diff --git a/Source/Core/Common/Src/LinearDiskCache.h b/Source/Core/Common/Src/LinearDiskCache.h
index bb6e21197e..93f3da981e 100644
--- a/Source/Core/Common/Src/LinearDiskCache.h
+++ b/Source/Core/Common/Src/LinearDiskCache.h
@@ -26,7 +26,7 @@
 // shader cache for every revision, graphics-related or not, which is simply annoying.
 enum
 {
-	LINEAR_DISKCACHE_VER = 6592
+	LINEAR_DISKCACHE_VER = 6618
 };
 
 // On disk format:
diff --git a/Source/Core/VideoCommon/Src/Debugger.cpp b/Source/Core/VideoCommon/Src/Debugger.cpp
index 94531e46f0..09ac246e70 100644
--- a/Source/Core/VideoCommon/Src/Debugger.cpp
+++ b/Source/Core/VideoCommon/Src/Debugger.cpp
@@ -101,21 +101,21 @@ void GFXDebuggerBase::DumpPixelShader(const char* path)
 	if (!useDstAlpha)
 	{
 		output = "Destination alpha disabled:\n";
-		output += GeneratePixelShaderCode(DSTALPHA_NONE, g_ActiveConfig.backend_info.APIType, g_nativeVertexFmt->m_components);
+		output += GeneratePixelShaderCode(DSTALPHA_NONE, g_ActiveConfig.backend_info.APIType, 65536, g_nativeVertexFmt->m_components);
 	}
 	else
 	{
 		if(g_ActiveConfig.backend_info.bSupportsDualSourceBlend)
 		{
 			output = "Using dual source blending for destination alpha:\n";
-			output += GeneratePixelShaderCode(DSTALPHA_DUAL_SOURCE_BLEND, g_ActiveConfig.backend_info.APIType, g_nativeVertexFmt->m_components);
+			output += GeneratePixelShaderCode(DSTALPHA_DUAL_SOURCE_BLEND, g_ActiveConfig.backend_info.APIType, 65536, g_nativeVertexFmt->m_components);
 		}
 		else
 		{
 			output = "Using two passes for emulating destination alpha:\n";
-			output += GeneratePixelShaderCode(DSTALPHA_NONE, g_ActiveConfig.backend_info.APIType, g_nativeVertexFmt->m_components);
+			output += GeneratePixelShaderCode(DSTALPHA_NONE, g_ActiveConfig.backend_info.APIType, 65536, g_nativeVertexFmt->m_components);
 			output += "\n\nDestination alpha pass shader:\n";
-			output += GeneratePixelShaderCode(DSTALPHA_ALPHA_PASS, g_ActiveConfig.backend_info.APIType, g_nativeVertexFmt->m_components);
+			output += GeneratePixelShaderCode(DSTALPHA_ALPHA_PASS, g_ActiveConfig.backend_info.APIType, 65536, g_nativeVertexFmt->m_components);
 		}
 	}
 
diff --git a/Source/Core/VideoCommon/Src/PixelShaderGen.cpp b/Source/Core/VideoCommon/Src/PixelShaderGen.cpp
index 57ec4cd2af..0752fa73be 100644
--- a/Source/Core/VideoCommon/Src/PixelShaderGen.cpp
+++ b/Source/Core/VideoCommon/Src/PixelShaderGen.cpp
@@ -160,7 +160,7 @@ void GetPixelShaderId(PIXELSHADERUID *uid, DSTALPHA_MODE dstAlphaMode)
 //   output is given by .outreg
 //   tevtemp is set according to swapmodetables and 
 
-static void WriteStage(char *&p, int n, API_TYPE ApiType);
+static void WriteStage(char *&p, int n, API_TYPE ApiType,int maxUniforms);
 static void SampleTexture(char *&p, const char *destination, const char *texcoords, const char *texswap, int texmap, API_TYPE ApiType);
 // static void WriteAlphaCompare(char *&p, int num, int comp);
 static bool WriteAlphaTest(char *&p, API_TYPE ApiType);
@@ -442,7 +442,7 @@ char *GeneratePixelLightShader(char *p, int index, const LitChannel& chan, const
 
 
 
-const char *GeneratePixelShaderCode(DSTALPHA_MODE dstAlphaMode, API_TYPE ApiType,u32 components)
+const char *GeneratePixelShaderCode(DSTALPHA_MODE dstAlphaMode, API_TYPE ApiType,u32 maxUniforms, u32 components)
 {
 	setlocale(LC_NUMERIC, "C"); // Reset locale for compilation
 	text[sizeof(text) - 1] = 0x7C;  // canary
@@ -504,10 +504,13 @@ const char *GeneratePixelShaderCode(DSTALPHA_MODE dstAlphaMode, API_TYPE ApiType
 	WRITE(p, "uniform float4 "I_ALPHA"[1] : register(c%d);\n", C_ALPHA);
 	WRITE(p, "uniform float4 "I_TEXDIMS"[8] : register(c%d);\n", C_TEXDIMS);
 	WRITE(p, "uniform float4 "I_ZBIAS"[2] : register(c%d);\n", C_ZBIAS);
-	WRITE(p, "uniform float4 "I_INDTEXSCALE"[2] : register(c%d);\n", C_INDTEXSCALE);
-	WRITE(p, "uniform float4 "I_INDTEXMTX"[6] : register(c%d);\n", C_INDTEXMTX);
-	WRITE(p, "uniform float4 "I_FOG"[2] : register(c%d);\n", C_FOG);
-	if(g_ActiveConfig.bEnablePixelLigting)
+	if(C_INDTEXSCALE + 2 <= maxUniforms)
+		WRITE(p, "uniform float4 "I_INDTEXSCALE"[2] : register(c%d);\n", C_INDTEXSCALE);
+	if(C_INDTEXMTX + 6 <= maxUniforms)
+		WRITE(p, "uniform float4 "I_INDTEXMTX"[6] : register(c%d);\n", C_INDTEXMTX);
+	if(C_FOG + 2 <= maxUniforms)
+		WRITE(p, "uniform float4 "I_FOG"[2] : register(c%d);\n", C_FOG);
+	if(g_ActiveConfig.bEnablePixelLigting && C_PLIGHTS + 40 <= maxUniforms && C_PMATERIALS + 4 <= maxUniforms)
 	{
 		WRITE(p,"typedef struct { float4 col; float4 cosatt; float4 distatt; float4 pos; float4 dir; } Light;\n");
 		WRITE(p,"typedef struct { Light lights[8]; } s_"I_PLIGHTS";\n");
@@ -588,7 +591,7 @@ const char *GeneratePixelShaderCode(DSTALPHA_MODE dstAlphaMode, API_TYPE ApiType
 			"  float2 wrappedcoord, tempcoord;\n"
 			"  float4 cc0, cc1, cc2, cprev,crastemp,ckonsttemp;\n\n");
 	
