First Commit

bin/resources/shaders/common/fxaa.fx

@@ -0,0 +1,498 @@
+// SPDX-FileCopyrightText: 2002-2025 PCSX2 Dev Team
+// SPDX-License-Identifier: GPL-3.0+
+
+#ifndef FXAA_HLSL
+    #define FXAA_HLSL 0
+#endif
+#ifndef FXAA_GLSL_130
+    #define FXAA_GLSL_130 0
+#endif
+#ifndef FXAA_GLSL_VK
+    #define FXAA_GLSL_VK 0
+#endif
+
+#define UHQ_FXAA 1          //High Quality Fast Approximate Anti Aliasing. Adapted for GS from Timothy Lottes FXAA 3.11.
+#define FxaaSubpixMax 0.0   //[0.00 to 1.00] Amount of subpixel aliasing removal. 0.00: Edge only antialiasing (no blurring)
+#define FxaaEarlyExit 1     //[0 or 1] Use Fxaa early exit pathing. When disabled, the entire scene is antialiased(FSAA). 0 is off, 1 is on.
+
+/*------------------------------------------------------------------------------
+							 [GLOBALS|FUNCTIONS]
+------------------------------------------------------------------------------*/
+#if (FXAA_GLSL_130 == 1)
+
+in vec2 PSin_t;
+
+layout(location = 0) out vec4 SV_Target0;
+layout(binding = 0) uniform sampler2D TextureSampler;
+
+#elif (FXAA_GLSL_VK == 1)
+
+layout(location = 0) in vec2 PSin_t;
+layout(location = 0) out vec4 SV_Target0;
+layout(set = 0, binding = 0) uniform sampler2D TextureSampler;
+
+#elif (FXAA_HLSL == 1)
+Texture2D Texture : register(t0);
+SamplerState TextureSampler : register(s0);
+
+struct VS_INPUT
+{
+	float4 p : POSITION;
+	float2 t : TEXCOORD0;
+};
+
+struct VS_OUTPUT
+{
+	float4 p : SV_Position;
+	float2 t : TEXCOORD0;
+};
+
+struct PS_OUTPUT
+{
+	float4 c : SV_Target0;
+};
+
+#elif defined(__METAL_VERSION__)
+static constexpr sampler MAIN_SAMPLER(coord::normalized, address::clamp_to_edge, filter::linear);
+#endif
+
+/*------------------------------------------------------------------------------
+                             [FXAA CODE SECTION]
+------------------------------------------------------------------------------*/
+
+#if (FXAA_HLSL == 1)
+struct FxaaTex { SamplerState smpl; Texture2D tex; };
+#define FxaaTexTop(t, p) t.tex.SampleLevel(t.smpl, p, 0.0)
+#define FxaaTexOff(t, p, o, r) t.tex.SampleLevel(t.smpl, p, 0.0, o)
+#define FxaaDiscard clip(-1)
+#define FxaaSat(x) saturate(x)
+
+#elif (FXAA_GLSL_130 == 1 || FXAA_GLSL_VK == 1)
+#define int2 ivec2
+#define float2 vec2
+#define float3 vec3
+#define float4 vec4
+#define FxaaDiscard discard
+#define FxaaSat(x) clamp(x, 0.0, 1.0)
+#define FxaaTex sampler2D
+#define FxaaTexTop(t, p) textureLod(t, p, 0.0)
+#define FxaaTexOff(t, p, o, r) textureLodOffset(t, p, 0.0, o)
+
+#elif defined(__METAL_VERSION__)
+#define FxaaTex texture2d<float>
+#define FxaaTexTop(t, p) t.sample(MAIN_SAMPLER, p)
+#define FxaaTexOff(t, p, o, r) t.sample(MAIN_SAMPLER, p, o)
+#define FxaaDiscard discard_fragment()
+#define FxaaSat(x) saturate(x)
+#endif
+
+#define FxaaEdgeThreshold 0.063
+#define FxaaEdgeThresholdMin 0.00
+#define FXAA_QUALITY_P0 1.0
+#define FXAA_QUALITY_P1 1.5
+#define FXAA_QUALITY_P2 2.0
+#define FXAA_QUALITY_P3 2.0
+#define FXAA_QUALITY_P4 2.0
+#define FXAA_QUALITY_P5 2.0
+#define FXAA_QUALITY_P6 2.0
+#define FXAA_QUALITY_P7 2.0
+#define FXAA_QUALITY_P8 2.0
+#define FXAA_QUALITY_P9 2.0
+#define FXAA_QUALITY_P10 4.0
+#define FXAA_QUALITY_P11 8.0
+#define FXAA_QUALITY_P12 8.0
+
+/*------------------------------------------------------------------------------
+                        [GAMMA PREPASS CODE SECTION]
+------------------------------------------------------------------------------*/
+float RGBLuminance(float3 color)
+{
+	const float3 lumCoeff = float3(0.2126729, 0.7151522, 0.0721750);
+	return dot(color.rgb, lumCoeff);
+}
+
+float3 RGBGammaToLinear(float3 color, float gamma)
+{
+	color = FxaaSat(color);
+	color.r = (color.r <= 0.0404482362771082) ?
+	color.r / 12.92 : pow((color.r + 0.055) / 1.055, gamma);
+	color.g = (color.g <= 0.0404482362771082) ?
+	color.g / 12.92 : pow((color.g + 0.055) / 1.055, gamma);
+	color.b = (color.b <= 0.0404482362771082) ?
+	color.b / 12.92 : pow((color.b + 0.055) / 1.055, gamma);
+
+	return color;
+}
+
+float3 LinearToRGBGamma(float3 color, float gamma)
+{
+	color = FxaaSat(color);
+	color.r = (color.r <= 0.00313066844250063) ?
+	color.r * 12.92 : 1.055 * pow(color.r, 1.0 / gamma) - 0.055;
+	color.g = (color.g <= 0.00313066844250063) ?
+	color.g * 12.92 : 1.055 * pow(color.g, 1.0 / gamma) - 0.055;
+	color.b = (color.b <= 0.00313066844250063) ?
