// SONIC ROBO BLAST 2 //----------------------------------------------------------------------------- // Copyright (C) 1998-2000 by DooM Legacy Team. // Copyright (C) 1999-2022 by Sonic Team Junior. // // This program is free software distributed under the // terms of the GNU General Public License, version 2. // See the 'LICENSE' file for more details. //----------------------------------------------------------------------------- /// \file hw_md2.c /// \brief 3D Model Handling /// Inspired from md2.c by Mete Ciragan (mete@swissquake.ch) #ifdef __GNUC__ #include #endif #include #include #include #include #include "../d_main.h" #include "../doomdef.h" #include "../doomstat.h" #include "../fastcmp.h" #ifdef HWRENDER #include "hw_drv.h" #include "hw_light.h" #include "hw_md2.h" #include "../d_main.h" #include "../r_bsp.h" #include "../r_fps.h" #include "../r_main.h" #include "../m_misc.h" #include "../w_wad.h" #include "../z_zone.h" #include "../r_things.h" #include "../r_draw.h" #include "../p_tick.h" #include "hw_model.h" #include "hw_main.h" #include "../v_video.h" #ifdef HAVE_PNG #ifndef _MSC_VER #ifndef _LARGEFILE64_SOURCE #define _LARGEFILE64_SOURCE #endif #endif #ifndef _LFS64_LARGEFILE #define _LFS64_LARGEFILE #endif #ifndef _FILE_OFFSET_BITS #define _FILE_OFFSET_BITS 0 #endif #include "png.h" #ifndef PNG_READ_SUPPORTED #undef HAVE_PNG #endif #if PNG_LIBPNG_VER < 100207 //#undef HAVE_PNG #endif #endif #ifndef errno #include "errno.h" #endif md2_t md2_models[NUMSPRITES]; md2_t md2_playermodels[MAXSKINS]; /* * free model */ #if 0 static void md2_freeModel (model_t *model) { UnloadModel(model); } #endif // // load model // // Hurdler: the current path is the Legacy.exe path static model_t *md2_readModel(const char *filename) { //Filename checking fixed ~Monster Iestyn and Golden if (FIL_FileExists(va("%s"PATHSEP"%s", srb2home, filename))) return LoadModel(va("%s"PATHSEP"%s", srb2home, filename), PU_STATIC); if (FIL_FileExists(va("%s"PATHSEP"%s", srb2path, filename))) return LoadModel(va("%s"PATHSEP"%s", srb2path, filename), PU_STATIC); return NULL; } static inline void md2_printModelInfo (model_t *model) { #if 0 INT32 i; CONS_Debug(DBG_RENDER, "magic:\t\t\t%c%c%c%c\n", model->header.magic>>24, (model->header.magic>>16)&0xff, (model->header.magic>>8)&0xff, model->header.magic&0xff); CONS_Debug(DBG_RENDER, "version:\t\t%d\n", model->header.version); CONS_Debug(DBG_RENDER, "skinWidth:\t\t%d\n", model->header.skinWidth); CONS_Debug(DBG_RENDER, "skinHeight:\t\t%d\n", model->header.skinHeight); CONS_Debug(DBG_RENDER, "frameSize:\t\t%d\n", model->header.frameSize); CONS_Debug(DBG_RENDER, "numSkins:\t\t%d\n", model->header.numSkins); CONS_Debug(DBG_RENDER, "numVertices:\t\t%d\n", model->header.numVertices); CONS_Debug(DBG_RENDER, "numTexCoords:\t\t%d\n", model->header.numTexCoords); CONS_Debug(DBG_RENDER, "numTriangles:\t\t%d\n", model->header.numTriangles); CONS_Debug(DBG_RENDER, "numGlCommands:\t\t%d\n", model->header.numGlCommands); CONS_Debug(DBG_RENDER, "numFrames:\t\t%d\n", model->header.numFrames); CONS_Debug(DBG_RENDER, "offsetSkins:\t\t%d\n", model->header.offsetSkins); CONS_Debug(DBG_RENDER, "offsetTexCoords:\t%d\n", model->header.offsetTexCoords); CONS_Debug(DBG_RENDER, "offsetTriangles:\t%d\n", model->header.offsetTriangles); CONS_Debug(DBG_RENDER, "offsetFrames:\t\t%d\n", model->header.offsetFrames); CONS_Debug(DBG_RENDER, "offsetGlCommands:\t%d\n", model->header.offsetGlCommands); CONS_Debug(DBG_RENDER, "offsetEnd:\t\t%d\n", model->header.offsetEnd); for (i = 0; i < model->header.numFrames; i++) CONS_Debug(DBG_RENDER, "%s ", model->frames[i].name); CONS_Debug(DBG_RENDER, "\n"); #else (void)model; #endif } #ifdef HAVE_PNG static void PNG_error(png_structp PNG, png_const_charp pngtext) { CONS_Debug(DBG_RENDER, "libpng error at %p: %s", PNG, pngtext); //I_Error("libpng error at %p: %s", PNG, pngtext); } static void PNG_warn(png_structp PNG, png_const_charp pngtext) { CONS_Debug(DBG_RENDER, "libpng warning at %p: %s", PNG, pngtext); } static GLTextureFormat_t PNG_Load(const char *filename, int *w, int *h, GLPatch_t *grpatch) { png_structp png_ptr; png_infop png_info_ptr; png_uint_32 width, height; int bit_depth, color_type; #ifdef PNG_SETJMP_SUPPORTED #ifdef USE_FAR_KEYWORD jmp_buf jmpbuf; #endif #endif volatile png_FILE_p png_FILE; //Filename checking fixed ~Monster Iestyn and Golden char *pngfilename = va("%s"PATHSEP"models"PATHSEP"%s", srb2home, filename); FIL_ForceExtension(pngfilename, ".png"); png_FILE = fopen(pngfilename, "rb"); if (!png_FILE) { pngfilename = va("%s"PATHSEP"models"PATHSEP"%s", srb2path, filename); FIL_ForceExtension(pngfilename, ".png"); png_FILE = fopen(pngfilename, "rb"); //CONS_Debug(DBG_RENDER, "M_SavePNG: Error on opening %s for loading\n", filename); if (!png_FILE) return 0; } png_ptr = png_create_read_struct(PNG_LIBPNG_VER_STRING, NULL, PNG_error, PNG_warn); if (!png_ptr) { CONS_Debug(DBG_RENDER, "PNG_Load: Error on initialize libpng\n"); fclose(png_FILE); return 0; } png_info_ptr = png_create_info_struct(png_ptr); if (!png_info_ptr) { CONS_Debug(DBG_RENDER, "PNG_Load: Error on allocate for libpng\n"); png_destroy_read_struct(&png_ptr, NULL, NULL); fclose(png_FILE); return 0; } #ifdef USE_FAR_KEYWORD if (setjmp(jmpbuf)) #else if (setjmp(png_jmpbuf(png_ptr))) #endif { //CONS_Debug(DBG_RENDER, "libpng load error on %s\n", filename); png_destroy_read_struct(&png_ptr, &png_info_ptr, NULL); fclose(png_FILE); Z_Free(grpatch->mipmap->data); return 