-	if(g_ActiveConfig.bEnablePixelLigting)
+	if(g_ActiveConfig.bEnablePixelLigting && C_PLIGHTS + 40 <= maxUniforms && C_PMATERIALS + 4 <= maxUniforms)
 	{
 		if (xfregs.numTexGens < 7) 
 		{
@@ -748,7 +751,7 @@ const char *GeneratePixelShaderCode(DSTALPHA_MODE dstAlphaMode, API_TYPE ApiType
 		{
 			int texcoord = bpmem.tevindref.getTexCoord(i);
 
-			if (texcoord < numTexgen)
+			if (texcoord < numTexgen && C_INDTEXSCALE + 2 <= maxUniforms)
 				WRITE(p, "tempcoord = uv%d.xy * "I_INDTEXSCALE"[%d].%s;\n", texcoord, i/2, (i&1)?"zw":"xy");
 			else
 				WRITE(p, "tempcoord = float2(0.0f, 0.0f);\n");
@@ -770,7 +773,7 @@ const char *GeneratePixelShaderCode(DSTALPHA_MODE dstAlphaMode, API_TYPE ApiType
 	}
 
 	for (int i = 0; i < numStages; i++)
-		WriteStage(p, i, ApiType); //build the equation for this stage
+		WriteStage(p, i, ApiType,maxUniforms); //build the equation for this stage
 
 	if(numStages)
 	{
@@ -832,7 +835,8 @@ const char *GeneratePixelShaderCode(DSTALPHA_MODE dstAlphaMode, API_TYPE ApiType
 			WRITE(p, "  ocol0 = float4(prev.rgb, "I_ALPHA"[0].a);\n");
 		else
 		{
-			WriteFog(p);
+			if(C_FOG + 2 <= maxUniforms)
+				WriteFog(p);
 			WRITE(p, "  ocol0 = prev;\n");
 		}
 
@@ -900,7 +904,7 @@ static const char *TEVCMPAlphaOPTable[16] =
 };
 
 
-static void WriteStage(char *&p, int n, API_TYPE ApiType)
+static void WriteStage(char *&p, int n, API_TYPE ApiType,int maxUniforms)
 {
 	char *rasswap = swapModeTable[bpmem.combiners[n].alphaC.rswap];
 	char *texswap = swapModeTable[bpmem.combiners[n].alphaC.tswap];
@@ -934,18 +938,18 @@ static void WriteStage(char *&p, int n, API_TYPE ApiType)
 		// multiply by offset matrix and scale
 		if (bpmem.tevind[n].mid != 0)
 		{
-			if (bpmem.tevind[n].mid <= 3)
+			if (bpmem.tevind[n].mid <= 3 && C_INDTEXMTX + 6 <= maxUniforms)
 			{
 				int mtxidx = 2*(bpmem.tevind[n].mid-1);
 				WRITE(p, "float2 indtevtrans%d = float2(dot("I_INDTEXMTX"[%d].xyz, indtevcrd%d), dot("I_INDTEXMTX"[%d].xyz, indtevcrd%d));\n",
 					n, mtxidx, n, mtxidx+1, n);
 			}
-			else if (bpmem.tevind[n].mid <= 7 && bHasTexCoord)
+			else if (bpmem.tevind[n].mid <= 7 && bHasTexCoord && C_INDTEXMTX + 6 <= maxUniforms)
 			{ // s matrix
 				int mtxidx = 2*(bpmem.tevind[n].mid-5);
 				WRITE(p, "float2 indtevtrans%d = "I_INDTEXMTX"[%d].ww * uv%d.xy * indtevcrd%d.xx;\n", n, mtxidx, texcoord, n);
 			}
-			else if (bpmem.tevind[n].mid <= 11 && bHasTexCoord)
+			else if (bpmem.tevind[n].mid <= 11 && bHasTexCoord && C_INDTEXMTX + 6 <= maxUniforms)
 			{ // t matrix
 				int mtxidx = 2*(bpmem.tevind[n].mid-9);
 				WRITE(p, "float2 indtevtrans%d = "I_INDTEXMTX"[%d].ww * uv%d.xy * indtevcrd%d.yy;\n", n, mtxidx, texcoord, n);
diff --git a/Source/Core/VideoCommon/Src/PixelShaderGen.h b/Source/Core/VideoCommon/Src/PixelShaderGen.h
index baf9db8aa6..9eec0b6dd2 100644
--- a/Source/Core/VideoCommon/Src/PixelShaderGen.h
+++ b/Source/Core/VideoCommon/Src/PixelShaderGen.h
@@ -112,7 +112,7 @@ enum DSTALPHA_MODE
 	DSTALPHA_DUAL_SOURCE_BLEND // Use dual-source blending
 };
 