+	color.b * 12.92 : 1.055 * pow(color.b, 1.0 / gamma) - 0.055;
+
+	return color;
+}
+
+float4 PreGammaPass(float4 color)
+{
+	const float GammaConst = 2.233;
+	color.rgb = RGBGammaToLinear(color.rgb, GammaConst);
+	color.rgb = LinearToRGBGamma(color.rgb, GammaConst);
+	color.a = RGBLuminance(color.rgb);
+
+	return color;
+}
+
+
+/*------------------------------------------------------------------------------
+                        [FXAA CODE SECTION]
+------------------------------------------------------------------------------*/
+
+float FxaaLuma(float4 rgba)
+{ 
+	rgba.w = RGBLuminance(rgba.xyz);
+	return rgba.w; 
+}
+
+float4 FxaaPixelShader(float2 pos, FxaaTex tex, float2 fxaaRcpFrame, float fxaaSubpix, float fxaaEdgeThreshold, float fxaaEdgeThresholdMin)
+{
+	float2 posM = pos;
+	float4 rgbyM = FxaaTexTop(tex, posM);
+	rgbyM.w = RGBLuminance(rgbyM.xyz);
+	#define lumaM rgbyM.w
+
+	float lumaS = FxaaLuma(FxaaTexOff(tex, posM, int2( 0, 1), fxaaRcpFrame.xy));
+	float lumaE = FxaaLuma(FxaaTexOff(tex, posM, int2( 1, 0), fxaaRcpFrame.xy));
+	float lumaN = FxaaLuma(FxaaTexOff(tex, posM, int2( 0,-1), fxaaRcpFrame.xy));
+	float lumaW = FxaaLuma(FxaaTexOff(tex, posM, int2(-1, 0), fxaaRcpFrame.xy));
+
+	float maxSM = max(lumaS, lumaM);
+	float minSM = min(lumaS, lumaM);
+	float maxESM = max(lumaE, maxSM);
+	float minESM = min(lumaE, minSM);
+	float maxWN = max(lumaN, lumaW);
+	float minWN = min(lumaN, lumaW);
+
+	float rangeMax = max(maxWN, maxESM);
+	float rangeMin = min(minWN, minESM);
+	float range = rangeMax - rangeMin;
+	float rangeMaxScaled = rangeMax * fxaaEdgeThreshold;
+	float rangeMaxClamped = max(fxaaEdgeThresholdMin, rangeMaxScaled);
+
+	#if (FxaaEarlyExit == 1)
+	// Potential optimization, early exit.
+	if (range < rangeMaxClamped)
+		return rgbyM;
+	#endif
+
+	float lumaNW = FxaaLuma(FxaaTexOff(tex, posM, int2(-1,-1), fxaaRcpFrame.xy));
+	float lumaSE = FxaaLuma(FxaaTexOff(tex, posM, int2( 1, 1), fxaaRcpFrame.xy));
+	float lumaNE = FxaaLuma(FxaaTexOff(tex, posM, int2( 1,-1), fxaaRcpFrame.xy));
+	float lumaSW = FxaaLuma(FxaaTexOff(tex, posM, int2(-1, 1), fxaaRcpFrame.xy));
+
+	float lumaNS = lumaN + lumaS;
+	float lumaWE = lumaW + lumaE;
+	float subpixRcpRange = 1.0/range;
+	float subpixNSWE = lumaNS + lumaWE;
+	float edgeHorz1 = (-2.0 * lumaM) + lumaNS;
+	float edgeVert1 = (-2.0 * lumaM) + lumaWE;
+	float lumaNESE = lumaNE + lumaSE;
+	float lumaNWNE = lumaNW + lumaNE;
+	float edgeHorz2 = (-2.0 * lumaE) + lumaNESE;
+	float edgeVert2 = (-2.0 * lumaN) + lumaNWNE;
+
+	float lumaNWSW = lumaNW + lumaSW;
+	float lumaSWSE = lumaSW + lumaSE;
+	float edgeHorz4 = (abs(edgeHorz1) * 2.0) + abs(edgeHorz2);
+	float edgeVert4 = (abs(edgeVert1) * 2.0) + abs(edgeVert2);
+	float edgeHorz3 = (-2.0 * lumaW) + lumaNWSW;
+	float edgeVert3 = (-2.0 * lumaS) + lumaSWSE;
+	float edgeHorz = abs(edgeHorz3) + edgeHorz4;
+	float edgeVert = abs(edgeVert3) + edgeVert4;
+
+	float subpixNWSWNESE = lumaNWSW + lumaNESE;
+	float lengthSign = fxaaRcpFrame.x;
+	bool horzSpan = edgeHorz >= edgeVert;
+	float subpixA = subpixNSWE * 2.0 + subpixNWSWNESE;
+	if(!horzSpan) lumaN = lumaW;
+	if(!horzSpan) lumaS = lumaE;
+	if(horzSpan) lengthSign = fxaaRcpFrame.y;
+	float subpixB = (subpixA * (1.0/12.0)) - lumaM;
+
+	float gradientN = lumaN - lumaM;
+	float gradientS = lumaS - lumaM;
+	float lumaNN = lumaN + lumaM;
+	float lumaSS = lumaS + lumaM;
+	bool pairN = abs(gradientN) >= abs(gradientS);
+	float gradient = max(abs(gradientN), abs(gradientS));
+	if(pairN) lengthSign = -lengthSign;
+	float subpixC = FxaaSat(abs(subpixB) * subpixRcpRange);
+
+	float2 posB;
+	posB.x = posM.x;
+	posB.y = posM.y;
+	float2 offNP;
+	offNP.x = (!horzSpan) ? 0.0 : fxaaRcpFrame.x;
+	offNP.y = ( horzSpan) ? 0.0 : fxaaRcpFrame.y;
+	if(!horzSpan) posB.x += lengthSign * 0.5;
+	if( horzSpan) posB.y += lengthSign * 0.5;
+
+	float2 posN;
+	posN.x = posB.x - offNP.x * FXAA_QUALITY_P0;
+	posN.y = posB.y - offNP.y * FXAA_QUALITY_P0;