0; } #ifdef USE_FAR_KEYWORD png_memcpy(png_jmpbuf(png_ptr), jmpbuf, sizeof jmp_buf); #endif png_init_io(png_ptr, png_FILE); #ifdef PNG_SET_USER_LIMITS_SUPPORTED png_set_user_limits(png_ptr, 2048, 2048); #endif png_read_info(png_ptr, png_info_ptr); png_get_IHDR(png_ptr, png_info_ptr, &width, &height, &bit_depth, &color_type, NULL, NULL, NULL); if (bit_depth == 16) png_set_strip_16(png_ptr); if (color_type == PNG_COLOR_TYPE_GRAY || color_type == PNG_COLOR_TYPE_GRAY_ALPHA) png_set_gray_to_rgb(png_ptr); else if (color_type == PNG_COLOR_TYPE_PALETTE) png_set_palette_to_rgb(png_ptr); if (png_get_valid(png_ptr, png_info_ptr, PNG_INFO_tRNS)) png_set_tRNS_to_alpha(png_ptr); else if (color_type != PNG_COLOR_TYPE_RGB_ALPHA && color_type != PNG_COLOR_TYPE_GRAY_ALPHA) { #if PNG_LIBPNG_VER < 10207 png_set_filler(png_ptr, 0xFF, PNG_FILLER_AFTER); #else png_set_add_alpha(png_ptr, 0xFF, PNG_FILLER_AFTER); #endif } png_read_update_info(png_ptr, png_info_ptr); { png_uint_32 i, pitch = png_get_rowbytes(png_ptr, png_info_ptr); png_bytep PNG_image = Z_Malloc(pitch*height, PU_HWRMODELTEXTURE, &grpatch->mipmap->data); png_bytepp row_pointers = png_malloc(png_ptr, height * sizeof (png_bytep)); for (i = 0; i < height; i++) row_pointers[i] = PNG_image + i*pitch; png_read_image(png_ptr, row_pointers); png_free(png_ptr, (png_voidp)row_pointers); } png_destroy_read_struct(&png_ptr, &png_info_ptr, NULL); fclose(png_FILE); *w = (int)width; *h = (int)height; return GL_TEXFMT_RGBA; } #endif typedef struct { UINT8 manufacturer; UINT8 version; UINT8 encoding; UINT8 bitsPerPixel; INT16 xmin; INT16 ymin; INT16 xmax; INT16 ymax; INT16 hDpi; INT16 vDpi; UINT8 colorMap[48]; UINT8 reserved; UINT8 numPlanes; INT16 bytesPerLine; INT16 paletteInfo; INT16 hScreenSize; INT16 vScreenSize; UINT8 filler[54]; } PcxHeader; static GLTextureFormat_t PCX_Load(const char *filename, int *w, int *h, GLPatch_t *grpatch) { PcxHeader header; #define PALSIZE 768 UINT8 palette[PALSIZE]; const UINT8 *pal; RGBA_t *image; size_t pw, ph, size, ptr = 0; INT32 ch, rep; FILE *file; //Filename checking fixed ~Monster Iestyn and Golden char *pcxfilename = va("%s"PATHSEP"models"PATHSEP"%s", srb2home, filename); FIL_ForceExtension(pcxfilename, ".pcx"); file = fopen(pcxfilename, "rb"); if (!file) { pcxfilename = va("%s"PATHSEP"models"PATHSEP"%s", srb2path, filename); FIL_ForceExtension(pcxfilename, ".pcx"); file = fopen(pcxfilename, "rb"); if (!file) return 0; } if (fread(&header, sizeof (PcxHeader), 1, file) != 1) { fclose(file); return 0; } if (header.bitsPerPixel != 8) { fclose(file); return 0; } fseek(file, -PALSIZE, SEEK_END); pw = *w = header.xmax - header.xmin + 1; ph = *h = header.ymax - header.ymin + 1; image = Z_Malloc(pw*ph*4, PU_HWRMODELTEXTURE, &grpatch->mipmap->data); if (fread(palette, sizeof (UINT8), PALSIZE, file) != PALSIZE) { Z_Free(image); fclose(file); return 0; } fseek(file, sizeof (PcxHeader), SEEK_SET); size = pw * ph; while (ptr < size) { ch = fgetc(file); //Hurdler: beurk if (ch >= 192) { rep = ch - 192; ch = fgetc(file); } else { rep = 1; } while (rep--) { pal = palette + ch*3; image[ptr].s.red = *pal++; image[ptr].s.green = *pal++; image[ptr].s.blue = *pal++; image[ptr].s.alpha = 0xFF; ptr++; } } fclose(file); return GL_TEXFMT_RGBA; } // -----------------+ // md2_loadTexture : Download a pcx or png texture for models // -----------------+ static void md2_loadTexture(md2_t *model) { patch_t *patch; GLPatch_t *grPatch = NULL; const char *filename = model->filename; if (model->grpatch) { patch = model->grpatch; grPatch = (GLPatch_t *)(patch->hardware); if (grPatch) Z_Free(grPatch->mipmap->data); } else model->grpatch = patch = Patch_Create(NULL, 0, NULL); if (!patch->hardware) Patch_AllocateHardwarePatch(patch); if (grPatch == NULL) grPatch = (GLPatch_t *)(patch->hardware); if (!grPatch->mipmap->downloaded && !grPatch->mipmap->data) { int w = 0, h = 0; UINT32 size; RGBA_t *image; #ifdef HAVE_PNG grPatch->mipmap->format = PNG_Load(filename, &w, &h, grPatch); if (grPatch->mipmap->format == 0) #endif grPatch->mipmap->format = PCX_Load(filename, &w, &h, grPatch); if (grPatch->mipmap->format == 0) { model->notexturefile = true; // mark it so its not searched for again repeatedly return; } grPatch->mipmap->downloaded = 0; grPatch->mipmap->flags = 0; patch->width = (INT16)w; patch->height = (INT16)h; grPatch->mipmap->width = (UINT16)w; grPatch->mipmap->height = (UINT16)h; // Lactozilla: Apply colour cube image = grPatch->mipmap->data; size = w*h; while (size--) { V_CubeApply(&image->s.red, &image->s.green, &image->s.blue); image++; } } HWD.pfnSetTexture(grPatch->mipmap); } // -----------------+ // md2_loadBlendTexture : Download a pcx or png texture for blending MD2 models // -----------------+ static void md2_loadBlendTexture(md2_t *model) { patch_t *patch; GLPatch_t *grPatch = NULL; char *filename = Z_Malloc(strlen(model->filename)+7, PU_STATIC, NULL); strcpy(filename, model->filename); FIL_ForceExtension(filename, "_blend.png"); if (model->blendgrpatch) { patch = model->blendgrpatch; grPatch = (GLPatch_t *)(patch->hardware); if (grPatch) Z_Free(grPatch->mipmap->data); } else model->blendgrpatch = patch = Patch_Create(NULL, 0, NULL); if (!patch->hardware) Patch_AllocateHardwarePatch(patch); if (grPatch == NULL) grPatch = (GLPatch_t *)(patch->hardware); if (!grPatch->mipmap->downloaded && !grPatch->mipmap->data) { int w = 0, h = 0; #ifdef HAVE_PNG