- NOTE: In this revision (possibly earlier), I've started to notice some intermittent geometry glitches in Spikeout and slowdown while playing certain characters.

- Added multi-texture fragment shader to repo.
- Added a multiTexture option (enabled by default) to use multi-texturing to decode textures.
- Added some comments regarding timing to the PowerPC execution loop.

Src/CPU/PowerPC/ppc603.c

@@ -252,6 +252,7 @@ void ppc_reset(void)
 int ppc_execute(int cycles)
 {
 	UINT32 opcode;
+	
 	ppc_icount = cycles;
 	ppc_tb_base_icount = cycles;
 	ppc_dec_base_icount = cycles + ppc.dec_frac;
@@ -290,8 +291,13 @@ int ppc_execute(int cycles)
 		ppc.pc = ppc.npc;
 		
 		// Debug breakpoints
-		//if (ppc.pc == 0x9e4d4)
+		/*
-		//	printf("%X R0=%08X\n", ppc.pc, REG(0));
+		if (ppc.pc == 0x9d40)
+		{
+			printf("%X R3=%08X R4=%08X\n", ppc.pc, REG(3), REG(4));			
+			
+		}
+		*/
 			
 		opcode = *ppc.op++;	// Supermodel byte reverses each aligned word (converting them to little endian) so they can be fetched directly
 		ppc.npc = ppc.pc + 4;
@@ -314,6 +320,9 @@ int ppc_execute(int cycles)
 		}
 
 		ppc_icount--;
+		
+		// Updating TB four times per core cycle fixes VF3 timing  but breaks other games (Daytona 2 too fast, Spikeout has some geometry flickering)
+		//ppc.tb += 4;
 
 		if(ppc_icount == ppc_dec_trigger_cycle)
 		{
@@ -332,6 +341,9 @@ int ppc_execute(int cycles)
 
 	// update timebase
 	// timebase is incremented once every four core clock cycles, so adjust the cycles accordingly
+	// NOTE: updating at the end of the time slice breaks things that try to wait on TBL. Performing
+	// the update inside the execution loop fixes VF3 and allows many decrementer patches to be 
+	// removed but it adversely affects Spikeout and other games.
 	ppc.tb += ((ppc_tb_base_icount - ppc_icount) / 4);
 
 	// update decrementer

Src/Graphics/Render3D.cpp

@@ -1110,17 +1110,25 @@ bool CRender3D::Init(unsigned xOffset, unsigned yOffset, unsigned xRes, unsigned
     totalXRes = totalXResParam;
     totalYRes = totalYResParam;
 
-	// Load shaders
-	const char *vsFile = g_Config.vertexShaderFile.size() ? g_Config.vertexShaderFile.c_str() : NULL;
-	const char *fsFile = g_Config.fragmentShaderFile.size() ? g_Config.fragmentShaderFile.c_str() : NULL;
-	if (OKAY != LoadShaderProgram(&shaderProgram,&vertexShader,&fragmentShader,vsFile,fsFile,vertexShaderSource,fragmentShaderSource))
-		return FAIL;
-	
 	// Query max number of texture units supported by video card to use as upper limit
 	GLint glMaxTexUnits;
 	glGetIntegerv(GL_MAX_TEXTURE_IMAGE_UNITS, &glMaxTexUnits);
 	unsigned maxTexUnits = max<int>(1, min<int>(g_Config.maxTexUnits, glMaxTexUnits));
 	
+	// Load shaders. Use multi-sheet shader if requested and possible.
+	const char *vsFile = g_Config.vertexShaderFile.size() ? g_Config.vertexShaderFile.c_str() : NULL;
+	const char *fsFile = g_Config.fragmentShaderFile.size() ? g_Config.fragmentShaderFile.c_str() : NULL;
+	const char *fragmentShaderSource = fragmentShaderSingleSheetSource;	// single texture shader
+	if (g_Config.multiTexture)
+	{ 
+		if (maxTexUnits >= 8)	// can we use the multi-sheet shader?
+			fragmentShaderSource = fragmentShaderMultiSheetSource;
+		else
+			ErrorLog("Your system has too few texture units. Reverting to single texture shader.");
+	}
+	if (OKAY != LoadShaderProgram(&shaderProgram,&vertexShader,&fragmentShader,vsFile,fsFile,vertexShaderSource,fragmentShaderSource))
+		return FAIL;
+	
 	// Try locating default "textureMap" uniform in shader program
 	glUseProgram(shaderProgram); // bind program
 	textureMapLoc = glGetUniformLocation(shaderProgram, "textureMap");
@@ -1157,7 +1165,7 @@ bool CRender3D::Init(unsigned xOffset, unsigned yOffset, unsigned xRes, unsigned
 
 	// Check located at least one uniform to bind to a texture unit
 	if (numTexUnits == 0)
-		return ErrorLog("Could not locate any textureMap uniforms in shader script.");
+		return ErrorLog("Fragment shader must contain at least one 'textureMap' uniform.");
 	InfoLog("Located and bound %u uniform(s) in GL shader script to %u texture unit(s).", numTexUnits, unitCount);
 