+	float2 posP;
+	posP.x = posB.x + offNP.x * FXAA_QUALITY_P0;
+	posP.y = posB.y + offNP.y * FXAA_QUALITY_P0;
+	float subpixD = ((-2.0)*subpixC) + 3.0;
+	float lumaEndN = FxaaLuma(FxaaTexTop(tex, posN));
+	float subpixE = subpixC * subpixC;
+	float lumaEndP = FxaaLuma(FxaaTexTop(tex, posP));
+
+	if(!pairN) lumaNN = lumaSS;
+	float gradientScaled = gradient * 1.0/4.0;
+	float lumaMM = lumaM - lumaNN * 0.5;
+	float subpixF = subpixD * subpixE;
+	bool lumaMLTZero = lumaMM < 0.0;
+	lumaEndN -= lumaNN * 0.5;
+	lumaEndP -= lumaNN * 0.5;
+	bool doneN = abs(lumaEndN) >= gradientScaled;
+	bool doneP = abs(lumaEndP) >= gradientScaled;
+	if(!doneN) posN.x -= offNP.x * FXAA_QUALITY_P1;
+	if(!doneN) posN.y -= offNP.y * FXAA_QUALITY_P1;
+	bool doneNP = (!doneN) || (!doneP);
+	if(!doneP) posP.x += offNP.x * FXAA_QUALITY_P1;
+	if(!doneP) posP.y += offNP.y * FXAA_QUALITY_P1;
+
+	if(doneNP) {
+	if(!doneN) lumaEndN = FxaaLuma(FxaaTexTop(tex, posN.xy));
+	if(!doneP) lumaEndP = FxaaLuma(FxaaTexTop(tex, posP.xy));
+	if(!doneN) lumaEndN = lumaEndN - lumaNN * 0.5;
+	if(!doneP) lumaEndP = lumaEndP - lumaNN * 0.5;
+	doneN = abs(lumaEndN) >= gradientScaled;
+	doneP = abs(lumaEndP) >= gradientScaled;
+	if(!doneN) posN.x -= offNP.x * FXAA_QUALITY_P2;
+	if(!doneN) posN.y -= offNP.y * FXAA_QUALITY_P2;
+	doneNP = (!doneN) || (!doneP);
+	if(!doneP) posP.x += offNP.x * FXAA_QUALITY_P2;
+	if(!doneP) posP.y += offNP.y * FXAA_QUALITY_P2;
+
+	if(doneNP) {
+	if(!doneN) lumaEndN = FxaaLuma(FxaaTexTop(tex, posN.xy));
+	if(!doneP) lumaEndP = FxaaLuma(FxaaTexTop(tex, posP.xy));
+	if(!doneN) lumaEndN = lumaEndN - lumaNN * 0.5;
+	if(!doneP) lumaEndP = lumaEndP - lumaNN * 0.5;
+	doneN = abs(lumaEndN) >= gradientScaled;
+	doneP = abs(lumaEndP) >= gradientScaled;
+	if(!doneN) posN.x -= offNP.x * FXAA_QUALITY_P3;
+	if(!doneN) posN.y -= offNP.y * FXAA_QUALITY_P3;
+	doneNP = (!doneN) || (!doneP);
+	if(!doneP) posP.x += offNP.x * FXAA_QUALITY_P3;
+	if(!doneP) posP.y += offNP.y * FXAA_QUALITY_P3;
+
+	if(doneNP) {
+	if(!doneN) lumaEndN = FxaaLuma(FxaaTexTop(tex, posN.xy));
+	if(!doneP) lumaEndP = FxaaLuma(FxaaTexTop(tex, posP.xy));
+	if(!doneN) lumaEndN = lumaEndN - lumaNN * 0.5;
+	if(!doneP) lumaEndP = lumaEndP - lumaNN * 0.5;
+	doneN = abs(lumaEndN) >= gradientScaled;
+	doneP = abs(lumaEndP) >= gradientScaled;
+	if(!doneN) posN.x -= offNP.x * FXAA_QUALITY_P4;
+	if(!doneN) posN.y -= offNP.y * FXAA_QUALITY_P4;
+	doneNP = (!doneN) || (!doneP);
+	if(!doneP) posP.x += offNP.x * FXAA_QUALITY_P4;
+	if(!doneP) posP.y += offNP.y * FXAA_QUALITY_P4;
+
+	if(doneNP) {
+	if(!doneN) lumaEndN = FxaaLuma(FxaaTexTop(tex, posN.xy));
+	if(!doneP) lumaEndP = FxaaLuma(FxaaTexTop(tex, posP.xy));
+	if(!doneN) lumaEndN = lumaEndN - lumaNN * 0.5;
+	if(!doneP) lumaEndP = lumaEndP - lumaNN * 0.5;
+	doneN = abs(lumaEndN) >= gradientScaled;
+	doneP = abs(lumaEndP) >= gradientScaled;
+	if(!doneN) posN.x -= offNP.x * FXAA_QUALITY_P5;
+	if(!doneN) posN.y -= offNP.y * FXAA_QUALITY_P5;
+	doneNP = (!doneN) || (!doneP);
+	if(!doneP) posP.x += offNP.x * FXAA_QUALITY_P5;
+	if(!doneP) posP.y += offNP.y * FXAA_QUALITY_P5;
+
+	if(doneNP) {
+	if(!doneN) lumaEndN = FxaaLuma(FxaaTexTop(tex, posN.xy));
+	if(!doneP) lumaEndP = FxaaLuma(FxaaTexTop(tex, posP.xy));
+	if(!doneN) lumaEndN = lumaEndN - lumaNN * 0.5;
+	if(!doneP) lumaEndP = lumaEndP - lumaNN * 0.5;
+	doneN = abs(lumaEndN) >= gradientScaled;
+	doneP = abs(lumaEndP) >= gradientScaled;
+	if(!doneN) posN.x -= offNP.x * FXAA_QUALITY_P6;
+	if(!doneN) posN.y -= offNP.y * FXAA_QUALITY_P6;
+	doneNP = (!doneN) || (!doneP);
+	if(!doneP) posP.x += offNP.x * FXAA_QUALITY_P6;
+	if(!doneP) posP.y += offNP.y * FXAA_QUALITY_P6;
+
+	if(doneNP) {
+	if(!doneN) lumaEndN = FxaaLuma(FxaaTexTop(tex, posN.xy));
+	if(!doneP) lumaEndP = FxaaLuma(FxaaTexTop(tex, posP.xy));
+	if(!doneN) lumaEndN = lumaEndN - lumaNN * 0.5;
+	if(!doneP) lumaEndP = lumaEndP - lumaNN * 0.5;
+	doneN = abs(lumaEndN) >= gradientScaled;