grPatch->mipmap->format = PNG_Load(filename, &w, &h, grPatch); if (grPatch->mipmap->format == 0) #endif grPatch->mipmap->format = PCX_Load(filename, &w, &h, grPatch); if (grPatch->mipmap->format == 0) { model->noblendfile = true; // mark it so its not searched for again repeatedly Z_Free(filename); return; } grPatch->mipmap->downloaded = 0; grPatch->mipmap->flags = 0; patch->width = (INT16)w; patch->height = (INT16)h; grPatch->mipmap->width = (UINT16)w; grPatch->mipmap->height = (UINT16)h; } HWD.pfnSetTexture(grPatch->mipmap); // We do need to do this so that it can be cleared and knows to recreate it when necessary Z_Free(filename); } // Don't spam the console, or the OS with fopen requests! static boolean nomd2s = false; void HWR_InitModels(void) { size_t i; INT32 s; FILE *f; char name[24], filename[32]; float scale, offset; size_t prefixlen; CONS_Printf("HWR_InitModels()...\n"); for (s = 0; s < MAXSKINS; s++) { md2_playermodels[s].scale = -1.0f; md2_playermodels[s].model = NULL; md2_playermodels[s].grpatch = NULL; md2_playermodels[s].notexturefile = false; md2_playermodels[s].noblendfile = false; md2_playermodels[s].skin = -1; md2_playermodels[s].notfound = true; md2_playermodels[s].error = false; } for (i = 0; i < NUMSPRITES; i++) { md2_models[i].scale = -1.0f; md2_models[i].model = NULL; md2_models[i].grpatch = NULL; md2_models[i].notexturefile = false; md2_models[i].noblendfile = false; md2_models[i].skin = -1; md2_models[i].notfound = true; md2_models[i].error = false; } // read the models.dat file //Filename checking fixed ~Monster Iestyn and Golden f = fopen(va("%s"PATHSEP"%s", srb2home, "models.dat"), "rt"); if (!f) { f = fopen(va("%s"PATHSEP"%s", srb2path, "models.dat"), "rt"); if (!f) { CONS_Printf("%s %s\n", M_GetText("Error while loading models.dat:"), strerror(errno)); nomd2s = true; return; } } // length of the player model prefix prefixlen = strlen(PLAYERMODELPREFIX); while (fscanf(f, "%25s %31s %f %f", name, filename, &scale, &offset) == 4) { char *skinname = name; size_t len = strlen(name); // check for the player model prefix. if (!strnicmp(name, PLAYERMODELPREFIX, prefixlen) && (len > prefixlen)) { skinname += prefixlen; goto addskinmodel; } // add sprite model if (len == 4) // must be 4 characters long exactly. otherwise it's not a sprite name. { for (i = 0; i < NUMSPRITES; i++) { if (stricmp(name, sprnames[i]) == 0) { md2_models[i].scale = scale; md2_models[i].offset = offset; md2_models[i].notfound = false; strcpy(md2_models[i].filename, filename); goto modelfound; } } } addskinmodel: // add player model for (s = 0; s < MAXSKINS; s++) { if (stricmp(skinname, skins[s].name) == 0) { md2_playermodels[s].skin = s; md2_playermodels[s].scale = scale; md2_playermodels[s].offset = offset; md2_playermodels[s].notfound = false; strcpy(md2_playermodels[s].filename, filename); goto modelfound; } } modelfound: // move on to next line... continue; } fclose(f); } void HWR_AddPlayerModel(int skin) // For skins that were added after startup { FILE *f; char name[24], filename[32]; float scale, offset; size_t prefixlen; if (nomd2s) return; //CONS_Printf("HWR_AddPlayerModel()...\n"); // read the models.dat file //Filename checking fixed ~Monster Iestyn and Golden f = fopen(va("%s"PATHSEP"%s", srb2home, "models.dat"), "rt"); if (!f) { f = fopen(va("%s"PATHSEP"%s", srb2path, "models.dat"), "rt"); if (!f) { CONS_Printf("%s %s\n", M_GetText("Error while loading models.dat:"), strerror(errno)); nomd2s = true; return; } } // length of the player model prefix prefixlen = strlen(PLAYERMODELPREFIX); // Check for any models that match the names of player skins! while (fscanf(f, "%25s %31s %f %f", name, filename, &scale, &offset) == 4) { char *skinname = name; size_t len = strlen(name); // ignore the player model prefix. if (!strnicmp(name, PLAYERMODELPREFIX, prefixlen) && (len > prefixlen)) skinname += prefixlen; if (stricmp(skinname, skins[skin].name) == 0) { md2_playermodels[skin].skin = skin; md2_playermodels[skin].scale = scale; md2_playermodels[skin].offset = offset; md2_playermodels[skin].notfound = false; strcpy(md2_playermodels[skin].filename, filename); goto playermodelfound; } } md2_playermodels[skin].notfound = true; playermodelfound: fclose(f); } void HWR_AddSpriteModel(size_t spritenum) // For sprites that were added after startup { FILE *f; // name[24] is used to check for names in the models.dat file that match with sprites or player skins // sprite names are always 4 characters long, and names is for player skins can be up to 19 characters long // PLAYERMODELPREFIX is 6 characters long char name[24], filename[32]; float scale, offset; if (nomd2s) return; if (spritenum == SPR_PLAY) // Handled already NEWMD2: Per sprite, per-skin check return; // Read the models.dat file //Filename checking fixed ~Monster Iestyn and Golden f = fopen(va("%s"PATHSEP"%s", srb2home, "models.dat"), "rt"); if (!f) { f = fopen(va("%s"PATHSEP"%s", srb2path, "models.dat"), "rt"); if (!f) { CONS_Printf("%s %s\n", M_GetText("Error while loading models.dat:"), strerror(errno)); nomd2s = true; return; } } // Check for any models that match the names of sprite names! while (fscanf(f, "%25s %31s %f %f", name, filename, &scale, &offset) == 4) { // length of the sprite name size_t len = strlen(name); if (len != 4) // must be 4 characters long exactly. otherwise it's not a sprite name. continue; // check for the player model prefix. if (!strnicmp(name, PLAYERMODELPREFIX, strlen(PLAYERMODELPREFIX))) continue; // that's not a sprite... if (stricmp(name, sprnames[spritenum]) == 0) { md2_models[spritenum].scale = scale; md2_models[spritenum].offset = offset; md2_models[spritenum].notfound = false; strcpy(md2_models[spritenum].filename, filename); goto spritemodelfound; } } md2_models[spritenum].notfound = true; spritemodelfound: fclose(f); } // Define for getting accurate color brightness readings according to how the human eye sees them. // https://en.wikipedia.org/wiki/Relative_luminance // 0.2126 to red // 0.7152 to green // 0.0722 to blue #define SETBRIGHTNESS(brightness,r,g,b) \ brightness = (UINT8)(((1063*(UINT16)(r))/5000) + ((3576*(UINT16)(g))/5000) + ((361*(UINT16)(b))/5000)) static void HWR_CreateBlendedTexture(patch_t *gpatch, patch_t *blendgpatch, GLMipmap_t *grMipmap, INT32 skinnum, skincolornum_t color) { GLPatch_t *hwrPatch = gpatch->hardware; GLPatch_t *hwrBlendPatch = blendgpatch->hardware; UINT16 w = gpatch->width, h = gpatch->height; UINT32 size = w*h; RGBA_t *image, *blendimage, *cur, blendcolor; UINT16 translation[16]; // First the color index UINT8 cutoff[16]; // Brightness cutoff before using the next color UINT8 translen = 0; UINT8 i; blendcolor = V_GetColor(0); // initialize memset(translation, 0, sizeof(translation)); memset(cutoff, 0, sizeof(cutoff)); if (grMipmap->width == 0) { grMipmap->width = gpatch->width; grMipmap->height = gpatch->height; // no wrap around, no chroma key grMipmap->flags = 0; // setup the texture info grMipmap->format = GL_TEXFMT_RGBA; } if (grMipmap->data) { Z_Free(grMipmap->data); grMipmap->data = NULL; } cur = Z_Malloc(size*4, PU_HWRMODELTEXTURE, &grMipmap->data); memset(cur, 0x00, size*4); image = hwrPatch->mipmap->data; blendimage = hwrBlendPatch->mipmap->data; // TC_METALSONIC includes an actual skincolor translation, on top of its flashing. if (skinnum == TC_METALSONIC) color = SKINCOLOR_COBALT; if (color != SKINCOLOR_NONE && color < numskincolors) { UINT8 numdupes = 1; translation[translen] = skincolors[color].ramp[0]; cutoff[translen] = 255; for (i = 1; i < 16; i++) { if (translation[translen] == skincolors[color].ramp[i]) { numdupes++; continue; } if (translen > 0) { cutoff[translen] = cutoff[translen-1] - (256 / (16 / numdupes)); } numdupes = 1; translen++; translation[translen] = (UINT16)skincolors[color].ramp[i]; } translen++; } while (size--) { if (skinnum == TC_ALLWHITE) { // Turn everything white cur->s.red = cur->s.green = cur->s.blue = 255; cur->s.alpha = image->s.alpha; } else { // Everything below requires a blend image if (blendimage == NULL) { cur->rgba = image->rgba; goto skippixel; } // Metal Sonic dash mode if (skinnum == TC_DASHMODE) { if (image->s.alpha == 0 && blendimage->s.alpha == 0) { // Don't bother with blending the pixel if the alpha of the blend pixel is 0 cur->rgba = image->rgba; } else { UINT8 ialpha = 255 - blendimage->s.alpha, balpha = blendimage->s.alpha; RGBA_t icolor = *image, bcolor; memset(&bcolor, 0x00, sizeof(RGBA_t)); if (blendimage->s.alpha) { bcolor.s.blue = 0; bcolor.s.red = 255; bcolor.s.green = (blendimage->s.red + blendimage->s.green + blendimage->s.blue) / 3; } if (image->s.alpha && image->s.red > image->s.green << 1) // this is pretty arbitrary, but it works well for Metal Sonic { icolor.s.red = image->s.blue; icolor.s.blue = image->s.red; } cur->s.red = (ialpha * icolor.s.red + balpha * bcolor.s.red)/255; cur->s.green = (ialpha * icolor.s.green + balpha * bcolor.s.green)/255; cur->s.blue = (ialpha * icolor.s.blue + balpha * bcolor.s.blue)/255; cur->s.alpha = image->s.alpha; } } else { // All settings that use skincolors! UINT16 brightness; if (translen <= 0) { cur->rgba = image->rgba; goto skippixel; } // Don't bother with blending the pixel if the alpha of the blend pixel is 0 if (skinnum == TC_RAINBOW) { if (image->s.alpha == 0 && blendimage->s.alpha == 0) { cur->rgba = image->rgba; goto skippixel; } else { UINT16 imagebright, blendbright; SETBRIGHTNESS(imagebright,image->s.red,image->s.green,image->s.blue); SETBRIGHTNESS(blendbright,blendimage->s.red,blendimage->s.green,blendimage->s.blue); // slightly dumb average between the blend image color and base image colour, usually one or the other will be fully opaque anyway brightness = (imagebright*(255-blendimage->s.alpha))/255 + (blendbright*blendimage->s.alpha)/255; } } else { if (blendimage->s.alpha == 0) { cur->rgba = image->rgba; goto skippixel; // for metal sonic blend } else { SETBRIGHTNESS(brightness,blendimage->s.red,blendimage->s.green,blendimage->s.blue); } } // Calculate a sort of "gradient" for the skincolor // (Me splitting this into a function didn't work, so I had to ruin this entire function's groove...) { RGBA_t nextcolor; UINT8 firsti, secondi, mul, mulmax; INT32 r, g, b; // Rainbow needs to find the closest match to the textures themselves, instead of matching brightnesses to other colors. // Ensue horrible mess. if (skinnum == TC_RAINBOW) { UINT16 brightdif = 256; UINT8 colorbrightnesses[16]; INT32 compare, m, d; // Ignore pure white & pitch black if (brightness > 253 || brightness < 2) { cur->rgba = image->rgba; cur++; image++; blendimage++; continue; } firsti = 0; mul = 0; mulmax = 1; for (i = 0; i < translen; i++) { RGBA_t tempc = V_GetColor(translation[i]); SETBRIGHTNESS(colorbrightnesses[i], tempc.s.red, tempc.s.green, tempc.s.blue); // store brightnesses for comparison } for (i = 0; i < translen; i++) { if (brightness > colorbrightnesses[i]) // don't allow greater matches (because calculating a makeshift gradient for this is already a huge mess as is) continue; compare = abs((INT16)(colorbrightnesses[i]) - (INT16)(brightness)); if (compare < brightdif) { brightdif = (UINT16)compare; firsti = i; // best matching color that's equal brightness or darker } } secondi = firsti+1; // next color in line if (secondi >= translen) { m = (INT16)brightness; // - 0; d = (INT16)colorbrightnesses[firsti]; // - 0; } else { m = (INT16)brightness - (INT16)colorbrightnesses[secondi]; d = (INT16)colorbrightnesses[firsti] - (INT16)colorbrightnesses[secondi]; } if (m >= d) m = d-1; mulmax = 16; // calculate the "gradient" multiplier based on how close this color is to the one next in line if (m <= 0 || d <= 0) mul = 0; else mul = (mulmax-1) - ((m * mulmax) / d); } else { // Just convert brightness to a skincolor value, use distance to next position to find the gradient multipler firsti = 0; for (i = 1; i < translen; i++) { if (brightness >= cutoff[i]) break; firsti = i; } secondi = firsti+1; mulmax = cutoff[firsti]; if (secondi < translen) mulmax -= cutoff[secondi]; mul = cutoff[firsti] - brightness; } blendcolor = V_GetColor(translation[firsti]); if (secondi >= translen) mul = 0; if (mul > 0) // If it's 0, then we only need the first color. { #if 0 if (secondi >= translen) { // blend to black nextcolor = V_GetColor(31); } else #endif nextcolor = V_GetColor(translation[secondi]); // Find difference between points r = (INT32)(nextcolor.s.red - blendcolor.s.red); g = (INT32)(nextcolor.s.green - blendcolor.s.green); b = (INT32)(nextcolor.s.blue - blendcolor.s.blue); // Find the gradient of the two points r = ((mul * r) / mulmax); g = ((mul * g) / mulmax); b = ((mul * b) / mulmax); // Add gradient value to color blendcolor.s.red += r; blendcolor.s.green += g; blendcolor.s.blue += b; } } if (skinnum == TC_RAINBOW) { UINT32 tempcolor; UINT16 colorbright; SETBRIGHTNESS(colorbright,blendcolor.s.red,blendcolor.s.green,blendcolor.s.blue); if (colorbright == 0) colorbright = 1; // no dividing by 0 please tempcolor = (brightness * blendcolor.s.red) / colorbright; tempcolor = min(255, tempcolor); cur->s.red = (UINT8)tempcolor; tempcolor = (brightness * blendcolor.s.green) / colorbright; tempcolor = min(255, tempcolor); cur->s.green = (UINT8)tempcolor; tempcolor = (brightness * blendcolor.s.blue) / colorbright; tempcolor = min(255, tempcolor); cur->s.blue = (UINT8)tempcolor; cur->s.alpha = image->s.alpha; } else { // Color strength depends on image alpha INT32 tempcolor; tempcolor = ((image->s.red * (255-blendimage->s.alpha)) / 255) + ((blendcolor.s.red * blendimage->s.alpha) / 255); tempcolor = min(255, tempcolor); cur->s.red = (UINT8)tempcolor; tempcolor = ((image->s.green * (255-blendimage->s.alpha)) / 255) + ((blendcolor.s.green * blendimage->s.alpha) / 255); tempcolor = min(255, tempcolor); cur->s.green = (UINT8)tempcolor; tempcolor = ((image->s.blue * (255-blendimage->s.alpha)) / 255) + ((blendcolor.s.blue * blendimage->s.alpha) / 255); tempcolor = min(255, tempcolor); cur->s.blue = (UINT8)tempcolor; cur->s.alpha = image->s.alpha; } skippixel: // *Now* we can do Metal Sonic's flashing if (skinnum == TC_METALSONIC) { // Blend dark blue into white if (cur->s.alpha > 0 && cur->s.red == 0 && cur->s.green == 0 && cur->s.blue < 255 && cur->s.blue > 31) { // Sal: Invert non-blue cur->s.red = cur->s.green = (255 - cur->s.blue); cur->s.blue = 255; } cur->s.alpha = image->s.alpha; } else if (skinnum == TC_BOSS) { // Turn everything below a certain threshold white if ((image->s.red == image->s.green) && (image->s.green == image->s.blue) && image->s.blue < 127) { // Lactozilla: Invert the colors cur->s.red = cur->s.green = cur->s.blue = (255 - image->s.blue); } } } } cur++; image++; if (blendimage != NULL) blendimage++; } return; } #undef SETBRIGHTNESS static void HWR_GetBlendedTexture(patch_t *patch, patch_t *blendpatch, INT32 skinnum, const UINT8 *colormap, skincolornum_t color) { // mostly copied from HWR_GetMappedPatch, hence the similarities and comment GLPatch_t *grPatch = patch->hardware; GLPatch_t *grBlendPatch = NULL; GLMipmap_t *grMipmap, *newMipmap; if (blendpatch == NULL || colormap == colormaps || colormap == NULL) { // Don't do any blending HWD.pfnSetTexture(grPatch->mipmap); return; } if ((blendpatch && (grBlendPatch = blendpatch->hardware) && grBlendPatch->mipmap->format) && (patch->width != blendpatch->width || patch->height != blendpatch->height)) { // Blend image exists, but it's bad. HWD.pfnSetTexture(grPatch->mipmap); return; } // search for the mipmap // skip the first (no colormap translated) for (grMipmap = grPatch->mipmap; grMipmap->nextcolormap; ) { grMipmap = grMipmap->nextcolormap; if (grMipmap->colormap && grMipmap->colormap->source == colormap) { if (grMipmap->downloaded && grMipmap->data) { if (memcmp(grMipmap->colormap->data, colormap, 256 * sizeof(UINT8))) { M_Memcpy(grMipmap->colormap->data, colormap, 256 * sizeof(UINT8)); HWR_CreateBlendedTexture(patch, blendpatch, grMipmap, skinnum, color); HWD.pfnUpdateTexture(grMipmap); } else HWD.pfnSetTexture(grMipmap); // found the colormap, set it to the correct texture Z_ChangeTag(grMipmap->data, PU_HWRMODELTEXTURE_UNLOCKED); return; } } } // If here, the blended texture has not been created // So we create it //BP: WARNING: don't free it manually without clearing the cache of harware renderer // (it have a liste of