp2[2]; |
540 | a[4]+=(int)book2[0]*inp2[3]; |
541 | a[4]+=(int)inp2[4]*(int)2048; |
542 | |
543 | a[5] =(int)book1[5]*(int)l1; |
544 | a[5]+=(int)book2[5]*(int)l2; |
545 | a[5]+=(int)book2[4]*inp2[0]; |
546 | a[5]+=(int)book2[3]*inp2[1]; |
547 | a[5]+=(int)book2[2]*inp2[2]; |
548 | a[5]+=(int)book2[1]*inp2[3]; |
549 | a[5]+=(int)book2[0]*inp2[4]; |
550 | a[5]+=(int)inp2[5]*(int)2048; |
551 | |
552 | a[6] =(int)book1[6]*(int)l1; |
553 | a[6]+=(int)book2[6]*(int)l2; |
554 | a[6]+=(int)book2[5]*inp2[0]; |
555 | a[6]+=(int)book2[4]*inp2[1]; |
556 | a[6]+=(int)book2[3]*inp2[2]; |
557 | a[6]+=(int)book2[2]*inp2[3]; |
558 | a[6]+=(int)book2[1]*inp2[4]; |
559 | a[6]+=(int)book2[0]*inp2[5]; |
560 | a[6]+=(int)inp2[6]*(int)2048; |
561 | |
562 | a[7] =(int)book1[7]*(int)l1; |
563 | a[7]+=(int)book2[7]*(int)l2; |
564 | a[7]+=(int)book2[6]*inp2[0]; |
565 | a[7]+=(int)book2[5]*inp2[1]; |
566 | a[7]+=(int)book2[4]*inp2[2]; |
567 | a[7]+=(int)book2[3]*inp2[3]; |
568 | a[7]+=(int)book2[2]*inp2[4]; |
569 | a[7]+=(int)book2[1]*inp2[5]; |
570 | a[7]+=(int)book2[0]*inp2[6]; |
571 | a[7]+=(int)inp2[7]*(int)2048; |
572 | |
573 | for(j=0;j<8;j++) |
574 | { |
575 | a[j^S]>>=11; |
576 | if(a[j^S]>32767) a[j^S]=32767; |
577 | else if(a[j^S]<-32768) a[j^S]=-32768; |
578 | *(out++)=a[j^S]; |
579 | //*(out+j+0x1f8)=a[j^S]; |
580 | } |
581 | l1=a[6]; |
582 | l2=a[7]; |
583 | |
584 | count-=32; |
585 | } |
586 | out-=16; |
587 | memcpy(&rsp.RDRAM[Address],out,32); |
588 | } |
589 | |
590 | static void RESAMPLE3 (u32 inst1, u32 inst2) { |
591 | unsigned char Flags=(u8)((inst2>>0x1e)); |
592 | unsigned int Pitch=((inst2>>0xe)&0xffff)<<1; |
593 | u32 addy = (inst1 & 0xffffff); |
594 | unsigned int Accum=0; |
595 | unsigned int location; |
596 | s16 *lut; |
597 | short *dst; |
598 | s16 *src; |
599 | dst=(short *)(BufferSpace); |
600 | src=(s16 *)(BufferSpace); |
601 | u32 srcPtr=((((inst2>>2)&0xfff)+0x4f0)/2); |
602 | u32 dstPtr;//=(AudioOutBuffer/2); |
603 | s32 temp; |
604 | s32 accum; |
605 | |
606 | //if (addy > (1024*1024*8)) |
607 | // addy = (inst2 & 0xffffff); |
608 | |
609 | srcPtr -= 4; |
610 | |
611 | if (inst2 & 0x3) { |
612 | dstPtr = 0x660/2; |
613 | } else { |
614 | dstPtr = 0x4f0/2; |
615 | } |
616 | |
617 | if ((Flags & 0x1) == 0) { |
618 | for (int x=0; x < 4; x++) //memcpy (src+srcPtr, rsp.RDRAM+addy, 0x8); |
619 | src[(srcPtr+x)^S] = ((u16 *)rsp.RDRAM)[((addy/2)+x)^S]; |
620 | Accum = *(u16 *)(rsp.RDRAM+addy+10); |
621 | } else { |
622 | for (int x=0; x < 4; x++) |
623 | src[(srcPtr+x)^S] = 0;//*(u16 *)(rsp.RDRAM+((addy+x)^2)); |
624 | } |
625 | |
626 | for(int i=0;i < 0x170/2;i++) { |
627 | location = (((Accum * 0x40) >> 0x10) * 8); |