-const char *GeneratePixelShaderCode(DSTALPHA_MODE dstAlphaMode, API_TYPE ApiType,u32 components);
+const char *GeneratePixelShaderCode(DSTALPHA_MODE dstAlphaMode, API_TYPE ApiType,u32 maxUniforms, u32 components);
 void GetPixelShaderId(PIXELSHADERUID *uid, DSTALPHA_MODE dstAlphaMode);
 
 extern PIXELSHADERUID last_pixel_shader_uid;
diff --git a/Source/Core/VideoCommon/Src/PixelShaderManager.cpp b/Source/Core/VideoCommon/Src/PixelShaderManager.cpp
index 76bc2e7be6..1e3ed1d321 100644
--- a/Source/Core/VideoCommon/Src/PixelShaderManager.cpp
+++ b/Source/Core/VideoCommon/Src/PixelShaderManager.cpp
@@ -215,7 +215,7 @@ void PixelShaderManager::SetConstants()
         s_bFogParamChanged = false;
     }
 
-	if (nLightsChanged[0] >= 0)
+	if (g_ActiveConfig.bEnablePixelLigting && nLightsChanged[0] >= 0) // config check added because the code in here was crashing for me inside SetPSConstant4f
 	{
 		// lights don't have a 1 to 1 mapping, the color component needs to be converted to 4 floats
 		int istart = nLightsChanged[0] / 0x10;
diff --git a/Source/Plugins/Plugin_VideoDX11/Src/PixelShaderCache.cpp b/Source/Plugins/Plugin_VideoDX11/Src/PixelShaderCache.cpp
index 0799e4e70f..5f5193f45b 100644
--- a/Source/Plugins/Plugin_VideoDX11/Src/PixelShaderCache.cpp
+++ b/Source/Plugins/Plugin_VideoDX11/Src/PixelShaderCache.cpp
@@ -353,7 +353,7 @@ bool PixelShaderCache::SetShader(DSTALPHA_MODE dstAlphaMode, u32 components)
 	}
 
 	// Need to compile a new shader
-	const char* code = GeneratePixelShaderCode(dstAlphaMode, API_D3D11, components);
+	const char* code = GeneratePixelShaderCode(dstAlphaMode, API_D3D11, 65536, components);
 
 	D3DBlob* pbytecode;
 	if (!D3D::CompilePixelShader(code, strlen(code), &pbytecode))
diff --git a/Source/Plugins/Plugin_VideoDX9/Src/PixelShaderCache.cpp b/Source/Plugins/Plugin_VideoDX9/Src/PixelShaderCache.cpp
index 6288f8c268..7eaa51bbfa 100644
--- a/Source/Plugins/Plugin_VideoDX9/Src/PixelShaderCache.cpp
+++ b/Source/Plugins/Plugin_VideoDX9/Src/PixelShaderCache.cpp
@@ -44,25 +44,29 @@ static LinearDiskCache<PIXELSHADERUID, u8> g_ps_disk_cache;
 static std::set<u32> unique_shaders;
 
 #define MAX_SSAA_SHADERS 3
+enum
+{
+	COPY_TYPE_DIRECT,
+	COPY_TYPE_MATRIXCOLOR,
+	NUM_COPY_TYPES
+};
 
-static LPDIRECT3DPIXELSHADER9 s_ColorMatrixProgram[MAX_SSAA_SHADERS];
-static LPDIRECT3DPIXELSHADER9 s_ColorCopyProgram[MAX_SSAA_SHADERS];
-static LPDIRECT3DPIXELSHADER9 s_DepthMatrixProgram[MAX_SSAA_SHADERS];
+static LPDIRECT3DPIXELSHADER9 s_CopyProgram[NUM_COPY_TYPES][PixelShaderCache::NUM_DEPTH_CONVERSION_TYPES][MAX_SSAA_SHADERS];
 static LPDIRECT3DPIXELSHADER9 s_ClearProgram = 0;
 
 LPDIRECT3DPIXELSHADER9 PixelShaderCache::GetColorMatrixProgram(int SSAAMode)
 {
-	return s_ColorMatrixProgram[SSAAMode % MAX_SSAA_SHADERS];
+	return s_CopyProgram[COPY_TYPE_MATRIXCOLOR][DEPTH_CONVERSION_TYPE_NONE][SSAAMode % MAX_SSAA_SHADERS];
 }
 
-LPDIRECT3DPIXELSHADER9 PixelShaderCache::GetDepthMatrixProgram(int SSAAMode)
+LPDIRECT3DPIXELSHADER9 PixelShaderCache::GetDepthMatrixProgram(int SSAAMode, int depthConversionType)
 {
-	return s_DepthMatrixProgram[SSAAMode % MAX_SSAA_SHADERS];
+	return s_CopyProgram[COPY_TYPE_MATRIXCOLOR][depthConversionType % NUM_DEPTH_CONVERSION_TYPES][SSAAMode % MAX_SSAA_SHADERS];
 }
 