 	// Now try creating texture sheets, one for each texture unit (memory permitting)

Src/Graphics/Render3D.h

@@ -175,15 +175,17 @@ struct ModelCache
 class CRender3DConfig
 {
 public:
-	string vertexShaderFile;	// path to vertex shader or "" to use internal shader
+	string 		vertexShaderFile;	// path to vertex shader or "" to use internal shader
-	string fragmentShaderFile;	// fragment shader
+	string 		fragmentShaderFile;	// fragment shader
-	unsigned maxTexUnits;       // maximum number of texture units to use (1-9)
+	unsigned	maxTexUnits;       	// maximum number of texture units to use (1-9)
+	bool		multiTexture;		// if enabled and no external fragment shader, select internal shader w/ multiple texture sheet support
 
 	// Defaults
 	CRender3DConfig(void)
 	{
 		// strings will be clear to begin with
 		maxTexUnits = 9;
+		multiTexture = true;
 	}
 };
 

Src/Graphics/Shaders/Fragment_MultiSheet.glsl

@@ -0,0 +1,237 @@
+/**
+ ** Supermodel
+ ** A Sega Model 3 Arcade Emulator.
+ ** Copyright 2011 Bart Trzynadlowski, Nik Henson 
+ **
+ ** This file is part of Supermodel.
+ **
+ ** Supermodel is free software: you can redistribute it and/or modify it under
+ ** the terms of the GNU General Public License as published by the Free 
+ ** Software Foundation, either version 3 of the License, or (at your option)
+ ** any later version.
+ **
+ ** Supermodel is distributed in the hope that it will be useful, but WITHOUT
+ ** ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ ** FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
+ ** more details.
+ **
+ ** You should have received a copy of the GNU General Public License along
+ ** with Supermodel.  If not, see <http://www.gnu.org/licenses/>.
+ **/
+ 
+/*
+ * Fragment_MultiSheet.glsl
+ *
+ * Fragment shader for 3D rendering. Uses 8 texture sheets to decode the 
+ * different possible formats.
+ */
+
+#version 120
+
+// Global uniforms
+uniform sampler2D	textureMap0;		// complete texture map (fmt 0), 2048x2048 texels
+uniform sampler2D	textureMap1;		// complete texture map (fmt 1), 2048x2048 texels
+uniform sampler2D	textureMap2;		// complete texture map (fmt 2), 2048x2048 texels
+uniform sampler2D	textureMap3;		// complete texture map (fmt 3), 2048x2048 texels
+uniform sampler2D	textureMap4;		// complete texture map (fmt 4), 2048x2048 texels
+uniform sampler2D	textureMap5;		// complete texture map (fmt 5), 2048x2048 texels
+uniform sampler2D	textureMap6;		// complete texture map (fmt 6), 2048x2048 texels
+uniform sampler2D	textureMap7;		// complete texture map (fmt 7), 2048x2048 texels
+uniform vec4		spotEllipse;		// spotlight ellipse position: .x=X position (screen coordinates), .y=Y position, .z=half-width, .w=half-height)
+uniform vec2		spotRange;			// spotlight Z range: .x=start (viewspace coordinates), .y=limit
+uniform vec3		spotColor;			// spotlight RGB color
+uniform vec3		lighting[2];		// lighting state (lighting[0] = sun direction, lighting[1].x,y = diffuse, ambient intensities from 0-1.0)
+
+// Inputs from vertex shader 
+varying vec4		fsSubTexture;	// .x=texture X, .y=texture Y, .z=texture width, .w=texture height (all in texels)
+varying vec4		fsTexParams;	// .x=texture enable (if 1, else 0), .y=use transparency (if > 0), .z=U wrap mode (1=mirror, 0=repeat), .w=V wrap mode
+varying float		fsTexFormat;	// T1RGB5 contour texture (if > 0)
+varying float		fsTransLevel;	// translucence level, 0.0 (transparent) to 1.0 (opaque)
+varying vec3		fsLightIntensity;	// lighting intensity 
+varying float		fsSpecularTerm;	// specular highlight
+varying float		fsFogFactor;	// fog factor
+varying float		fsViewZ;		// Z distance to fragment from viewpoint at origin
+
+/*
+ * WrapTexelCoords():
+ *
+ * Computes the normalized OpenGL S,T coordinates within the 2048x2048 texture
+ * sheet, taking into account wrapping behavior.
+ *
+ * Computing normalized OpenGL texture coordinates (0 to 1) within the 