+	doneP = abs(lumaEndP) >= gradientScaled;
+	if(!doneN) posN.x -= offNP.x * FXAA_QUALITY_P7;
+	if(!doneN) posN.y -= offNP.y * FXAA_QUALITY_P7;
+	doneNP = (!doneN) || (!doneP);
+	if(!doneP) posP.x += offNP.x * FXAA_QUALITY_P7;
+	if(!doneP) posP.y += offNP.y * FXAA_QUALITY_P7;
+
+	if(doneNP) {
+	if(!doneN) lumaEndN = FxaaLuma(FxaaTexTop(tex, posN.xy));
+	if(!doneP) lumaEndP = FxaaLuma(FxaaTexTop(tex, posP.xy));
+	if(!doneN) lumaEndN = lumaEndN - lumaNN * 0.5;
+	if(!doneP) lumaEndP = lumaEndP - lumaNN * 0.5;
+	doneN = abs(lumaEndN) >= gradientScaled;
+	doneP = abs(lumaEndP) >= gradientScaled;
+	if(!doneN) posN.x -= offNP.x * FXAA_QUALITY_P8;
+	if(!doneN) posN.y -= offNP.y * FXAA_QUALITY_P8;
+	doneNP = (!doneN) || (!doneP);
+	if(!doneP) posP.x += offNP.x * FXAA_QUALITY_P8;
+	if(!doneP) posP.y += offNP.y * FXAA_QUALITY_P8;
+
+	if(doneNP) {
+	if(!doneN) lumaEndN = FxaaLuma(FxaaTexTop(tex, posN.xy));
+	if(!doneP) lumaEndP = FxaaLuma(FxaaTexTop(tex, posP.xy));
+	if(!doneN) lumaEndN = lumaEndN - lumaNN * 0.5;
+	if(!doneP) lumaEndP = lumaEndP - lumaNN * 0.5;
+	doneN = abs(lumaEndN) >= gradientScaled;
+	doneP = abs(lumaEndP) >= gradientScaled;
+	if(!doneN) posN.x -= offNP.x * FXAA_QUALITY_P9;
+	if(!doneN) posN.y -= offNP.y * FXAA_QUALITY_P9;
+	doneNP = (!doneN) || (!doneP);
+	if(!doneP) posP.x += offNP.x * FXAA_QUALITY_P9;
+	if(!doneP) posP.y += offNP.y * FXAA_QUALITY_P9;
+
+	if(doneNP) {
+	if(!doneN) lumaEndN = FxaaLuma(FxaaTexTop(tex, posN.xy));
+	if(!doneP) lumaEndP = FxaaLuma(FxaaTexTop(tex, posP.xy));
+	if(!doneN) lumaEndN = lumaEndN - lumaNN * 0.5;
+	if(!doneP) lumaEndP = lumaEndP - lumaNN * 0.5;
+	doneN = abs(lumaEndN) >= gradientScaled;
+	doneP = abs(lumaEndP) >= gradientScaled;
+	if(!doneN) posN.x -= offNP.x * FXAA_QUALITY_P10;
+	if(!doneN) posN.y -= offNP.y * FXAA_QUALITY_P10;
+	doneNP = (!doneN) || (!doneP);
+	if(!doneP) posP.x += offNP.x * FXAA_QUALITY_P10;
+	if(!doneP) posP.y += offNP.y * FXAA_QUALITY_P10;
+
+	if(doneNP) {
+	if(!doneN) lumaEndN = FxaaLuma(FxaaTexTop(tex, posN.xy));
+	if(!doneP) lumaEndP = FxaaLuma(FxaaTexTop(tex, posP.xy));
+	if(!doneN) lumaEndN = lumaEndN - lumaNN * 0.5;
+	if(!doneP) lumaEndP = lumaEndP - lumaNN * 0.5;
+	doneN = abs(lumaEndN) >= gradientScaled;
+	doneP = abs(lumaEndP) >= gradientScaled;
+	if(!doneN) posN.x -= offNP.x * FXAA_QUALITY_P11;
+	if(!doneN) posN.y -= offNP.y * FXAA_QUALITY_P11;
+	doneNP = (!doneN) || (!doneP);
+	if(!doneP) posP.x += offNP.x * FXAA_QUALITY_P11;
+	if(!doneP) posP.y += offNP.y * FXAA_QUALITY_P11;
+
+	if(doneNP) {
+	if(!doneN) lumaEndN = FxaaLuma(FxaaTexTop(tex, posN.xy));
+	if(!doneP) lumaEndP = FxaaLuma(FxaaTexTop(tex, posP.xy));
+	if(!doneN) lumaEndN = lumaEndN - lumaNN * 0.5;
+	if(!doneP) lumaEndP = lumaEndP - lumaNN * 0.5;
+	doneN = abs(lumaEndN) >= gradientScaled;
+	doneP = abs(lumaEndP) >= gradientScaled;
+	if(!doneN) posN.x -= offNP.x * FXAA_QUALITY_P12;
+	if(!doneN) posN.y -= offNP.y * FXAA_QUALITY_P12;
+	doneNP = (!doneN) || (!doneP);
+	if(!doneP) posP.x += offNP.x * FXAA_QUALITY_P12;
+	if(!doneP) posP.y += offNP.y * FXAA_QUALITY_P12;
+	}}}}}}}}}}}
+
+	float dstN = posM.x - posN.x;
+	float dstP = posP.x - posM.x;
+	if(!horzSpan) dstN = posM.y - posN.y;
+	if(!horzSpan) dstP = posP.y - posM.y;
+
+	bool goodSpanN = (lumaEndN < 0.0) != lumaMLTZero;
+	float spanLength = (dstP + dstN);
+	bool goodSpanP = (lumaEndP < 0.0) != lumaMLTZero;
+	float spanLengthRcp = 1.0/spanLength;
+
+	bool directionN = dstN < dstP;
+	float dst = min(dstN, dstP);
+	bool goodSpan = directionN ? goodSpanN : goodSpanP;
+	float subpixG = subpixF * subpixF;
+	float pixelOffset = (dst * (-spanLengthRcp)) + 0.5;
+	float subpixH = subpixG * fxaaSubpix;
+
+	float pixelOffsetGood = goodSpan ? pixelOffset : 0.0;
+	float pixelOffsetSubpix = max(pixelOffsetGood, subpixH);
+	if(!horzSpan) posM.x += pixelOffsetSubpix * lengthSign;
+	if( horzSpan) posM.y += pixelOffsetSubpix * lengthSign;
+
+	return float4(FxaaTexTop(tex, posM).xyz, lumaM);
+}
+