 LPDIRECT3DPIXELSHADER9 PixelShaderCache::GetColorCopyProgram(int SSAAMode)
 {
-	return s_ColorCopyProgram[SSAAMode % MAX_SSAA_SHADERS];
+	return s_CopyProgram[COPY_TYPE_DIRECT][DEPTH_CONVERSION_TYPE_NONE][SSAAMode % MAX_SSAA_SHADERS];
 }
 
 LPDIRECT3DPIXELSHADER9 PixelShaderCache::GetClearProgram()
@@ -95,132 +99,120 @@ public:
 	}
 };
 
+#define WRITE p+=sprintf
+
+static LPDIRECT3DPIXELSHADER9 CreateCopyShader(int copyMatrixType, int depthConversionType, int SSAAMode)
+{
+	//Used for Copy/resolve the color buffer
+	//Color conversion Programs
+	//Depth copy programs
+	// this should create the same shaders as before (plus some extras added for DF16), just... more manageably than listing the full program for each combination
+	char text[3072];
+
+	setlocale(LC_NUMERIC, "C"); // Reset locale for compilation
+	text[sizeof(text) - 1] = 0x7C;  // canary
+
+	char* p = text;
+	WRITE(p, "// Copy/Color Matrix/Depth Matrix shader (matrix=%d, depth=%d, ssaa=%d)\n", copyMatrixType, depthConversionType, SSAAMode);
+
+	WRITE(p, "uniform sampler samp0 : register(s0);\n");
+	if(copyMatrixType == COPY_TYPE_MATRIXCOLOR)
+		WRITE(p, "uniform float4 cColMatrix[5] : register(c%d);\n", C_COLORMATRIX);
+	WRITE(p, "void main(\n"
+	         "out float4 ocol0 : COLOR0,\n");
+
+	switch(SSAAMode % MAX_SSAA_SHADERS)
+	{
+	case 0: // 1 Sample
+		WRITE(p, "in float2 uv0 : TEXCOORD0){\n"
+		         "float4 texcol = tex2D(samp0,uv0);\n");
+		break;
+	case 1: // 1 Samples SSAA
+		WRITE(p, "in float4 uv0 : TEXCOORD0,\n"
+		         "in float4 uv1 : TEXCOORD1){\n"
+		         "float4 texcol = tex2D(samp0,uv0.xy);\n");
+		break;
+	case 2: // 4 Samples SSAA
+		WRITE(p, "in float4 uv0 : TEXCOORD0,\n"
+		         "in float4 uv1 : TEXCOORD1,\n"
+		         "in float4 uv2 : TEXCOORD2,\n"
+		         "in float4 uv3 : TEXCOORD3){\n"
+		         "float4 texcol = (tex2D(samp0,uv1.xy) + tex2D(samp0,uv1.wz) + tex2D(samp0,uv2.xy) + tex2D(samp0,uv2.wz))*0.25f;\n");
+		break;
+	}
+
+	switch(depthConversionType % PixelShaderCache::NUM_DEPTH_CONVERSION_TYPES)
+	{
+	case PixelShaderCache::DEPTH_CONVERSION_TYPE_NONE:
+		break;
+	case PixelShaderCache::DEPTH_CONVERSION_TYPE_16BIT:
+		// this is probably wrong. but it works better than the 24-bit conversion we used to generate in this case.
+		WRITE(p, "float4 EncodedDepth = frac((texcol.r * (65535.0f/65536.0f)) * float4(1.0f,255.0f,255.0f*255.0f,255.0f*255.0f*255.0f));\n"
+		         "texcol = float4((EncodedDepth.rgb * (65536.0f/65535.0f)),1.0f);\n");
+		break;
+	case PixelShaderCache::DEPTH_CONVERSION_TYPE_24BIT:
+		WRITE(p, "float4 EncodedDepth = frac((texcol.r * (16777215.0f/16777216.0f)) * float4(1.0f,255.0f,255.0f*255.0f,255.0f*255.0f*255.0f));\n"
+		         "texcol = float4((EncodedDepth.rgb * (16777216.0f/16777215.0f)),1.0f);\n");
+		break;
+	}
+
+	if(copyMatrixType == COPY_TYPE_MATRIXCOLOR)
+		WRITE(p, "ocol0 = float4(dot(texcol,cColMatrix[0]),dot(texcol,cColMatrix[1]),dot(texcol,cColMatrix[2]),dot(texcol,cColMatrix[3])) + cColMatrix[4];\n");
+	else
+		WRITE(p, "ocol0 = texcol;\n");
+
+	WRITE(p, "}\n");
+	if (text[sizeof(text) - 1] != 0x7C)
+		PanicAlert("PixelShaderCache copy shader generator - buffer too small, canary has been eaten!");
+
+	setlocale(LC_NUMERIC, ""); // restore locale
+	return D3D::CompileAndCreatePixelShader(text, (int)strlen(text));	
+}
+
 void PixelShaderCache::Init()
 {
 	//program used for clear screen
-	char pprog[3072];
-	sprintf(pprog, "void main(\n"
-						"out float4 ocol0 : COLOR0,\n"
-						" in float4 incol0 : COLOR0){\n"
-						"ocol0 = incol0;\n"
-						"}\n");
-	s_ClearProgram = D3D::CompileAndCreatePixelShader(pprog, (int)strlen(pprog));	
+	{
+		char pprog[3072];
+		sprintf(pprog, "void main(\n"
+							"out float4 ocol0 : COLOR0,\n"
+							" in float4 incol0 : COLOR0){\n"
+							"ocol0 = incol0;\n"
+							"}\n");
+		s_ClearProgram = D3D::CompileAndCreatePixelShader(pprog, (int)strlen(pprog));	
+	}
 