+ * Real3D texture sheet:
+ *
+ * If the texture is not mirrored, we simply have to clamp the
+ * coordinates to fit within the texture dimensions, add the texture
+ * X, Y position to select the appropriate one, and normalize by 2048
+ * (the dimensions of the Real3D texture sheet).
+ *
+ *		= [(u,v)%(w,h)+(x,y)]/(2048,2048)
+ *
+ * If mirroring is enabled, textures are mirrored every odd multiple of
+ * the original texture. To detect whether we are in an odd multiple, 
+ * simply divide the coordinate by the texture dimension and check 
+ * whether the result is odd. Then, clamp the coordinates as before but
+ * subtract from the last texel to mirror them:
+ *
+ * 		= [M*((w-1,h-1)-(u,v)%(w,h)) + (1-M)*(u,v)%(w,h) + (x,y)]/(2048,2048)
+ *		where M is 1.0 if the texture must be mirrored.
+ *
+ * As an optimization, this function computes TWO texture coordinates
+ * simultaneously. The first is texCoord.xy, the second is in .zw. The other
+ * parameters must have .xy = .zw.
+ */
+vec4 WrapTexelCoords(vec4 texCoord, vec4 texOffset, vec4 texSize, vec4 mirrorEnable)
+{
+	vec4	clampedCoord, mirror, glTexCoord;
+	
+	clampedCoord = mod(texCoord,texSize);						// clamp coordinates to within texture size
+	mirror = mirrorEnable * mod(floor(texCoord/texSize),2.0);	// whether this texel needs to be mirrored
+
+	glTexCoord = (	mirror*(texSize-clampedCoord) +
+					(vec4(1.0,1.0,1.0,1.0)-mirror)*clampedCoord +
+					texOffset
+				 ) / 2048.0;
+	return glTexCoord;
+}
+
+/*
+ * main():
+ *
+ * Fragment shader entry point.
+ */
+
+void main(void)
+{	
+	vec4	uv_top, uv_bot, c[4];
+	vec2	r;
+	vec4	fragColor;
+	vec2	ellipse;
+	vec3	lightIntensity;
+	float	insideSpot;
+	int		x;
+	
+	// Get polygon color for untextured polygons (textured polygons will overwrite)
+	if (fsTexParams.x < 0.5)
+		fragColor = gl_Color;		
+	else
+	// Textured polygons: set fragment color to texel value
+	{			
+		/*
+		 * Bilinear Filtering
+		 *
+		 * In order to get this working on ATI, the number of operations is
+		 * reduced by putting everything into vec4s. uv_top holds the UV 
+		 * coordinates for the top two texels (.xy=left, .zw=right) and uv_bot
+		 * is for the lower two.
+		 */
+
+		// Compute fractional blending factor, r, and lower left corner of texel 0
+		uv_bot.xy = gl_TexCoord[0].st-vec2(0.5,0.5);	// move into the lower left blending texel 
+		r = uv_bot.xy-floor(uv_bot.xy);					// fractional part
+		uv_bot.xy = floor(uv_bot.xy);					// integral part
+		
+		// Compute texel coordinates
+		uv_bot.xy += vec2(0.5,0.5);	// offset to center of pixel (should not be needed but it fixes a lot of glitches, esp. on Nvidia)
+		uv_bot.zw = uv_bot.xy + vec2(1.0,0.0);			// compute coordinates of the other three neighbors
+		uv_top = uv_bot + vec4(0.0,1.0,0.0,1.0);
+
+		// Compute the properly wrapped texel coordinates
+		uv_top = WrapTexelCoords(uv_top,vec4(fsSubTexture.xy,fsSubTexture.xy),vec4(fsSubTexture.zw,fsSubTexture.zw), vec4(fsTexParams.zw,fsTexParams.zw));
+		uv_bot = WrapTexelCoords(uv_bot,vec4(fsSubTexture.xy,fsSubTexture.xy),vec4(fsSubTexture.zw,fsSubTexture.zw), vec4(fsTexParams.zw,fsTexParams.zw));
+
+		// Fetch the texels from the texture map that corresponds to the current texture format
+		if (fsTexFormat < 0.5f)	{
+			c[0]=texture2D(textureMap0, uv_bot.xy);       // bottom-left (base texel)
+			c[1]=texture2D(textureMap0, uv_bot.zw);       // bottom-right
+			c[2]=texture2D(textureMap0, uv_top.xy);       // top-left
+			c[3]=texture2D(textureMap0, uv_top.zw);       // top-right
+		} else if (fsTexFormat < 1.5f) {
+            c[0]=texture2D(textureMap1, uv_bot.xy);       // bottom-left (base texel)
+			c[1]=texture2D(textureMap1, uv_bot.zw);       // bottom-right
+			c[2]=texture2D(textureMap1, uv_top.xy);       // top-left
+			c[3]=texture2D(textureMap1, uv_top.zw);       // top-right
+		} else if (fsTexFormat < 2.5f) {