+#if (FXAA_GLSL_130 == 1 || FXAA_GLSL_VK == 1)
+float4 FxaaPass(float4 FxaaColor, float2 uv0)
+#elif (FXAA_HLSL == 1)
+float4 FxaaPass(float4 FxaaColor : COLOR0, float2 uv0 : TEXCOORD0)
+#elif defined(__METAL_VERSION__)
+float4 FxaaPass(float4 FxaaColor, float2 uv0, texture2d<float> tex)
+#endif
+{
+
+	#if (FXAA_HLSL == 1)
+	FxaaTex tex;
+	tex.tex = Texture;
+	tex.smpl = TextureSampler;
+
+	float2 PixelSize;
+	Texture.GetDimensions(PixelSize.x, PixelSize.y);
+	FxaaColor = FxaaPixelShader(uv0, tex, 1.0/PixelSize.xy, FxaaSubpixMax, FxaaEdgeThreshold, FxaaEdgeThresholdMin);
+
+	#elif (FXAA_GLSL_130 == 1 || FXAA_GLSL_VK == 1)
+	vec2 PixelSize = vec2(textureSize(TextureSampler, 0));
+	FxaaColor = FxaaPixelShader(uv0, TextureSampler, 1.0/PixelSize.xy, FxaaSubpixMax, FxaaEdgeThreshold, FxaaEdgeThresholdMin);
+	#elif defined(__METAL_VERSION__)
+	float2 PixelSize = float2(tex.get_width(), tex.get_height());
+	FxaaColor = FxaaPixelShader(uv0, tex, 1.f/PixelSize, FxaaSubpixMax, FxaaEdgeThreshold, FxaaEdgeThresholdMin);
+	#endif
+
+	return FxaaColor;
+}
+
+/*------------------------------------------------------------------------------
+                      [MAIN() & COMBINE PASS CODE SECTION]
+------------------------------------------------------------------------------*/
+#if (FXAA_GLSL_130 == 1 || FXAA_GLSL_VK == 1)
+
+void main()
+{
+	vec4 color = texture(TextureSampler, PSin_t);
+	color      = PreGammaPass(color);
+	color      = FxaaPass(color, PSin_t);
+
+	SV_Target0 = float4(color.rgb, 1.0);
+}
+
+#elif (FXAA_HLSL == 1)
+PS_OUTPUT main(VS_OUTPUT input)
+{
+	PS_OUTPUT output;
+
+	float4 color = Texture.Sample(TextureSampler, input.t);
+
+	color = PreGammaPass(color);
+	color = FxaaPass(color, input.t);
+
+	output.c = float4(color.rgb, 1.0);
+	
+	return output;
+}
+
+// Metal main function in in fxaa.metal
+#endif

bin/resources/shaders/dx11/cas.hlsl

@@ -0,0 +1,75 @@
+// Based on CAS_Shader.hlsl
+//
+// Copyright(c) 2019 Advanced Micro Devices, Inc.All rights reserved.
+// Permission is hereby granted, free of charge, to any person obtaining a copy
+// of this software and associated documentation files(the "Software"), to deal
+// in the Software without restriction, including without limitation the rights
+// to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+// copies of the Software, and to permit persons to whom the Software is
+// furnished to do so, subject to the following conditions :
+// The above copyright notice and this permission notice shall be included in
+// all copies or substantial portions of the Software.
+// THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+// IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+// FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.IN NO EVENT SHALL THE
+// AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+// LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+// OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
+// THE SOFTWARE.
+
+cbuffer cb : register(b0)
+{
+  uint4 const0;
+  uint4 const1;
+  int2 srcOffset;
+};
+
+Texture2D InputTexture : register(t0);
+RWTexture2D<float4> OutputTexture : register(u0);
+
+#define A_GPU 1
+#define A_HLSL 1
+
+#include "ffx_a.h"
+
+AF3 CasLoad(ASU2 p)
+{
+  return InputTexture.Load(int3(srcOffset, 0) + int3(p, 0)).rgb;
+}
+
+// Lets you transform input from the load into a linear color space between 0 and 1. See ffx_cas.h
+// In this case, our input is already linear and between 0 and 1
+void CasInput(inout AF1 r, inout AF1 g, inout AF1 b) {}
+
+#include "ffx_cas.h"
+
+[numthreads(64, 1, 1)]
+void main(uint3 LocalThreadId : SV_GroupThreadID, uint3 WorkGroupId : SV_GroupID)
+{
+  // Do remapping of local xy in workgroup for a more PS-like swizzle pattern.
+  AU2 gxy = ARmp8x8(LocalThreadId.x) + AU2(WorkGroupId.x << 4u, WorkGroupId.y << 4u);
+
+#if CAS_SHARPEN_ONLY
+  const bool sharpenOnly = true;
+#else
+  const bool sharpenOnly = false;
+#endif
+
+  // Filter.