-	//Used for Copy/resolve the color buffer
-	//1 Sample
-	sprintf(pprog, "uniform sampler samp0 : register(s0);\n"
-						"void main(\n"
-						"out float4 ocol0 : COLOR0,\n"
-						"in float2 uv0 : TEXCOORD0){\n"
-						"ocol0 = tex2D(samp0,uv0);\n"						
-						"}\n");
-	s_ColorCopyProgram[0] = D3D::CompileAndCreatePixelShader(pprog, (int)strlen(pprog));		
+	int shaderModel = ((D3D::GetCaps().PixelShaderVersion >> 8) & 0xFF);
+	int maxConstants = (shaderModel < 3) ? 32 : ((shaderModel < 4) ? 224 : 65536);
+	bool canUseColorMatrix = (C_COLORMATRIX + 5 <= maxConstants);
 
-	//1 Samples SSAA
-	sprintf(pprog, "uniform sampler samp0 : register(s0);\n"
-					"void main(\n"
-					"out float4 ocol0 : COLOR0,\n"
-					"in float4 uv0 : TEXCOORD0,\n"
-					"in float4 uv1 : TEXCOORD1){\n"
-					"ocol0 = tex2D(samp0,uv0.xy);\n"
-					"}\n");
-	s_ColorCopyProgram[1] = D3D::CompileAndCreatePixelShader(pprog, (int)strlen(pprog));
-
-	//4 Samples SSAA
-	sprintf(pprog, "uniform sampler samp0 : register(s0);\n"					
-					"void main(\n"
-					"out float4 ocol0 : COLOR0,\n"
-					"in float4 uv0 : TEXCOORD0,\n"
-					"in float4 uv1 : TEXCOORD1,\n"
-					"in float4 uv2 : TEXCOORD2,\n"
-					"in float4 uv3 : TEXCOORD3){\n"
-					"ocol0 = (tex2D(samp0,uv1.xy) + tex2D(samp0,uv1.wz) + tex2D(samp0,uv2.xy) + tex2D(samp0,uv2.wz))*0.25;\n"
-					"}\n");
-	s_ColorCopyProgram[2] = D3D::CompileAndCreatePixelShader(pprog, (int)strlen(pprog));
-
-
-
-	//Color conversion Programs
-	//1 sample
-	sprintf(pprog, "uniform sampler samp0 : register(s0);\n"
-						"uniform float4 cColMatrix[5] : register(c%d);\n"
-						"void main(\n"
-						"out float4 ocol0 : COLOR0,\n"
-						" in float2 uv0 : TEXCOORD0){\n"
-						"float4 texcol = tex2D(samp0,uv0);\n"
-						"ocol0 = float4(dot(texcol,cColMatrix[0]),dot(texcol,cColMatrix[1]),dot(texcol,cColMatrix[2]),dot(texcol,cColMatrix[3])) + cColMatrix[4];\n"
-						"}\n",C_COLORMATRIX);
-	s_ColorMatrixProgram[0] = D3D::CompileAndCreatePixelShader(pprog, (int)strlen(pprog));		
-	
-	//1 samples SSAA
-	sprintf(pprog, "uniform sampler samp0 : register(s0);\n"
-					"uniform float4 cColMatrix[5] : register(c%d);\n"
-					"void main(\n"
-					"out float4 ocol0 : COLOR0,\n"
-					"in float4 uv0 : TEXCOORD0,\n"
-					"in float4 uv1 : TEXCOORD1){\n"
-					"float4 texcol = tex2D(samp0,uv0.xy);\n"
-					"ocol0 = float4(dot(texcol,cColMatrix[0]),dot(texcol,cColMatrix[1]),dot(texcol,cColMatrix[2]),dot(texcol,cColMatrix[3])) + cColMatrix[4];\n"
-					"}\n",C_COLORMATRIX);
-	s_ColorMatrixProgram[1] = D3D::CompileAndCreatePixelShader(pprog, (int)strlen(pprog));
-
-	//4 samples SSAA
-	sprintf(pprog, "uniform sampler samp0 : register(s0);\n"
-					"uniform float4 cColMatrix[5] : register(c%d);\n"
-					"void main(\n"
-					"out float4 ocol0 : COLOR0,\n"
-					"in float4 uv0 : TEXCOORD0,\n"
-					"in float4 uv1 : TEXCOORD1,\n"
-					"in float4 uv2 : TEXCOORD2,\n"
-					"in float4 uv3 : TEXCOORD3){\n"
-					"float4 texcol = (tex2D(samp0,uv1.xy) + tex2D(samp0,uv1.wz) + tex2D(samp0,uv2.xy) + tex2D(samp0,uv2.wz))*0.25f;\n"
-					"ocol0 = float4(dot(texcol,cColMatrix[0]),dot(texcol,cColMatrix[1]),dot(texcol,cColMatrix[2]),dot(texcol,cColMatrix[3])) + cColMatrix[4];\n"
-					"}\n",C_COLORMATRIX);
-	s_ColorMatrixProgram[2] = D3D::CompileAndCreatePixelShader(pprog, (int)strlen(pprog));
-
-	//Depth copy programs
-	//1 sample
-	sprintf(pprog, "uniform sampler samp0 : register(s0);\n"
-						"uniform float4 cColMatrix[5] : register(c%d);\n"
-						"void main(\n"
-						"out float4 ocol0 : COLOR0,\n"