+            c[0]=texture2D(textureMap2, uv_bot.xy);       // bottom-left (base texel)
+			c[1]=texture2D(textureMap2, uv_bot.zw);       // bottom-right
+			c[2]=texture2D(textureMap2, uv_top.xy);       // top-left
+			c[3]=texture2D(textureMap2, uv_top.zw);       // top-right
+		} else if (fsTexFormat < 3.5f) {
+            c[0]=texture2D(textureMap3, uv_bot.xy);       // bottom-left (base texel)
+			c[1]=texture2D(textureMap3, uv_bot.zw);       // bottom-right
+			c[2]=texture2D(textureMap3, uv_top.xy);       // top-left
+			c[3]=texture2D(textureMap3, uv_top.zw);       // top-right
+		} else if (fsTexFormat < 4.5f) {
+            c[0]=texture2D(textureMap4, uv_bot.xy);       // bottom-left (base texel)
+			c[1]=texture2D(textureMap4, uv_bot.zw);       // bottom-right
+			c[2]=texture2D(textureMap4, uv_top.xy);       // top-left
+			c[3]=texture2D(textureMap4, uv_top.zw);       // top-right
+		} else if (fsTexFormat < 5.5f) {
+            c[0]=texture2D(textureMap5, uv_bot.xy);       // bottom-left (base texel)
+			c[1]=texture2D(textureMap5, uv_bot.zw);       // bottom-right
+			c[2]=texture2D(textureMap5, uv_top.xy);       // top-left
+			c[3]=texture2D(textureMap5, uv_top.zw);       // top-right
+		} else if (fsTexFormat < 6.5f) {
+			c[0]=texture2D(textureMap6, uv_bot.xy);       // bottom-left (base texel)
+			c[1]=texture2D(textureMap6, uv_bot.zw);       // bottom-right
+			c[2]=texture2D(textureMap6, uv_top.xy);       // top-left
+			c[3]=texture2D(textureMap6, uv_top.zw);       // top-right
+		} else {
+            c[0]=texture2D(textureMap7, uv_bot.xy);       // bottom-left (base texel)
+			c[1]=texture2D(textureMap7, uv_bot.zw);       // bottom-right
+			c[2]=texture2D(textureMap7, uv_top.xy);       // top-left
+			c[3]=texture2D(textureMap7, uv_top.zw);       // top-right
+		}                      
+
+		// Interpolate texels and blend result with material color to determine final (unlit) fragment color
+		// fragColor = (c[0]*(1.0-r.s)*(1.0-r.t) + c[1]*r.s*(1.0-r.t) + c[2]*(1.0-r.s)*r.t + c[3]*r.s*r.t);
+		// Faster method:
+		c[0] += (c[1]-c[0])*r.s;			// 2 alu
+		c[2] += (c[3]-c[2])*r.s;			// 2 alu
+		fragColor = c[0]+(c[2]-c[0])*r.t;	// 2 alu
+	
+		/*
+		 * T1RGB5:
+		 *
+		 * The transparency bit determines whether to discard pixels (if set).
+		 * What is unknown is how this bit behaves when interpolated. OpenGL
+		 * processes it as an alpha value, so it might concievably be blended
+		 * with neighbors. Here, an arbitrary threshold is chosen.
+		 *
+		 * To-do: blending could probably enabled and this would work even
+		 * better with a hard threshold.
+		 *
+		 * Countour processing also seems to be enabled for RGBA4 textures.
+		 * When the alpha value is 0.0 (or close), pixels are discarded 
+		 * entirely.
+		 */
+		if (fsTexParams.y > 0.5)	// contour processing enabled
+		{
+			if (fragColor.a < 0.01)	// discard anything with alpha == 0
+				discard;
+		}
+		
+		// If contour texture and not discarded, force alpha to 1.0 because will later be modified by polygon translucency
+		if (fsTexFormat < 0.5)		// contour (T1RGB5) texture map
+			fragColor.a = 1.0;
+	}
+
+	// Compute spotlight and apply lighting
+	ellipse = (gl_FragCoord.xy-spotEllipse.xy)/spotEllipse.zw;
+	insideSpot = dot(ellipse,ellipse);
+	if ((insideSpot <= 1.0) &&  (fsViewZ>=spotRange.x) && (fsViewZ<spotRange.y))
+		lightIntensity = fsLightIntensity+(1.0-insideSpot)*spotColor;
+	else
+		lightIntensity = fsLightIntensity;
+	fragColor.rgb *= lightIntensity;
+	fragColor.rgb += vec3(fsSpecularTerm,fsSpecularTerm,fsSpecularTerm);
+	
+	// Translucency (modulates existing alpha channel for RGBA4 texels)
+	fragColor.a *= fsTransLevel;
+
+	// Apply fog under the control of fog factor setting from polygon header
+	fragColor.rgb = mix(gl_Fog.color.rgb, fragColor.rgb, fsFogFactor);
+
+	// Store final color
+	gl_FragColor = fragColor;
+}