mipmap) // this malloc is cleared in HWR_FreeColormapCache // (...) unfortunately z_malloc fragment alot the memory :(so malloc is better newMipmap = calloc(1, sizeof (*newMipmap)); if (newMipmap == NULL) I_Error("%s: Out of memory", "HWR_GetBlendedTexture"); grMipmap->nextcolormap = newMipmap; newMipmap->colormap = Z_Calloc(sizeof(*newMipmap->colormap), PU_HWRPATCHCOLMIPMAP, NULL); newMipmap->colormap->source = colormap; M_Memcpy(newMipmap->colormap->data, colormap, 256 * sizeof(UINT8)); HWR_CreateBlendedTexture(patch, blendpatch, newMipmap, skinnum, color); HWD.pfnSetTexture(newMipmap); Z_ChangeTag(newMipmap->data, PU_HWRMODELTEXTURE_UNLOCKED); } #define NORMALFOG 0x00000000 #define FADEFOG 0x19000000 static boolean HWR_AllowModel(mobj_t *mobj) { // Signpost overlay. Not needed. if (mobj->state-states == S_PLAY_SIGN) return false; // Otherwise, render the model. return true; } static boolean HWR_CanInterpolateModel(mobj_t *mobj, model_t *model) { if (cv_glmodelinterpolation.value == 2) // Always interpolate return true; return model->interpolate[(mobj->frame & FF_FRAMEMASK)]; } static boolean HWR_CanInterpolateSprite2(modelspr2frames_t *spr2frame) { if (cv_glmodelinterpolation.value == 2) // Always interpolate return true; return spr2frame->interpolate; } // // HWR_GetModelSprite2 (see P_GetSkinSprite2) // For non-super players, tries each sprite2's immediate predecessor until it finds one with a number of frames or ends up at standing. // For super players, does the same as above - but tries the super equivalent for each sprite2 before the non-super version. // static UINT8 HWR_GetModelSprite2(md2_t *md2, skin_t *skin, UINT8 spr2, player_t *player) { UINT8 super = 0, i = 0; if (!md2 || !md2->model || !md2->model->spr2frames || !skin) return 0; if ((playersprite_t)(spr2 & ~FF_SPR2SUPER) >= free_spr2) return 0; while (!md2->model->spr2frames[spr2].numframes && spr2 != SPR2_STND && ++i != 32) // recursion limiter { if (spr2 & FF_SPR2SUPER) { super = FF_SPR2SUPER; spr2 &= ~FF_SPR2SUPER; continue; } switch(spr2) { // Normal special cases. case SPR2_JUMP: spr2 = ((player ? player->charflags : skin->flags) & SF_NOJUMPSPIN) ? SPR2_SPNG : SPR2_ROLL; break; case SPR2_TIRE: spr2 = ((player ? player->charability : skin->ability) == CA_SWIM) ? SPR2_SWIM : SPR2_FLY; break; // Use the handy list, that's what it's there for! default: spr2 = spr2defaults[spr2]; break; } spr2 |= super; } if (i >= 32) // probably an infinite loop... return 0; return spr2; } // Adjust texture coords of model to fit into a patch's max_s and max_t static void adjustTextureCoords(model_t *model, patch_t *patch) { int i; GLPatch_t *gpatch = ((GLPatch_t *)patch->hardware); for (i = 0; i < model->numMeshes; i++) { int j; mesh_t *mesh = &model->meshes[i]; int numVertices; float *uvReadPtr = mesh->originaluvs; float *uvWritePtr; // i dont know if this is actually possible, just logical conclusion of structure in CreateModelVBOs if (!mesh->frames && !mesh->tinyframes) continue; if (mesh->frames) // again CreateModelVBO and CreateModelVBOTiny iterate like this so I'm gonna do that too numVertices = mesh->numTriangles * 3; else numVertices = mesh->numVertices; // if originaluvs points to uvs, we need to allocate new memory for adjusted uvs // the old uvs are kept around for use in possible readjustments if (mesh->uvs == mesh->originaluvs) mesh->uvs = Z_Malloc(numVertices * 2 * sizeof(float), PU_STATIC, NULL); uvWritePtr = mesh->uvs; // fix uvs (texture coordinates) to take into account that the actual texture // has empty space added until the next power of two for (j = 0; j < numVertices; j++) { *uvWritePtr++ = *uvReadPtr++ * gpatch->max_s; *uvWritePtr++ = *uvReadPtr++ * gpatch->max_t; } } // Save the values we adjusted the uvs for model->max_s = gpatch->max_s; model->max_t = gpatch->max_t; } // // HWR_DrawModel // boolean HWR_DrawModel(gl_vissprite_t *spr) { md2_t *md2; char filename[64]; INT32 frame = 0; INT32 nextFrame = -1; UINT8 spr2 = 0; FTransform p; FSurfaceInfo Surf; if (!cv_glmodels.value) return false; if (spr->precip) return false; // Lactozilla: Disallow certain models from rendering if (!HWR_AllowModel(spr->mobj)) return false; memset(&p, 0x00, sizeof(FTransform)); // MD2 colormap fix // colormap test if (spr->mobj->subsector) { sector_t *sector = spr->mobj->subsector->sector; UINT8 lightlevel = 255; extracolormap_t *colormap = NULL; if (sector->numlights) { INT32 light; light = R_GetPlaneLight(sector, spr->mobj->z + spr->mobj->height, false); // Always use the light at the top instead of whatever I was doing before if (R_ThingIsFullDark(spr->mobj)) lightlevel = 0; else if (R_ThingIsSemiBright(spr->mobj)) lightlevel = 128 + (*sector->lightlist[light].lightlevel>>1); else if (!R_ThingIsFullBright(spr->mobj)) lightlevel = *sector->lightlist[light].lightlevel > 255 ? 255 : *sector->lightlist[light].lightlevel; if (*sector->lightlist[light].extra_colormap) colormap = *sector->lightlist[light].extra_colormap; } else { if (R_ThingIsFullDark(spr->mobj)) lightlevel = 0; else if (R_ThingIsSemiBright(spr->mobj)) lightlevel = 128 + (sector->lightlevel>>1); else if (!R_ThingIsFullBright(spr->mobj)) lightlevel = sector->lightlevel > 255 ? 