+  AF3 c = (float3)0.0f;
+
+  CasFilter(c.r, c.g, c.b, gxy, const0, const1, sharpenOnly);
+  OutputTexture[ASU2(gxy)] = AF4(c, 1);
+  gxy.x += 8u;
+
+  CasFilter(c.r, c.g, c.b, gxy, const0, const1, sharpenOnly);
+  OutputTexture[ASU2(gxy)] = AF4(c, 1);
+  gxy.y += 8u;
+
+  CasFilter(c.r, c.g, c.b, gxy, const0, const1, sharpenOnly);
+  OutputTexture[ASU2(gxy)] = AF4(c, 1);
+  gxy.x -= 8u;
+
+  CasFilter(c.r, c.g, c.b, gxy, const0, const1, sharpenOnly);
+  OutputTexture[ASU2(gxy)] = AF4(c, 1);
+}

bin/resources/shaders/dx11/convert.fx

@@ -0,0 +1,583 @@
+// SPDX-FileCopyrightText: 2002-2025 PCSX2 Dev Team
+// SPDX-License-Identifier: GPL-3.0+
+
+struct VS_INPUT
+{
+	float4 p : POSITION;
+	float2 t : TEXCOORD0;
+	float4 c : COLOR;
+};
+
+struct VS_OUTPUT
+{
+	float4 p : SV_Position;
+	float2 t : TEXCOORD0;
+	float4 c : COLOR;
+};
+
+cbuffer cb0 : register(b0)
+{
+	float4 BGColor;
+	int EMODA;
+	int EMODC;
+	int DOFFSET;
+};
+
+static const float3x3 rgb2yuv =
+{
+	{0.587, 0.114, 0.299},
+	{-0.311, 0.500, -0.169},
+	{-0.419, -0.081, 0.500}
+};
+
+Texture2D Texture;
+SamplerState TextureSampler;
+
+float4 sample_c(float2 uv)
+{
+	return Texture.Sample(TextureSampler, uv);
+}
+
+struct PS_INPUT
+{
+	float4 p : SV_Position;
+	float2 t : TEXCOORD0;
+	float4 c : COLOR;
+};
+
+struct PS_OUTPUT
+{
+	float4 c : SV_Target0;
+};
+
+VS_OUTPUT vs_main(VS_INPUT input)
+{
+	VS_OUTPUT output;
+
+	output.p = input.p;
+	output.t = input.t;
+	output.c = input.c;
+
+	return output;
+}
+
+PS_OUTPUT ps_copy(PS_INPUT input)
+{
+	PS_OUTPUT output;
+	
+	output.c = sample_c(input.t);
+
+	return output;
+}
+
+float ps_depth_copy(PS_INPUT input) : SV_Depth
+{
+	return sample_c(input.t).r;
+}
+
+PS_OUTPUT ps_downsample_copy(PS_INPUT input)
+{
+	int DownsampleFactor = DOFFSET;
+	int2 ClampMin = int2(EMODA, EMODC);
+	float Weight = BGColor.x;
+
+	int2 coord = max(int2(input.p.xy) * DownsampleFactor, ClampMin);
+
+	PS_OUTPUT output;
+	output.c = (float4)0;
+	for (int yoff = 0; yoff < DownsampleFactor; yoff++)
+	{
+		for (int xoff = 0; xoff < DownsampleFactor; xoff++)
+			output.c += Texture.Load(int3(coord + int2(xoff, yoff), 0));
+	}
+	output.c /= Weight;
+	return output;
+}
+
+PS_OUTPUT ps_filter_transparency(PS_INPUT input)
+{
+	PS_OUTPUT output;
+	float4 c = sample_c(input.t);
+	output.c = float4(c.rgb, 1.0);
+	return output;
+}
+
+// Need to be careful with precision here, it can break games like Spider-Man 3 and Dogs Life
+uint ps_convert_rgba8_16bits(PS_INPUT input) : SV_Target0
+{
+	uint4 i = sample_c(input.t) * float4(255.5f, 255.5f, 255.5f, 255.5f);
+
+	return ((i.x & 0x00F8u) >> 3) | ((i.y & 0x00F8u) << 2) | ((i.z & 0x00f8u) << 7) | ((i.w & 0x80u) << 8);
+}
+
+void ps_datm1(PS_INPUT input)
+{
+	clip(sample_c(input.t).a - 127.5f / 255); // >= 0x80 pass
+}
+
+void ps_datm0(PS_INPUT input)
+{
+	clip(127.5f / 255 - sample_c(input.t).a); // < 0x80 pass (== 0x80 should not pass)
+}
+
+void ps_datm1_rta_correction(PS_INPUT input)
+{
+	clip(sample_c(input.t).a - 254.5f / 255); // >= 0x80 pass
+}
+
+void ps_datm0_rta_correction(PS_INPUT input)
+{
+	clip(254.5f / 255 - sample_c(input.t).a); // < 0x80 pass (== 0x80 should not pass)
+}
+
+PS_OUTPUT ps_rta_correction(PS_INPUT input)
+{
+	PS_OUTPUT output;
+	float4 value = sample_c(input.t);
+	output.c = float4(value.rgb, value.a / (128.25f / 255.0f));
+	return output;
+}
+
+PS_OUTPUT ps_rta_decorrection(PS_INPUT input)
+{
+	PS_OUTPUT output;
+	float4 value = sample_c(input.t);
+	output.c = float4(value.rgb, value.a * (128.25f / 255.0f));
+	return output;
+}
+
+PS_OUTPUT ps_colclip_init(PS_INPUT input)
+{
+	PS_OUTPUT output;
+	float4 value = sample_c(input.t);
+	output.c = float4(round(value.rgb * 255) / 65535, value.a);
+	return output;
+}
+
+PS_OUTPUT ps_colclip_resolve(PS_INPUT input)
+{
+	PS_OUTPUT output;
+	float4 value = sample_c(input.t);
+	output.c = float4(float3(uint3(value.rgb * 65535.5) & 255) / 255, value.a);
+	return output;
+}
+
+uint ps_convert_float32_32bits(PS_INPUT input) : SV_Target0
+{
+	// Convert a FLOAT32 depth texture into a 32 bits UINT texture
+	return uint(exp2(32.0f) * sample_c(input.t).r);
+}
+
+PS_OUTPUT ps_convert_float32_rgba8(PS_INPUT input)
+{
+	PS_OUTPUT output;
+
+	// Convert a FLOAT32 depth texture into a RGBA color texture
+	uint d = uint(sample_c(input.t).r * exp2(32.0f));
+	output.c = float4(uint4((d & 0xFFu), ((d >> 8) & 0xFFu), ((d >> 16) & 0xFFu), (d >> 24))) / 255.0f;
+
+	return output;
+}
+
+PS_OUTPUT ps_convert_float16_rgb5a1(PS_INPUT input)
+{
+	PS_OUTPUT output;
+
+	// Convert a FLOAT32 (only 16 lsb) depth into a RGB5A1 color texture
+	uint d = uint(sample_c(input.t).r * exp2(32.0f));
+	output.c = float4(uint4(d << 3, d >> 2, d >> 7, d >> 8) & uint4(0xf8, 0xf8, 0xf8, 0x80)) / 255.0f;
+	return output;
+}
+
+float rgba8_to_depth32(float4 val)
+{
+	uint4 c = uint4(val * 255.5f);