Src/Graphics/Shaders3D.h

@@ -212,8 +212,8 @@ static const char vertexShaderSource[] =
 };
 	
 	
-// Fragment shader
+// Fragment shader (single texture sheet)
-static const char fragmentShaderSource[] = 
+static const char fragmentShaderSingleSheetSource[] = 
 {
 "/**\n"
 " ** Supermodel\n"
@@ -409,4 +409,248 @@ static const char fragmentShaderSource[] =
 "}\n"
 };
 
+
+// Fragment shader (8 texture sheets)
+static const char fragmentShaderMultiSheetSource[] = 
+{
+"/**\n"
+" ** Supermodel\n"
+" ** A Sega Model 3 Arcade Emulator.\n"
+" ** Copyright 2011 Bart Trzynadlowski, Nik Henson \n"
+" **\n"
+" ** This file is part of Supermodel.\n"
+" **\n"
+" ** Supermodel is free software: you can redistribute it and/or modify it under\n"
+" ** the terms of the GNU General Public License as published by the Free \n"
+" ** Software Foundation, either version 3 of the License, or (at your option)\n"
+" ** any later version.\n"
+" **\n"
+" ** Supermodel is distributed in the hope that it will be useful, but WITHOUT\n"
+" ** ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or\n"
+" ** FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for\n"
+" ** more details.\n"
+" **\n"
+" ** You should have received a copy of the GNU General Public License along\n"
+" ** with Supermodel.  If not, see <http://www.gnu.org/licenses/>.\n"
+" **/\n"
+" \n"
+"/*\n"
+" * Fragment_MultiSheet.glsl\n"
+" *\n"
+" * Fragment shader for 3D rendering. Uses 8 texture sheets to decode the \n"
+" * different possible formats.\n"
+" */\n"
+"\n"
+"#version 120\n"
+"\n"
+"// Global uniforms\n"
+"uniform sampler2D\ttextureMap0;\t\t// complete texture map (fmt 0), 2048x2048 texels\n"
+"uniform sampler2D\ttextureMap1;\t\t// complete texture map (fmt 1), 2048x2048 texels\n"
+"uniform sampler2D\ttextureMap2;\t\t// complete texture map (fmt 2), 2048x2048 texels\n"
+"uniform sampler2D\ttextureMap3;\t\t// complete texture map (fmt 3), 2048x2048 texels\n"
+"uniform sampler2D\ttextureMap4;\t\t// complete texture map (fmt 4), 2048x2048 texels\n"
+"uniform sampler2D\ttextureMap5;\t\t// complete texture map (fmt 5), 2048x2048 texels\n"
+"uniform sampler2D\ttextureMap6;\t\t// complete texture map (fmt 6), 2048x2048 texels\n"
+"uniform sampler2D\ttextureMap7;\t\t// complete texture map (fmt 7), 2048x2048 texels\n"
+"uniform vec4\t\tspotEllipse;\t\t// spotlight ellipse position: .x=X position (screen coordinates), .y=Y position, .z=half-width, .w=half-height)\n"
+"uniform vec2\t\tspotRange;\t\t\t// spotlight Z range: .x=start (viewspace coordinates), .y=limit\n"
+"uniform vec3\t\tspotColor;\t\t\t// spotlight RGB color\n"
+"uniform vec3\t\tlighting[2];\t\t// lighting state (lighting[0] = sun direction, lighting[1].x,y = diffuse, ambient intensities from 0-1.0)\n"
+"\n"
+"// Inputs from vertex shader \n"
+"varying vec4\t\tfsSubTexture;\t// .x=texture X, .y=texture Y, .z=texture width, .w=texture height (all in texels)\n"
+"varying vec4\t\tfsTexParams;\t// .x=texture enable (if 1, else 0), .y=use transparency (if > 0), .z=U wrap mode (1=mirror, 0=repeat), .w=V wrap mode\n"
+"varying float\t\tfsTexFormat;\t// T1RGB5 contour texture (if > 0)\n"
+"varying float\t\tfsTransLevel;\t// translucence level, 0.0 (transparent) to 1.0 (opaque)\n"
+"varying vec3\t\tfsLightIntensity;\t// lighting intensity \n"
+"varying float\t\tfsSpecularTerm;\t// specular highlight\n"
+"varying float\t\tfsFogFactor;\t// fog factor\n"
+"varying float\t\tfsViewZ;\t\t// Z distance to fragment from viewpoint at origin\n"
+"\n"
+"/*\n"
+" * WrapTexelCoords():\n"
+" *\n"
+" * Computes the normalized OpenGL S,T coordinates within the 2048x2048 texture\n"
+" * sheet, taking into account wrapping behavior.\n"
+" *\n"
+" * Computing normalized OpenGL texture coordinates (0 to 1) within the \n"
+" * Real3D texture sheet:\n"
+" *\n"
+" * If the texture is not mirrored, we simply have to clamp the\n"
+" * coordinates to fit within the texture dimensions, add the texture\n"
+" * X, Y position to select the appropriate one, and normalize by 2048\n"
+" * (the dimensions of the Real3D texture sheet).\n"
+" *\n"
+" *\t\t= [(u,v)%(w,h)+(x,y)]/(2048,2048)\n"
+" *\n"
+" * If mirroring is enabled, textures are mirrored every odd multiple of\n"