255 : sector->lightlevel; if (sector->extra_colormap) colormap = sector->extra_colormap; } HWR_Lighting(&Surf, lightlevel, colormap); } else Surf.PolyColor.rgba = 0xFFFFFFFF; // Look at HWR_ProjectSprite for more { patch_t *gpatch, *blendgpatch; GLPatch_t *hwrPatch = NULL, *hwrBlendPatch = NULL; float durs = (float)spr->mobj->state->tics; float tics = (float)spr->mobj->tics; const boolean papersprite = (R_ThingIsPaperSprite(spr->mobj) && !R_ThingIsFloorSprite(spr->mobj)); const UINT8 flip = (UINT8)(!(spr->mobj->eflags & MFE_VERTICALFLIP) != !R_ThingVerticallyFlipped(spr->mobj)); const UINT8 hflip = (UINT8)(!(spr->mobj->mirrored) != !R_ThingHorizontallyFlipped(spr->mobj)); spritedef_t *sprdef; spriteframe_t *sprframe; spriteinfo_t *sprinfo; angle_t ang; INT32 mod; float finalscale; interpmobjstate_t interp; if (R_UsingFrameInterpolation() && !paused) { R_InterpolateMobjState(spr->mobj, rendertimefrac, &interp); } else { R_InterpolateMobjState(spr->mobj, FRACUNIT, &interp); } // Apparently people don't like jump frames like that, so back it goes //if (tics > durs) //durs = tics; INT32 blendmode; if (spr->mobj->frame & FF_BLENDMASK) blendmode = ((spr->mobj->frame & FF_BLENDMASK) >> FF_BLENDSHIFT) + 1; else blendmode = spr->mobj->blendmode; if (spr->mobj->frame & FF_TRANSMASK) Surf.PolyFlags = HWR_SurfaceBlend(blendmode, (spr->mobj->frame & FF_TRANSMASK)>>FF_TRANSSHIFT, &Surf); else { Surf.PolyColor.s.alpha = (spr->mobj->flags2 & MF2_SHADOW) ? 0x40 : 0xff; Surf.PolyFlags = HWR_GetBlendModeFlag(blendmode); } // don't forget to enable the depth test because we can't do this // like before: model polygons are not sorted // 1. load model+texture if not already loaded // 2. draw model with correct position, rotation,... if (spr->mobj->skin && spr->mobj->sprite == SPR_PLAY) // Use the player MD2 list if the mobj has a skin and is using the player sprites { md2 = &md2_playermodels[(skin_t*)spr->mobj->skin-skins]; md2->skin = (skin_t*)spr->mobj->skin-skins; sprinfo = &((skin_t *)spr->mobj->skin)->sprinfo[spr->mobj->sprite2]; } else { md2 = &md2_models[spr->mobj->sprite]; sprinfo = &spriteinfo[spr->mobj->sprite]; } // texture loading before model init, so it knows if sprite graphics are used, which // means that texture coordinates have to be adjusted gpatch = md2->grpatch; if (gpatch) hwrPatch = ((GLPatch_t *)gpatch->hardware); if (!gpatch || !hwrPatch || ((!hwrPatch->mipmap->format || !hwrPatch->mipmap->downloaded) && !md2->notexturefile)) md2_loadTexture(md2); // Load it again, because it isn't being loaded into gpatch after md2_loadtexture... gpatch = md2->grpatch; if (gpatch) hwrPatch = ((GLPatch_t *)gpatch->hardware); // Load blend texture blendgpatch = md2->blendgrpatch; if (blendgpatch) hwrBlendPatch = ((GLPatch_t *)blendgpatch->hardware); if ((gpatch && hwrPatch && hwrPatch->mipmap->format) // don't load the blend texture if the base texture isn't available && (!blendgpatch || !hwrBlendPatch || ((!hwrBlendPatch->mipmap->format || !hwrBlendPatch->mipmap->downloaded) && !md2->noblendfile))) md2_loadBlendTexture(md2); // Load it again, because it isn't being loaded into blendgpatch after md2_loadblendtexture... blendgpatch = md2->blendgrpatch; if (blendgpatch) hwrBlendPatch = ((GLPatch_t *)blendgpatch->hardware); if (md2->error) return false; // we already failed loading this before :( if (!md2->model) { //CONS_Debug(DBG_RENDER, "Loading model... (%s)", sprnames[spr->mobj->sprite]); sprintf(filename, "models/%s", md2->filename); md2->model = md2_readModel(filename); if (md2->model) { md2_printModelInfo(md2->model); // If model uses sprite patch as texture, then // adjust texture coordinates to take power of two textures into account if (!gpatch || !hwrPatch->mipmap->format) adjustTextureCoords(md2->model, spr->gpatch); // note down the max_s and max_t that end up in the VBO md2->model->vbo_max_s = md2->model->max_s; md2->model->vbo_max_t = md2->model->max_t; HWD.pfnCreateModelVBOs(md2->model); } else { //CONS_Debug(DBG_RENDER, " FAILED\n"); md2->error = true; // prevent endless fail return false; } } //HWD.pfnSetBlend(blend); // This seems to actually break translucency? finalscale = md2->scale; //Hurdler: arf, I don't like that implementation at all... too much crappy if (gpatch && hwrPatch && hwrPatch->mipmap->format) // else if meant that if a texture couldn't be loaded, it would just end up using something else's texture { INT32 skinnum = TC_DEFAULT; if ((spr->mobj->flags & (MF_ENEMY|MF_BOSS)) && (spr->mobj->flags2 & MF2_FRET) && !(spr->mobj->flags & MF_GRENADEBOUNCE) && (leveltime & 1)) // Bosses "flash" { if (spr->mobj->type == MT_CYBRAKDEMON || spr->mobj->colorized) skinnum = TC_ALLWHITE; else if (spr->mobj->type == MT_METALSONIC_BATTLE) skinnum = TC_METALSONIC; else skinnum = TC_BOSS; } else if ((skincolornum_t)spr->mobj->color != SKINCOLOR_NONE) { if (spr->mobj->colorized) skinnum = TC_RAINBOW; else if (spr->mobj->player && spr->mobj->player->dashmode >= DASHMODE_THRESHOLD && (spr->mobj->player->charflags & SF_DASHMODE) && ((leveltime/2) & 1)) { if (spr->mobj->player->charflags & SF_MACHINE) skinnum = TC_DASHMODE; else skinnum = TC_RAINBOW; } else if (spr->mobj->skin && spr->mobj->sprite == SPR_PLAY) skinnum = (INT32)((skin_t*)spr->mobj->skin-skins); else skinnum = TC_DEFAULT; } // Translation or skin number found HWR_GetBlendedTexture(gpatch, blendgpatch, skinnum, spr->colormap, (skincolornum_t)spr->mobj->color); } else // Sprite { // Check if sprite dimensions are different from previously used sprite. // If so, uvs need to be readjusted. // Comparing floats with the != operator here should be okay because they // are just copies of glpatches' max_s and max_t values. // Instead