+	return float(c.r | (c.g << 8) | (c.b << 16) | (c.a << 24)) * exp2(-32.0f);
+}
+
+float rgba8_to_depth24(float4 val)
+{
+	uint3 c = uint3(val.rgb * 255.5f);
+	return float(c.r | (c.g << 8) | (c.b << 16)) * exp2(-32.0f);
+}
+
+float rgba8_to_depth16(float4 val)
+{
+	uint2 c = uint2(val.rg * 255.5f);
+	return float(c.r | (c.g << 8)) * exp2(-32.0f);
+}
+
+float rgb5a1_to_depth16(float4 val)
+{
+	uint4 c = uint4(val * 255.5f);
+	return float(((c.r & 0xF8u) >> 3) | ((c.g & 0xF8u) << 2) | ((c.b & 0xF8u) << 7) | ((c.a & 0x80u) << 8)) * exp2(-32.0f);
+}
+
+float ps_convert_float32_float24(PS_INPUT input) : SV_Depth
+{
+	// Truncates depth value to 24bits
+	uint d = uint(sample_c(input.t).r * exp2(32.0f)) & 0xFFFFFFu;
+	return float(d) * exp2(-32.0f);
+}
+
+float ps_convert_rgba8_float32(PS_INPUT input) : SV_Depth
+{
+	// Convert an RGBA texture into a float depth texture
+	return rgba8_to_depth32(sample_c(input.t));
+}
+
+float ps_convert_rgba8_float24(PS_INPUT input) : SV_Depth
+{
+	// Same as above but without the alpha channel (24 bits Z)
+
+	// Convert an RGBA texture into a float depth texture
+	return rgba8_to_depth24(sample_c(input.t));
+}
+
+float ps_convert_rgba8_float16(PS_INPUT input) : SV_Depth
+{
+	// Same as above but without the A/B channels (16 bits Z)
+
+	// Convert an RGBA texture into a float depth texture
+	return rgba8_to_depth16(sample_c(input.t));
+}
+
+float ps_convert_rgb5a1_float16(PS_INPUT input) : SV_Depth
+{
+	// Convert an RGB5A1 (saved as RGBA8) color to a 16 bit Z
+	return rgb5a1_to_depth16(sample_c(input.t));
+}
+
+#define SAMPLE_RGBA_DEPTH_BILN(CONVERT_FN) \
+	uint width, height; \
+	Texture.GetDimensions(width, height); \
+	float2 top_left_f = input.t * float2(width, height) - 0.5f; \
+	int2 top_left = int2(floor(top_left_f)); \
+	int4 coords = clamp(int4(top_left, top_left + 1), int4(0, 0, 0, 0), int2(width - 1, height - 1).xyxy); \
+	float2 mix_vals = frac(top_left_f); \
+	float depthTL = CONVERT_FN(Texture.Load(int3(coords.xy, 0))); \
+	float depthTR = CONVERT_FN(Texture.Load(int3(coords.zy, 0))); \
+	float depthBL = CONVERT_FN(Texture.Load(int3(coords.xw, 0))); \
+	float depthBR = CONVERT_FN(Texture.Load(int3(coords.zw, 0))); \
+	return lerp(lerp(depthTL, depthTR, mix_vals.x), lerp(depthBL, depthBR, mix_vals.x), mix_vals.y);
+
+float ps_convert_rgba8_float32_biln(PS_INPUT input) : SV_Depth
+{
+	// Convert an RGBA texture into a float depth texture
+	SAMPLE_RGBA_DEPTH_BILN(rgba8_to_depth32);
+}
+
+float ps_convert_rgba8_float24_biln(PS_INPUT input) : SV_Depth
+{
+	// Same as above but without the alpha channel (24 bits Z)
+
+	// Convert an RGBA texture into a float depth texture
+	SAMPLE_RGBA_DEPTH_BILN(rgba8_to_depth24);
+}
+
+float ps_convert_rgba8_float16_biln(PS_INPUT input) : SV_Depth
+{
+	// Same as above but without the A/B channels (16 bits Z)
+
+	// Convert an RGBA texture into a float depth texture
+	SAMPLE_RGBA_DEPTH_BILN(rgba8_to_depth16);
+}
+
+float ps_convert_rgb5a1_float16_biln(PS_INPUT input) : SV_Depth
+{
+	// Convert an RGB5A1 (saved as RGBA8) color to a 16 bit Z
+	SAMPLE_RGBA_DEPTH_BILN(rgb5a1_to_depth16);
+}
+
+PS_OUTPUT ps_convert_rgb5a1_8i(PS_INPUT input)
+{
+	PS_OUTPUT output;
+
+	// Convert a RGB5A1 texture into a 8 bits packed texture
+	// Input column: 16x2 RGB5A1 pixels
+	// 0: 16 RGBA
+	// 1: 16 RGBA
+	// Output column: 16x4 Index pixels
+	// 0: 16 R5G2
+	// 1: 16 R5G2
+	// 2: 16 G2B5A1
+	// 3: 16 G2B5A1
+	uint2 pos = uint2(input.p.xy);
+
+	// Collapse separate R G B A areas into their base pixel
+	uint2 column = (pos & ~uint2(0u, 3u)) / uint2(1,2);
+	uint2 subcolumn = (pos & uint2(0u, 1u));
+	column.x -= (column.x / 128) * 64;
+	column.y += (column.y / 32) * 32;
+	
+	uint PSM = uint(DOFFSET);
+	
+	// Deal with swizzling differences
+	if ((PSM & 0x8) != 0) // PSMCT16S
+	{
+		if ((pos.x & 32) != 0)
+		{
+			column.y += 32; // 4 columns high times 4 to get bottom 4 blocks
+			column.x &= ~32;
+		}
+		
+		if ((pos.x & 64) != 0)
+		{
+			column.x -= 32;
+		}
+		
+		if (((pos.x & 16) != 0) != ((pos.y & 16) != 0))
+		{
+			column.x ^= 16; 
+			column.y ^= 8;
+		}
+		
+		if ((PSM & 0x30) != 0) // PSMZ16S - Untested but hopefully ok if anything uses it.
+		{
+			column.x ^= 32;
+			column.y ^= 16;
+		}
+	}
+	else // PSMCT16
+	{
+		if ((pos.y & 32) != 0)
+		{
+			column.y -= 16;
+			column.x += 32;
+		}
+		
+		if ((pos.x & 96) != 0)
+		{
+			uint multi = (pos.x & 96) / 32;
+			column.y += 16 * multi; // 4 columns high times 4 to get bottom 4 blocks
+			column.x -= (pos.x & 96);
+		}
+		
+		if (((pos.x & 16) != 0) != ((pos.y & 16) != 0))
+		{
+			column.x ^= 16; 
+			column.y ^= 8;
+		}
+		
+		if ((PSM & 0x30) != 0) // PSMZ16 - Untested but hopefully ok if anything uses it.
+		{
+			column.x ^= 32;
+			column.y ^= 32;
+		}
+	}
+	
+	uint2 coord = column | subcolumn;
+
+	// Compensate for potentially differing page pitch.