+" * the original texture. To detect whether we are in an odd multiple, \n"
+" * simply divide the coordinate by the texture dimension and check \n"
+" * whether the result is odd. Then, clamp the coordinates as before but\n"
+" * subtract from the last texel to mirror them:\n"
+" *\n"
+" * \t\t= [M*((w-1,h-1)-(u,v)%(w,h)) + (1-M)*(u,v)%(w,h) + (x,y)]/(2048,2048)\n"
+" *\t\twhere M is 1.0 if the texture must be mirrored.\n"
+" *\n"
+" * As an optimization, this function computes TWO texture coordinates\n"
+" * simultaneously. The first is texCoord.xy, the second is in .zw. The other\n"
+" * parameters must have .xy = .zw.\n"
+" */\n"
+"vec4 WrapTexelCoords(vec4 texCoord, vec4 texOffset, vec4 texSize, vec4 mirrorEnable)\n"
+"{\n"
+"\tvec4\tclampedCoord, mirror, glTexCoord;\n"
+"\t\n"
+"\tclampedCoord = mod(texCoord,texSize);\t\t\t\t\t\t// clamp coordinates to within texture size\n"
+"\tmirror = mirrorEnable * mod(floor(texCoord/texSize),2.0);\t// whether this texel needs to be mirrored\n"
+"\n"
+"\tglTexCoord = (\tmirror*(texSize-clampedCoord) +\n"
+"\t\t\t\t\t(vec4(1.0,1.0,1.0,1.0)-mirror)*clampedCoord +\n"
+"\t\t\t\t\ttexOffset\n"
+"\t\t\t\t ) / 2048.0;\n"
+"\treturn glTexCoord;\n"
+"}\n"
+"\n"
+"/*\n"
+" * main():\n"
+" *\n"
+" * Fragment shader entry point.\n"
+" */\n"
+"\n"
+"void main(void)\n"
+"{\t\n"
+"\tvec4\tuv_top, uv_bot, c[4];\n"
+"\tvec2\tr;\n"
+"\tvec4\tfragColor;\n"
+"\tvec2\tellipse;\n"
+"\tvec3\tlightIntensity;\n"
+"\tfloat\tinsideSpot;\n"
+"\tint\t\tx;\n"
+"\t\n"
+"\t// Get polygon color for untextured polygons (textured polygons will overwrite)\n"
+"\tif (fsTexParams.x < 0.5)\n"
+"\t\tfragColor = gl_Color;\t\t\n"
+"\telse\n"
+"\t// Textured polygons: set fragment color to texel value\n"
+"\t{\t\t\t\n"
+"\t\t/*\n"
+"\t\t * Bilinear Filtering\n"
+"\t\t *\n"
+"\t\t * In order to get this working on ATI, the number of operations is\n"
+"\t\t * reduced by putting everything into vec4s. uv_top holds the UV \n"
+"\t\t * coordinates for the top two texels (.xy=left, .zw=right) and uv_bot\n"
+"\t\t * is for the lower two.\n"
+"\t\t */\n"
+"\n"
+"\t\t// Compute fractional blending factor, r, and lower left corner of texel 0\n"
+"\t\tuv_bot.xy = gl_TexCoord[0].st-vec2(0.5,0.5);\t// move into the lower left blending texel \n"
+"\t\tr = uv_bot.xy-floor(uv_bot.xy);\t\t\t\t\t// fractional part\n"
+"\t\tuv_bot.xy = floor(uv_bot.xy);\t\t\t\t\t// integral part\n"
+"\t\t\n"
+"\t\t// Compute texel coordinates\n"
+"\t\tuv_bot.xy += vec2(0.5,0.5);\t// offset to center of pixel (should not be needed but it fixes a lot of glitches, esp. on Nvidia)\n"
+"\t\tuv_bot.zw = uv_bot.xy + vec2(1.0,0.0);\t\t\t// compute coordinates of the other three neighbors\n"
+"\t\tuv_top = uv_bot + vec4(0.0,1.0,0.0,1.0);\n"
+"\n"
+"\t\t// Compute the properly wrapped texel coordinates\n"
+"\t\tuv_top = WrapTexelCoords(uv_top,vec4(fsSubTexture.xy,fsSubTexture.xy),vec4(fsSubTexture.zw,fsSubTexture.zw), vec4(fsTexParams.zw,fsTexParams.zw));\n"
+"\t\tuv_bot = WrapTexelCoords(uv_bot,vec4(fsSubTexture.xy,fsSubTexture.xy),vec4(fsSubTexture.zw,fsSubTexture.zw), vec4(fsTexParams.zw,fsTexParams.zw));\n"
+"\n"
+"\t\t// Fetch the texels from the texture map that corresponds to the current texture format\n"
+"\t\tif (fsTexFormat < 0.5f)\t{\n"
+"\t\t\tc[0]=texture2D(textureMap0, uv_bot.xy);       // bottom-left (base texel)\n"
+"\t\t\tc[1]=texture2D(textureMap0, uv_bot.zw);       // bottom-right\n"
+"\t\t\tc[2]=texture2D(textureMap0, uv_top.xy);       // top-left\n"
+"\t\t\tc[3]=texture2D(textureMap0, uv_top.zw);       // top-right\n"
+"\t\t} else if (fsTexFormat < 1.5f) {\n"
+"            c[0]=texture2D(textureMap1, uv_bot.xy);       // bottom-left (base texel)\n"
+"\t\t\tc[1]=texture2D(textureMap1, uv_bot.zw);       // bottom-right\n"
+"\t\t\tc[2]=texture2D(textureMap1, uv_top.xy);       // top-left\n"
+"\t\t\tc[3]=texture2D(textureMap1, uv_top.zw);       // top-right\n"
+"\t\t} else if (fsTexFormat < 2.5f) {\n"
+"            c[0]=texture2D(textureMap2, uv_bot.xy);       // bottom-left (base texel)\n"
+"\t\t\tc[1]=texture2D(textureMap2, uv_bot.zw);       // bottom-right\n"
+"\t\t\tc[2]=texture2D(textureMap2, uv_top.xy);       // top-left\n"