of the != operator, memcmp is used to avoid a compiler warning. if (memcmp(&(hwrPatch->max_s), &(md2->model->max_s), sizeof(md2->model->max_s)) != 0 || memcmp(&(hwrPatch->max_t), &(md2->model->max_t), sizeof(md2->model->max_t)) != 0) adjustTextureCoords(md2->model, spr->gpatch); HWR_GetMappedPatch(spr->gpatch, spr->colormap); } if (spr->mobj->frame & FF_ANIMATE) { // set duration and tics to be the correct values for FF_ANIMATE states durs = (float)spr->mobj->state->var2; tics = (float)spr->mobj->anim_duration; } frame = (spr->mobj->frame & FF_FRAMEMASK); if (spr->mobj->skin && spr->mobj->sprite == SPR_PLAY && md2->model->spr2frames) { spr2 = HWR_GetModelSprite2(md2, spr->mobj->skin, spr->mobj->sprite2, spr->mobj->player); mod = md2->model->spr2frames[spr2].numframes; #ifndef DONTHIDEDIFFANIMLENGTH // by default, different anim length is masked by the mod if (mod > (INT32)((skin_t *)spr->mobj->skin)->sprites[spr2].numframes) mod = ((skin_t *)spr->mobj->skin)->sprites[spr2].numframes; #endif if (!mod) mod = 1; frame = md2->model->spr2frames[spr2].frames[frame%mod]; } else { mod = md2->model->meshes[0].numFrames; if (!mod) mod = 1; } #ifdef USE_MODEL_NEXTFRAME // Interpolate the model interpolation. (lol) tics -= FixedToFloat(rendertimefrac); #define INTERPOLERATION_LIMIT (TICRATE * 0.25f) if (cv_glmodelinterpolation.value && tics <= durs && tics <= INTERPOLERATION_LIMIT) { if (durs > INTERPOLERATION_LIMIT) durs = INTERPOLERATION_LIMIT; if (spr->mobj->skin && spr->mobj->sprite == SPR_PLAY && md2->model->spr2frames) { if (HWR_CanInterpolateSprite2(&md2->model->spr2frames[spr2]) && (spr->mobj->frame & FF_ANIMATE || (spr->mobj->state->nextstate != S_NULL && states[spr->mobj->state->nextstate].sprite == SPR_PLAY && ((P_GetSkinSprite2(spr->mobj->skin, (((spr->mobj->player && spr->mobj->player->powers[pw_super]) ? FF_SPR2SUPER : 0)|states[spr->mobj->state->nextstate].frame) & FF_FRAMEMASK, spr->mobj->player) == spr->mobj->sprite2))))) { nextFrame = (spr->mobj->frame & FF_FRAMEMASK) + 1; if (nextFrame >= mod) { if (spr->mobj->state->frame & FF_SPR2ENDSTATE) nextFrame--; else nextFrame = 0; } if (frame || !(spr->mobj->state->frame & FF_SPR2ENDSTATE)) nextFrame = md2->model->spr2frames[spr2].frames[nextFrame]; else nextFrame = -1; } } else if (HWR_CanInterpolateModel(spr->mobj, md2->model)) { // frames are handled differently for states with FF_ANIMATE, so get the next frame differently for the interpolation if (spr->mobj->frame & FF_ANIMATE) { nextFrame = (spr->mobj->frame & FF_FRAMEMASK) + 1; if (nextFrame >= (INT32)(spr->mobj->state->var1 + (spr->mobj->state->frame & FF_FRAMEMASK))) nextFrame = (spr->mobj->state->frame & FF_FRAMEMASK) % mod; } else { if (spr->mobj->state->nextstate != S_NULL && states[spr->mobj->state->nextstate].sprite != SPR_NULL && !(spr->mobj->player && (spr->mobj->state->nextstate == S_PLAY_WAIT) && spr->mobj->state == &states[S_PLAY_STND])) nextFrame = (states[spr->mobj->state->nextstate].frame & FF_FRAMEMASK) % mod; } } } #undef INTERPOLERATION_LIMIT #endif //Hurdler: it seems there is still a small problem with mobj angle p.x = FIXED_TO_FLOAT(interp.x); p.y = FIXED_TO_FLOAT(interp.y)+md2->offset; if (flip) p.z = FIXED_TO_FLOAT(interp.z + spr->mobj->height); else p.z = FIXED_TO_FLOAT(interp.z); if (spr->mobj->skin && spr->mobj->sprite == SPR_PLAY) sprdef = &((skin_t *)spr->mobj->skin)->sprites[spr->mobj->sprite2]; else sprdef = &sprites[spr->mobj->sprite]; sprframe = &sprdef->spriteframes[spr->mobj->frame & FF_FRAMEMASK]; if (sprframe->rotate || papersprite) { fixed_t anglef = AngleFixed(interp.angle); p.angley = FIXED_TO_FLOAT(anglef); } else { const fixed_t anglef = AngleFixed((R_PointToAngle(interp.x, interp.y))-ANGLE_180); p.angley = FIXED_TO_FLOAT(anglef); } p.rollangle = 0.0f; p.rollflip = 1; p.rotaxis = 0; if (spr->mobj->rollangle) { fixed_t anglef = AngleFixed(spr->mobj->rollangle); p.rollangle = FIXED_TO_FLOAT(anglef); p.roll = true; // rotation pivot p.centerx = FIXED_TO_FLOAT(spr->mobj->radius/2); p.centery = FIXED_TO_FLOAT(spr->mobj->height/(flip ? -2 : 2)); // rotation axis if (sprinfo->available) p.rotaxis = (UINT8)(sprinfo->pivot[(spr->mobj->frame & FF_FRAMEMASK)].rotaxis); // for NiGHTS specifically but should work everywhere else ang = R_PointToAngle (interp.x, interp.y) - interp.angle; if ((sprframe->rotate & SRF_RIGHT) && (ang < ANGLE_180)) // See from right p.rollflip = 1; else if ((sprframe->rotate & SRF_LEFT) && (ang >= ANGLE_180)) // See from left p.rollflip = -1; if (flip) p.rollflip *= -1; } p.anglex = 0.0f; #ifdef USE_FTRANSFORM_ANGLEZ // Slope rotation from Kart p.anglez = 0.0f; if (spr->mobj->standingslope) { fixed_t tempz = spr->mobj->standingslope->normal.z; fixed_t tempy = spr->mobj->standingslope->normal.y; fixed_t tempx = spr->mobj->standingslope->normal.x; fixed_t tempangle = AngleFixed(R_PointToAngle2(0, 0, FixedSqrt(FixedMul(tempy, tempy) + FixedMul(tempz, tempz)), tempx)); p.anglez = FIXED_TO_FLOAT(tempangle); tempangle = -AngleFixed(R_PointToAngle2(0, 0, tempz, tempy)); p.anglex = FIXED_TO_FLOAT(tempangle); } #endif // SRB2CBTODO: MD2 scaling support finalscale *= FIXED_TO_FLOAT(spr->mobj->scale); p.flip = atransform.flip; #ifdef USE_FTRANSFORM_MIRROR p.mirror = atransform.mirror; // from Kart #endif HWD.pfnSetShader(SHADER_MODEL); // model shader HWD.pfnDrawModel(md2->model, frame, durs, tics, nextFrame, &p, finalscale, flip, hflip, &Surf); } return true; } #endif //HWRENDER