628 | //location = (Accum >> 0xa) << 0x3; |
629 | lut = (s16 *)(((u8 *)ResampleLUT) + location); |
630 | |
631 | temp = ((s32)*(s16*)(src+((srcPtr+0)^S))*((s32)((s16)lut[0]))); |
632 | accum = (s32)(temp >> 15); |
633 | |
634 | temp = ((s32)*(s16*)(src+((srcPtr+1)^S))*((s32)((s16)lut[1]))); |
635 | accum += (s32)(temp >> 15); |
636 | |
637 | temp = ((s32)*(s16*)(src+((srcPtr+2)^S))*((s32)((s16)lut[2]))); |
638 | accum += (s32)(temp >> 15); |
639 | |
640 | temp = ((s32)*(s16*)(src+((srcPtr+3)^S))*((s32)((s16)lut[3]))); |
641 | accum += (s32)(temp >> 15); |
642 | /* temp = ((s64)*(s16*)(src+((srcPtr+0)^S))*((s64)((s16)lut[0]<<1))); |
643 | if (temp & 0x8000) temp = (temp^0x8000) + 0x10000; |
644 | else temp = (temp^0x8000); |
645 | temp = (s32)(temp >> 16); |
646 | if ((s32)temp > 32767) temp = 32767; |
647 | if ((s32)temp < -32768) temp = -32768; |
648 | accum = (s32)(s16)temp; |
649 | |
650 | temp = ((s64)*(s16*)(src+((srcPtr+1)^S))*((s64)((s16)lut[1]<<1))); |
651 | if (temp & 0x8000) temp = (temp^0x8000) + 0x10000; |
652 | else temp = (temp^0x8000); |
653 | temp = (s32)(temp >> 16); |
654 | if ((s32)temp > 32767) temp = 32767; |
655 | if ((s32)temp < -32768) temp = -32768; |
656 | accum += (s32)(s16)temp; |
657 | |
658 | temp = ((s64)*(s16*)(src+((srcPtr+2)^S))*((s64)((s16)lut[2]<<1))); |
659 | if (temp & 0x8000) temp = (temp^0x8000) + 0x10000; |
660 | else temp = (temp^0x8000); |
661 | temp = (s32)(temp >> 16); |
662 | if ((s32)temp > 32767) temp = 32767; |
663 | if ((s32)temp < -32768) temp = -32768; |
664 | accum += (s32)(s16)temp; |
665 | |
666 | temp = ((s64)*(s16*)(src+((srcPtr+3)^S))*((s64)((s16)lut[3]<<1))); |
667 | if (temp & 0x8000) temp = (temp^0x8000) + 0x10000; |
668 | else temp = (temp^0x8000); |
669 | temp = (s32)(temp >> 16); |
670 | if ((s32)temp > 32767) temp = 32767; |
671 | if ((s32)temp < -32768) temp = -32768; |
672 | accum += (s32)(s16)temp;*/ |
673 | |
674 | if (accum > 32767) accum = 32767; |
675 | if (accum < -32768) accum = -32768; |
676 | |
677 | dst[dstPtr^S] = (accum); |
678 | dstPtr++; |
679 | Accum += Pitch; |
680 | srcPtr += (Accum>>16); |
681 | Accum&=0xffff; |
682 | } |
683 | for (int x=0; x < 4; x++) |
684 | ((u16 *)rsp.RDRAM)[((addy/2)+x)^S] = src[(srcPtr+x)^S]; |
685 | *(u16 *)(rsp.RDRAM+addy+10) = Accum; |
686 | } |
687 | |
688 | static void INTERLEAVE3 (u32 inst1, u32 inst2) { // Needs accuracy verification... |
689 | //u32 inL, inR; |
690 | u16 *outbuff = (u16 *)(BufferSpace + 0x4f0);//(u16 *)(AudioOutBuffer+dmem); |
691 | u16 *inSrcR; |
692 | u16 *inSrcL; |
693 | u16 Left, Right, Left2, Right2; |
694 | |
695 | //inR = inst2 & 0xFFFF; |
696 | //inL = (inst2 >> 16) & 0xFFFF; |
697 | |