-						" in float2 uv0 : TEXCOORD0){\n"
-						"float4 texcol = tex2D(samp0,uv0);\n"
-						"float4 EncodedDepth = frac((texcol.r * (16777215.0f/16777216.0f)) * float4(1.0f,255.0f,255.0f*255.0f,255.0f*255.0f*255.0f));\n"
-						"texcol = float4((EncodedDepth.rgb * (16777216.0f/16777215.0f)),1.0f);\n"
-						"ocol0 = float4(dot(texcol,cColMatrix[0]),dot(texcol,cColMatrix[1]),dot(texcol,cColMatrix[2]),dot(texcol,cColMatrix[3])) + cColMatrix[4];\n"
-						"}\n",C_COLORMATRIX);
-	s_DepthMatrixProgram[0] = D3D::CompileAndCreatePixelShader(pprog, (int)strlen(pprog));
-	
-	//1 sample SSAA
-	sprintf(pprog, "uniform sampler samp0 : register(s0);\n"					
-					"uniform float4 cColMatrix[5] : register(c%d);\n"
-					"void main(\n"
-					"out float4 ocol0 : COLOR0,\n"
-					"in float4 uv0 : TEXCOORD0,\n"
-					"in float4 uv1 : TEXCOORD1){\n"
-					"float4 texcol = tex2D(samp0,uv0.xy);\n"
-					"float4 EncodedDepth = frac((texcol.r * (16777215.0f/16777216.0f)) * float4(1.0f,255.0f,255.0f*255.0f,255.0f*255.0f*255.0f));\n"
-					"texcol = float4((EncodedDepth.rgb * (16777216.0f/16777215.0f)),1.0f);\n"
-					"ocol0 = float4(dot(texcol,cColMatrix[0]),dot(texcol,cColMatrix[1]),dot(texcol,cColMatrix[2]),dot(texcol,cColMatrix[3])) + cColMatrix[4];\n"						
-					"}\n",C_COLORMATRIX);
-	s_DepthMatrixProgram[1] = D3D::CompileAndCreatePixelShader(pprog, (int)strlen(pprog));
-
-	//4 sample SSAA
-	sprintf(pprog, "uniform sampler samp0 : register(s0);\n"					
-					"uniform float4 cColMatrix[5] : register(c%d);\n"
-					"void main(\n"
-					"out float4 ocol0 : COLOR0,\n"
-					"in float4 uv0 : TEXCOORD0,\n"
-					"in float4 uv1 : TEXCOORD1,\n"
-					"in float4 uv2 : TEXCOORD2,\n"
-					"in float4 uv3 : TEXCOORD3){\n"
-					"float4 texcol = (tex2D(samp0,uv1.xy) + tex2D(samp0,uv1.wz) + tex2D(samp0,uv2.xy) + tex2D(samp0,uv2.wz))*0.25f;\n"
-					"float4 EncodedDepth = frac((texcol.r * (16777215.0f/16777216.0f)) * float4(1.0f,255.0f,255.0f*255.0f,255.0f*255.0f*255.0f));\n"
-					"texcol = float4((EncodedDepth.rgb * (16777216.0f/16777215.0f)),1.0f);\n"
-					"ocol0 = float4(dot(texcol,cColMatrix[0]),dot(texcol,cColMatrix[1]),dot(texcol,cColMatrix[2]),dot(texcol,cColMatrix[3])) + cColMatrix[4];\n"						
-					"}\n",C_COLORMATRIX);
-	s_DepthMatrixProgram[2] = D3D::CompileAndCreatePixelShader(pprog, (int)strlen(pprog));
+	// other screen copy/convert programs
+	for(int copyMatrixType = 0; copyMatrixType < NUM_COPY_TYPES; copyMatrixType++)
+	{
+		for(int depthType = 0; depthType < NUM_DEPTH_CONVERSION_TYPES; depthType++)
+		{
+			for(int ssaaMode = 0; ssaaMode < MAX_SSAA_SHADERS; ssaaMode++)
+			{
+				if(ssaaMode && !s_CopyProgram[copyMatrixType][depthType][ssaaMode-1]
+				|| depthType && !s_CopyProgram[copyMatrixType][depthType-1][ssaaMode]
+				|| copyMatrixType && !s_CopyProgram[copyMatrixType-1][depthType][ssaaMode])
+				{
+					// if it failed at a lower setting, it's going to fail here for the same reason it did there,
+					// so skip this attempt to avoid duplicate error messages.
+					s_CopyProgram[copyMatrixType][depthType][ssaaMode] = NULL;
+				}
+				else if(copyMatrixType == COPY_TYPE_MATRIXCOLOR && !canUseColorMatrix)
+				{
+					// color matrix not supported, so substitute the nearest equivalent program that doesn't use it.
+					s_CopyProgram[copyMatrixType][depthType][ssaaMode] = s_CopyProgram[COPY_TYPE_DIRECT][depthType][ssaaMode];
+				}
+				else
+				{
+					s_CopyProgram[copyMatrixType][depthType][ssaaMode] = CreateCopyShader(copyMatrixType, depthType, ssaaMode);
+				}
+			}
+		}
+	}
 
 	Clear();
 