+"\t\t\tc[3]=texture2D(textureMap2, uv_top.zw);       // top-right\n"
+"\t\t} else if (fsTexFormat < 3.5f) {\n"
+"            c[0]=texture2D(textureMap3, uv_bot.xy);       // bottom-left (base texel)\n"
+"\t\t\tc[1]=texture2D(textureMap3, uv_bot.zw);       // bottom-right\n"
+"\t\t\tc[2]=texture2D(textureMap3, uv_top.xy);       // top-left\n"
+"\t\t\tc[3]=texture2D(textureMap3, uv_top.zw);       // top-right\n"
+"\t\t} else if (fsTexFormat < 4.5f) {\n"
+"            c[0]=texture2D(textureMap4, uv_bot.xy);       // bottom-left (base texel)\n"
+"\t\t\tc[1]=texture2D(textureMap4, uv_bot.zw);       // bottom-right\n"
+"\t\t\tc[2]=texture2D(textureMap4, uv_top.xy);       // top-left\n"
+"\t\t\tc[3]=texture2D(textureMap4, uv_top.zw);       // top-right\n"
+"\t\t} else if (fsTexFormat < 5.5f) {\n"
+"            c[0]=texture2D(textureMap5, uv_bot.xy);       // bottom-left (base texel)\n"
+"\t\t\tc[1]=texture2D(textureMap5, uv_bot.zw);       // bottom-right\n"
+"\t\t\tc[2]=texture2D(textureMap5, uv_top.xy);       // top-left\n"
+"\t\t\tc[3]=texture2D(textureMap5, uv_top.zw);       // top-right\n"
+"\t\t} else if (fsTexFormat < 6.5f) {\n"
+"\t\t\tc[0]=texture2D(textureMap6, uv_bot.xy);       // bottom-left (base texel)\n"
+"\t\t\tc[1]=texture2D(textureMap6, uv_bot.zw);       // bottom-right\n"
+"\t\t\tc[2]=texture2D(textureMap6, uv_top.xy);       // top-left\n"
+"\t\t\tc[3]=texture2D(textureMap6, uv_top.zw);       // top-right\n"
+"\t\t} else {\n"
+"            c[0]=texture2D(textureMap7, uv_bot.xy);       // bottom-left (base texel)\n"
+"\t\t\tc[1]=texture2D(textureMap7, uv_bot.zw);       // bottom-right\n"
+"\t\t\tc[2]=texture2D(textureMap7, uv_top.xy);       // top-left\n"
+"\t\t\tc[3]=texture2D(textureMap7, uv_top.zw);       // top-right\n"
+"\t\t}                      \n"
+"\n"
+"\t\t// Interpolate texels and blend result with material color to determine final (unlit) fragment color\n"
+"\t\t// fragColor = (c[0]*(1.0-r.s)*(1.0-r.t) + c[1]*r.s*(1.0-r.t) + c[2]*(1.0-r.s)*r.t + c[3]*r.s*r.t);\n"
+"\t\t// Faster method:\n"
+"\t\tc[0] += (c[1]-c[0])*r.s;\t\t\t// 2 alu\n"
+"\t\tc[2] += (c[3]-c[2])*r.s;\t\t\t// 2 alu\n"
+"\t\tfragColor = c[0]+(c[2]-c[0])*r.t;\t// 2 alu\n"
+"\t\n"
+"\t\t/*\n"
+"\t\t * T1RGB5:\n"
+"\t\t *\n"
+"\t\t * The transparency bit determines whether to discard pixels (if set).\n"
+"\t\t * What is unknown is how this bit behaves when interpolated. OpenGL\n"
+"\t\t * processes it as an alpha value, so it might concievably be blended\n"
+"\t\t * with neighbors. Here, an arbitrary threshold is chosen.\n"
+"\t\t *\n"
+"\t\t * To-do: blending could probably enabled and this would work even\n"
+"\t\t * better with a hard threshold.\n"
+"\t\t *\n"
+"\t\t * Countour processing also seems to be enabled for RGBA4 textures.\n"
+"\t\t * When the alpha value is 0.0 (or close), pixels are discarded \n"
+"\t\t * entirely.\n"
+"\t\t */\n"
+"\t\tif (fsTexParams.y > 0.5)\t// contour processing enabled\n"
+"\t\t{\n"
+"\t\t\tif (fragColor.a < 0.01)\t// discard anything with alpha == 0\n"
+"\t\t\t\tdiscard;\n"
+"\t\t}\n"
+"\t\t\n"
+"\t\t// If contour texture and not discarded, force alpha to 1.0 because will later be modified by polygon translucency\n"
+"\t\tif (fsTexFormat < 0.5)\t\t// contour (T1RGB5) texture map\n"
+"\t\t\tfragColor.a = 1.0;\n"
+"\t}\n"
+"\n"
+"\t// Compute spotlight and apply lighting\n"
+"\tellipse = (gl_FragCoord.xy-spotEllipse.xy)/spotEllipse.zw;\n"
+"\tinsideSpot = dot(ellipse,ellipse);\n"
+"\tif ((insideSpot <= 1.0) &&  (fsViewZ>=spotRange.x) && (fsViewZ<spotRange.y))\n"
+"\t\tlightIntensity = fsLightIntensity+(1.0-insideSpot)*spotColor;\n"
+"\telse\n"
+"\t\tlightIntensity = fsLightIntensity;\n"
+"\tfragColor.rgb *= lightIntensity;\n"
+"\tfragColor.rgb += vec3(fsSpecularTerm,fsSpecularTerm,fsSpecularTerm);\n"
+"\t\n"
+"\t// Translucency (modulates existing alpha channel for RGBA4 texels)\n"
+"\tfragColor.a *= fsTransLevel;\n"
+"\n"
+"\t// Apply fog under the control of fog factor setting from polygon header\n"
+"\tfragColor.rgb = mix(gl_Fog.color.rgb, fragColor.rgb, fsFogFactor);\n"
+"\n"
+"\t// Store final color\n"
+"\tgl_FragColor = fragColor;\n"
+"}\n"
+};
+
+
 #endif	// INCLUDED_SHADERS3D_H