698 | inSrcR = (u16 *)(BufferSpace+0xb40); |
699 | inSrcL = (u16 *)(BufferSpace+0x9d0); |
700 | |
701 | for (int x = 0; x < (0x170/4); x++) { |
702 | Left=*(inSrcL++); |
703 | Right=*(inSrcR++); |
704 | Left2=*(inSrcL++); |
705 | Right2=*(inSrcR++); |
706 | |
707 | #ifdef M64P_BIG_ENDIAN |
708 | *(outbuff++)=Right; |
709 | *(outbuff++)=Left; |
710 | *(outbuff++)=Right2; |
711 | *(outbuff++)=Left2; |
712 | #else |
713 | *(outbuff++)=Right2; |
714 | *(outbuff++)=Left2; |
715 | *(outbuff++)=Right; |
716 | *(outbuff++)=Left; |
717 | #endif |
718 | /* |
719 | Left=*(inSrcL++); |
720 | Right=*(inSrcR++); |
721 | *(outbuff++)=(u16)Left; |
722 | Left >>= 16; |
723 | *(outbuff++)=(u16)Right; |
724 | Right >>= 16; |
725 | *(outbuff++)=(u16)Left; |
726 | *(outbuff++)=(u16)Right;*/ |
727 | } |
728 | } |
729 | |
730 | //static void UNKNOWN (u32 inst1, u32 inst2); |
731 | /* |
732 | typedef struct { |
733 | unsigned char sync; |
734 | |
735 | unsigned char error_protection : 1; // 0=yes, 1=no |
736 | unsigned char lay : 2; // 4-lay = layerI, II or III |
737 | unsigned char version : 1; // 3=mpeg 1.0, 2=mpeg 2.5 0=mpeg 2.0 |
738 | unsigned char sync2 : 4; |
739 | |
740 | unsigned char extension : 1; // Unknown |
741 | unsigned char padding : 1; // padding |
742 | unsigned char sampling_freq : 2; // see table below |
743 | unsigned char bitrate_index : 4; // see table below |
744 | |
745 | unsigned char emphasis : 2; //see table below |
746 | unsigned char original : 1; // 0=no 1=yes |
747 | unsigned char copyright : 1; // 0=no 1=yes |
748 | unsigned char mode_ext : 2; // used with "joint stereo" mode |
749 | unsigned char mode : 2; // Channel Mode |
750 | } mp3struct; |
751 | |
752 | mp3struct mp3; |
753 | FILE *mp3dat; |
754 | */ |
755 | |
756 | static void WHATISTHIS (u32 inst1, u32 inst2) { |
757 | } |
758 | |
759 | //static FILE *fp = fopen ("d:\\mp3info.txt", "wt"); |
760 | u32 setaddr; |
761 | static void MP3ADDY (u32 inst1, u32 inst2) { |
762 | setaddr = (inst2 & 0xffffff); |
763 | } |
764 | |
765 | extern "C" { |
766 | void rsp_run(void); |
767 | void mp3setup (unsigned int inst1, unsigned int inst2, unsigned int t8); |
768 | } |
769 | |
770 | extern u32 base, dmembase; |
771 | extern "C" { |
772 | extern char *pDMEM; |
773 | } |
774 | void MP3 (u32 inst1, u32 inst2); |
775 | /* |
776 | { |
777 | // return; |
778 | // Setup Registers... |
779 | mp3setup (inst1, inst2, 0xFA0); |
780 | |
781 | // Setup Memory Locations... |
782 | //u32 base = ((u32*)dmem)[0xFD0/4]; // Should be 000291A0 |
783 | memcpy (BufferSpace, dmembase+rsp.RDRAM, 0x10); |
784 | ((u32*)BufferSpace)[0x0] = base; |
785 | ((u32*)BufferSpace)[0x008/4] += base; |