@@ -248,15 +240,18 @@ void PixelShaderCache::Clear()
 
 void PixelShaderCache::Shutdown()
 {
-	for(int i = 0;i < MAX_SSAA_SHADERS; i++)
-	{
-		if (s_ColorMatrixProgram[i]) s_ColorMatrixProgram[i]->Release();
-		s_ColorMatrixProgram[i] = NULL;
-		if (s_ColorCopyProgram[i]) s_ColorCopyProgram[i]->Release();
-		s_ColorCopyProgram[i] = NULL;
-		if (s_DepthMatrixProgram[i]) s_DepthMatrixProgram[i]->Release();
-		s_DepthMatrixProgram[i] = NULL;
-	}
+	for(int copyMatrixType = 0; copyMatrixType < NUM_COPY_TYPES; copyMatrixType++)
+		for(int depthType = 0; depthType < NUM_DEPTH_CONVERSION_TYPES; depthType++)
+			for(int ssaaMode = 0; ssaaMode < MAX_SSAA_SHADERS; ssaaMode++)
+				if(s_CopyProgram[copyMatrixType][depthType][ssaaMode]
+				&& (copyMatrixType == 0 || s_CopyProgram[copyMatrixType][depthType][ssaaMode] != s_CopyProgram[copyMatrixType-1][depthType][ssaaMode]))
+					s_CopyProgram[copyMatrixType][depthType][ssaaMode]->Release();
+
+	for(int copyMatrixType = 0; copyMatrixType < NUM_COPY_TYPES; copyMatrixType++)
+		for(int depthType = 0; depthType < NUM_DEPTH_CONVERSION_TYPES; depthType++)
+			for(int ssaaMode = 0; ssaaMode < MAX_SSAA_SHADERS; ssaaMode++)
+				s_CopyProgram[copyMatrixType][depthType][ssaaMode] = NULL;
+
 	if (s_ClearProgram)	s_ClearProgram->Release();
 	s_ClearProgram = NULL;	
 	
@@ -296,8 +291,11 @@ bool PixelShaderCache::SetShader(DSTALPHA_MODE dstAlphaMode, u32 components)
 		return (entry.shader != NULL);
 	}
 
+	int shaderModel = ((D3D::GetCaps().PixelShaderVersion >> 8) & 0xFF);
+	int maxConstants = (shaderModel < 3) ? 32 : ((shaderModel < 4) ? 224 : 65536);
+
 	// Need to compile a new shader
-	const char *code = GeneratePixelShaderCode(dstAlphaMode, API_D3D9, components);
+	const char *code = GeneratePixelShaderCode(dstAlphaMode, API_D3D9, maxConstants, components);
 
 	u32 code_hash = HashAdler32((const u8 *)code, strlen(code));
 	unique_shaders.insert(code_hash);
diff --git a/Source/Plugins/Plugin_VideoDX9/Src/PixelShaderCache.h b/Source/Plugins/Plugin_VideoDX9/Src/PixelShaderCache.h
index cc27585d62..7568d76fa6 100644
--- a/Source/Plugins/Plugin_VideoDX9/Src/PixelShaderCache.h
+++ b/Source/Plugins/Plugin_VideoDX9/Src/PixelShaderCache.h
@@ -62,7 +62,14 @@ public:
 	static bool InsertByteCode(const PIXELSHADERUID &uid, const u8 *bytecode, int bytecodelen, bool activate);
 	static LPDIRECT3DPIXELSHADER9 GetColorMatrixProgram(int SSAAMode);
 	static LPDIRECT3DPIXELSHADER9 GetColorCopyProgram(int SSAAMode);
-	static LPDIRECT3DPIXELSHADER9 GetDepthMatrixProgram(int SSAAMode);
+	enum
+	{
+		DEPTH_CONVERSION_TYPE_NONE,
+		DEPTH_CONVERSION_TYPE_16BIT,
+		DEPTH_CONVERSION_TYPE_24BIT,
+		NUM_DEPTH_CONVERSION_TYPES
+	};
+	static LPDIRECT3DPIXELSHADER9 GetDepthMatrixProgram(int SSAAMode, int depthConversionType);
 	static LPDIRECT3DPIXELSHADER9 GetClearProgram();	
 };
 
diff --git a/Source/Plugins/Plugin_VideoDX9/Src/Render.cpp b/Source/Plugins/Plugin_VideoDX9/Src/Render.cpp
index 3703362d68..f769f3315f 100644
--- a/Source/Plugins/Plugin_VideoDX9/Src/Render.cpp
+++ b/Source/Plugins/Plugin_VideoDX9/Src/Render.cpp
@@ -589,13 +589,22 @@ u32 Renderer::AccessEFB(EFBAccessType type, u32 x, u32 y, u32 poke_data)
 		
 		D3D::ChangeSamplerState(0, D3DSAMP_MINFILTER, D3DTEXF_POINT);
 
+		D3DFORMAT bformat = FramebufferManager::GetEFBDepthRTSurfaceFormat();
+		int depthConversionType;
+		if(bformat == FOURCC_RAWZ)
+			depthConversionType = PixelShaderCache::DEPTH_CONVERSION_TYPE_NONE;
+		else if(bformat == FOURCC_DF16)
+			depthConversionType = PixelShaderCache::DEPTH_CONVERSION_TYPE_16BIT;
+		else
+			depthConversionType = PixelShaderCache::DEPTH_CONVERSION_TYPE_24BIT;
+
 		D3D::drawShadedTexQuad(
 			read_texture,
 			&RectToLock,
 			Renderer::GetFullTargetWidth(),
 			Renderer::GetFullTargetHeight(),
 			4, 4,
-			(FramebufferManager::GetEFBDepthRTSurfaceFormat() == FOURCC_RAWZ) ? PixelShaderCache::GetColorMatrixProgram(0) : PixelShaderCache::GetDepthMatrixProgram(0),
+			PixelShaderCache::GetDepthMatrixProgram(0, depthConversionType),
 			VertexShaderCache::GetSimpleVertexShader(0));
 