Src/OSD/SDL/Main.cpp

@@ -413,7 +413,9 @@ static void ApplySettings(CINIFile *INI, const char *section)
 	if (OKAY == INI->Get(section, "FullScreen", x))
 		g_Config.fullScreen = x ? true : false;
 	if (OKAY == INI->Get(section, "WideScreen", x))
-		g_Config.wideScreen = x ? true : false;			
+		g_Config.wideScreen = x ? true : false;
+	if (OKAY == INI->Get(section, "MultiTexture", x))
+		g_Config.multiTexture = x ? true : false;
 	if (OKAY == INI->Get(section, "Throttle", x))
 		g_Config.throttle = x ? true : false;
 	if (OKAY == INI->Get(section, "ShowFrameRate", x))
@@ -460,6 +462,7 @@ static void LogConfig(void)
 	InfoLog("\tYResolution                   = %d", g_Config.yRes);
 	InfoLog("\tFullScreen                    = %d", g_Config.fullScreen);
 	InfoLog("\tWideScreen                    = %d", g_Config.wideScreen);
+	InfoLog("\tMultiTexture                  = %d", g_Config.multiTexture);
 	InfoLog("\tThrottle                      = %d", g_Config.throttle);
 	InfoLog("\tShowFrameRate                 = %d", g_Config.showFPS);
 #ifdef SUPERMODEL_DEBUGGER
@@ -1263,6 +1266,8 @@ static void Help(void)
 	puts("    -window                Windowed mode [Default]");
 	puts("    -fullscreen            Full screen mode");
 	puts("    -wide-screen           Expand 3D field of view to screen width");
+	puts("    -multi-texture         Use 8 texture maps for accurate decoding [Default]");
+	puts("    -no-multi-texture      Decode to a single texture map");
 	puts("    -no-throttle           Disable 60 Hz frame rate lock");
 	puts("    -show-fps              Display frame rate in window title bar");
 	puts("    -vert-shader=<file>    Load 3D vertex shader from external file");
@@ -1480,6 +1485,16 @@ int main(int argc, char **argv)
 			n = 1;
 			CmdLine.Set("Global", "WideScreen", n);
 		}
+		else if (!strcmp(argv[i],"-multi-texture"))
+		{
+			n = 1;
+			CmdLine.Set("Global", "MultiTexture", n);
+		}
+		else if (!strcmp(argv[i],"-no-multi-texture"))
+		{
+			n = 0;
+			CmdLine.Set("Global", "MultiTexture", n);
+		}
 		else if (!strcmp(argv[i],"-no-throttle"))
 		{
 			n = 0;