786 | ((u32*)BufferSpace)[0xFFC/4] = loopval; |
787 | ((u32*)BufferSpace)[0xFF8/4] = dmembase; |
788 | |
789 | memcpy (imem+0x238, rsp.RDRAM+((u32*)BufferSpace)[0x008/4], 0x9C0); |
790 | ((u32*)BufferSpace)[0xFF4/4] = setaddr; |
791 | pDMEM = (char *)BufferSpace; |
792 | rsp_run (void); |
793 | dmembase = ((u32*)BufferSpace)[0xFF8/4]; |
794 | loopval = ((u32*)BufferSpace)[0xFFC/4]; |
795 | //0x1A98 SW S1, 0x0FF4 (R0) |
796 | //0x1A9C SW S0, 0x0FF8 (R0) |
797 | //0x1AA0 SW T7, 0x0FFC (R0) |
798 | //0x1AA4 SW T3, 0x0FF0 (R0) |
799 | //fprintf (fp, "mp3: inst1: %08X, inst2: %08X\n", inst1, inst2); |
800 | }*/ |
801 | /* |
802 | FFT = Fast Fourier Transform |
803 | DCT = Discrete Cosine Transform |
804 | MPEG-1 Layer 3 retains Layer 2's 1152-sample window, as well as the FFT polyphase filter for |
805 | backward compatibility, but adds a modified DCT filter. DCT's advantages over DFTs (discrete |
806 | Fourier transforms) include half as many multiply-accumulate operations and half the |
807 | generated coefficients because the sinusoidal portion of the calculation is absent, and DCT |
808 | generally involves simpler math. The finite lengths of a conventional DCTs' bandpass impulse |
809 | responses, however, may result in block-boundary effects. MDCTs overlap the analysis blocks |
810 | and lowpass-filter the decoded audio to remove aliases, eliminating these effects. MDCTs also |
811 | have a higher transform coding gain than the standard DCT, and their basic functions |
812 | correspond to better bandpass response. |
813 | |
814 | MPEG-1 Layer 3's DCT sub-bands are unequally sized, and correspond to the human auditory |
815 | system's critical bands. In Layer 3 decoders must support both constant- and variable-bit-rate |
816 | bit streams. (However, many Layer 1 and 2 decoders also handle variable bit rates). Finally, |
817 | Layer 3 encoders Huffman-code the quantized coefficients before archiving or transmission for |
818 | additional lossless compression. Bit streams range from 32 to 320 kbps, and 128-kbps rates |
819 | achieve near-CD quality, an important specification to enable dual-channel ISDN |
820 | (integrated-services-digital-network) to be the future high-bandwidth pipe to the home. |
821 | |
822 | */ |
823 | static void DISABLE (u32 inst1, u32 inst2) { |
824 | //MessageBox (NULL, "Help", "ABI 3 Command 0", MB_OK); |
825 | //ChangeABI (5); |
826 | } |
827 | |
828 | |
829 | extern "C" const acmd_callback_t ABI3[0x10] = { |
830 | DISABLE , ADPCM3 , CLEARBUFF3, ENVMIXER3 , LOADBUFF3, RESAMPLE3 , SAVEBUFF3, MP3, |
831 | MP3ADDY, SETVOL3, DMEMMOVE3 , LOADADPCM3 , MIXER3 , INTERLEAVE3, WHATISTHIS , SETLOOP3 |
832 | }; |
833 | |
834 | |