 		D3D::RefreshSamplerState(0, D3DSAMP_MINFILTER);
diff --git a/Source/Plugins/Plugin_VideoDX9/Src/TextureCache.cpp b/Source/Plugins/Plugin_VideoDX9/Src/TextureCache.cpp
index aa215ae4cc..627e6e0f4b 100644
--- a/Source/Plugins/Plugin_VideoDX9/Src/TextureCache.cpp
+++ b/Source/Plugins/Plugin_VideoDX9/Src/TextureCache.cpp
@@ -131,12 +131,18 @@ void TextureCache::TCacheEntry::FromRenderTarget(bool bFromZBuffer, bool bScaleB
 		D3DFORMAT bformat = FramebufferManager::GetEFBDepthRTSurfaceFormat();
 		int SSAAMode = g_ActiveConfig.iMultisampleMode;
 
+		int depthConversionType;
+		if(bformat == FOURCC_RAWZ || bformat == D3DFMT_D24X8 || !bFromZBuffer)
+			depthConversionType = PixelShaderCache::DEPTH_CONVERSION_TYPE_NONE;
+		else if(bformat == FOURCC_DF16)
+			depthConversionType = PixelShaderCache::DEPTH_CONVERSION_TYPE_16BIT;
+		else
+			depthConversionType = PixelShaderCache::DEPTH_CONVERSION_TYPE_24BIT;
+
 		D3D::drawShadedTexQuad(read_texture, &sourcerect, 
 			Renderer::GetFullTargetWidth(), Renderer::GetFullTargetHeight(),
 			virtualW, virtualH,
-			((bformat != FOURCC_RAWZ && bformat != D3DFMT_D24X8) && bFromZBuffer) ?
-				PixelShaderCache::GetDepthMatrixProgram(SSAAMode) :
-				PixelShaderCache::GetColorMatrixProgram(SSAAMode),
+			PixelShaderCache::GetDepthMatrixProgram(SSAAMode, depthConversionType),
 			VertexShaderCache::GetSimpleVertexShader(SSAAMode));
 
 		Rendersurf->Release();
diff --git a/Source/Plugins/Plugin_VideoDX9/Src/TextureConverter.cpp b/Source/Plugins/Plugin_VideoDX9/Src/TextureConverter.cpp
index b734e21c91..3d1808c6e1 100644
--- a/Source/Plugins/Plugin_VideoDX9/Src/TextureConverter.cpp
+++ b/Source/Plugins/Plugin_VideoDX9/Src/TextureConverter.cpp
@@ -52,6 +52,7 @@ static LPDIRECT3DPIXELSHADER9 s_yuyvToRgbProgram = NULL;
 // Not all slots are taken - but who cares.
 const u32 NUM_ENCODING_PROGRAMS = 64;
 static LPDIRECT3DPIXELSHADER9 s_encodingPrograms[NUM_ENCODING_PROGRAMS];
+static bool s_encodingProgramsFailed[NUM_ENCODING_PROGRAMS];
 
 void CreateRgbToYuyvProgram()
 {
@@ -121,6 +122,13 @@ LPDIRECT3DPIXELSHADER9 GetOrCreateEncodingShader(u32 format)
 
 	if (!s_encodingPrograms[format])
 	{
+		if(s_encodingProgramsFailed[format])
+		{
+			// we already failed to create a shader for this format,
+			// so instead of re-trying and showing the same error message every frame, just return.
+			return NULL;
+		}
+
 		const char* shader = TextureConversionShader::GenerateEncodingShader(format,API_D3D9);
 
 #if defined(_DEBUG) || defined(DEBUGFAST)
@@ -135,6 +143,7 @@ LPDIRECT3DPIXELSHADER9 GetOrCreateEncodingShader(u32 format)
 		s_encodingPrograms[format] = D3D::CompileAndCreatePixelShader(shader, (int)strlen(shader));
 		if (!s_encodingPrograms[format]) {
 			ERROR_LOG(VIDEO, "Failed to create encoding fragment program");
+			s_encodingProgramsFailed[format] = true;
 		}
     }
 	return s_encodingPrograms[format];
@@ -145,6 +154,7 @@ void Init()
 	for (unsigned int i = 0; i < NUM_ENCODING_PROGRAMS; i++)
 	{
 		s_encodingPrograms[i] = NULL;
+		s_encodingProgramsFailed[i] = false;
 	}
 	for (unsigned int i = 0; i < NUM_TRANSFORM_BUFFERS; i++)
 	{
diff --git a/Source/Plugins/Plugin_VideoOGL/Src/PixelShaderCache.cpp b/Source/Plugins/Plugin_VideoOGL/Src/PixelShaderCache.cpp
index 15b5668df0..a24fc93893 100644
--- a/Source/Plugins/Plugin_VideoOGL/Src/PixelShaderCache.cpp
+++ b/Source/Plugins/Plugin_VideoOGL/Src/PixelShaderCache.cpp
@@ -219,7 +219,7 @@ FRAGMENTSHADER* PixelShaderCache::SetShader(DSTALPHA_MODE dstAlphaMode, u32 comp
 	PSCacheEntry& newentry = PixelShaders[uid];
 	newentry.frameCount = frameCount;
 	pShaderLast = &newentry.shader;
-	const char *code = GeneratePixelShaderCode(dstAlphaMode, API_OPENGL, components);
+	const char *code = GeneratePixelShaderCode(dstAlphaMode, API_OPENGL, 65536, components);
 
 #if defined(_DEBUG) || defined(DEBUGFAST)
 	if (g_ActiveConfig.iLog & CONF_SAVESHADERS && code) {