+	uint SBW = uint(EMODA);
+	uint DBW = uint(EMODC);
+	uint2 block_xy = coord / uint2(64,64);
+	uint block_num = (block_xy.y * (DBW / 128)) + block_xy.x;
+	uint2 block_offset = uint2((block_num % (SBW / 64)) * 64, (block_num / (SBW / 64)) * 64);
+	coord = (coord % uint2(64, 64)) + block_offset;
+
+	// Apply offset to cols 1 and 2
+	uint is_col23 = pos.y & 4u;
+	uint is_col13 = pos.y & 2u;
+	uint is_col12 = is_col23 ^ (is_col13 << 1);
+	coord.x ^= is_col12; // If cols 1 or 2, flip bit 3 of x
+
+	float ScaleFactor = BGColor.x;
+	if (floor(ScaleFactor) != ScaleFactor)
+		coord = uint2(float2(coord) * ScaleFactor);
+	else
+		coord *= uint(ScaleFactor);
+
+	float4 pixel = Texture.Load(int3(int2(coord), 0));
+	uint4 denorm_c = (uint4)(pixel * 255.5f);
+	if ((pos.y & 2u) == 0u)
+	{
+		uint red = (denorm_c.r >> 3) & 0x1Fu;
+		uint green = (denorm_c.g >> 3) & 0x1Fu;
+		float sel0 = (float)(((green << 5) | red) & 0xFF) / 255.0f;
+		
+		output.c = (float4)(sel0);
+	}
+	else
+	{
+		uint green = (denorm_c.g >> 3) & 0x1Fu;
+		uint blue = (denorm_c.b >> 3) & 0x1Fu;
+		uint alpha = denorm_c.a & 0x80u;
+		float sel0 = (float)((alpha | (blue << 2) | (green >> 3)) & 0xFF) / 255.0f;
+
+		output.c = (float4)(sel0);
+	}
+	return output;
+}
+
+PS_OUTPUT ps_convert_rgba_8i(PS_INPUT input)
+{
+	PS_OUTPUT output;
+
+	// Convert a RGBA texture into a 8 bits packed texture
+	// Input column: 8x2 RGBA pixels
+	// 0: 8 RGBA
+	// 1: 8 RGBA
+	// Output column: 16x4 Index pixels
+	// 0: 8 R | 8 B
+	// 1: 8 R | 8 B
+	// 2: 8 G | 8 A
+	// 3: 8 G | 8 A
+	uint2 pos = uint2(input.p.xy);
+
+	// Collapse separate R G B A areas into their base pixel
+	uint2 block = (pos & ~uint2(15u, 3u)) >> 1;
+	uint2 subblock = pos & uint2(7u, 1u);
+	uint2 coord = block | subblock;
+
+	// Compensate for potentially differing page pitch.
+	uint SBW = uint(EMODA);
+	uint DBW = uint(EMODC);
+	uint2 block_xy = coord / uint2(64, 32);
+	uint block_num = (block_xy.y * (DBW / 128)) + block_xy.x;
+	uint2 block_offset = uint2((block_num % (SBW / 64)) * 64, (block_num / (SBW / 64)) * 32);
+	coord = (coord % uint2(64, 32)) + block_offset;
+
+	// Apply offset to cols 1 and 2
+	uint is_col23 = pos.y & 4u;
+	uint is_col13 = pos.y & 2u;
+	uint is_col12 = is_col23 ^ (is_col13 << 1);
+	coord.x ^= is_col12; // If cols 1 or 2, flip bit 3 of x
+
+	float ScaleFactor = BGColor.x;
+	if (floor(ScaleFactor) != ScaleFactor)
+		coord = uint2(float2(coord) * ScaleFactor);
+	else
+		coord *= uint(ScaleFactor);
+
+	float4 pixel = Texture.Load(int3(int2(coord), 0));
+	float2 sel0 = (pos.y & 2u) == 0u ? pixel.rb : pixel.ga;
+	float  sel1 = (pos.x & 8u) == 0u ? sel0.x : sel0.y;
+	output.c = (float4)(sel1); // Divide by something here?
+	return output;
+}
+
+PS_OUTPUT ps_convert_clut_4(PS_INPUT input)
+{
+	// Borrowing the YUV constant buffer.
+	float scale = BGColor.x;
+	uint2 offset = uint2(uint(EMODA), uint(EMODC)) + uint(DOFFSET);
+
+	// CLUT4 is easy, just two rows of 8x8.
+	uint index = uint(input.p.x);
+	uint2 pos = uint2(index % 8u, index / 8u);
+
+	int2 final = int2(floor(float2(offset + pos) * scale));
+	PS_OUTPUT output;
+	output.c = Texture.Load(int3(final, 0), 0);
+	return output;
+}
+
+PS_OUTPUT ps_convert_clut_8(PS_INPUT input)
+{
+	float scale = BGColor.x;
+	uint2 offset = uint2(uint(EMODA), uint(EMODC));
+	uint index = min(uint(input.p.x) + uint(DOFFSET), 255u);
+
+	// CLUT is arranged into 8 groups of 16x2, with the top-right and bottom-left quadrants swapped.
+	// This can probably be done better..
+	uint subgroup = (index / 8u) % 4u;
+	uint2 pos;
+	pos.x = (index % 8u) + ((subgroup >= 2u) ? 8u : 0u);
+	pos.y = ((index / 32u) * 2u) + (subgroup % 2u);
+
+	int2 final = int2(floor(float2(offset + pos) * scale));
+	PS_OUTPUT output;
+	output.c = Texture.Load(int3(final, 0), 0);
+	return output;
+}
+
+PS_OUTPUT ps_yuv(PS_INPUT input)
+{
+	PS_OUTPUT output;
+
+	float4 i = sample_c(input.t);
+	float3 yuv = mul(rgb2yuv, i.gbr);
+
+	float Y = float(0xDB) / 255.0f * yuv.x + float(0x10) / 255.0f;
+	float Cr = float(0xE0) / 255.0f * yuv.y + float(0x80) / 255.0f;
+	float Cb = float(0xE0) / 255.0f * yuv.z + float(0x80) / 255.0f;
+
+	switch (EMODA)
+	{
+		case 0:
+			output.c.a = i.a;
+			break;
+		case 1:
+			output.c.a = Y;
+			break;
+		case 2:
+			output.c.a = Y / 2.0f;
+			break;
+		case 3:
+		default:
+			output.c.a = 0.0f;
+			break;
+	}
+
+	switch (EMODC)
+	{
+		case 0:
+			output.c.rgb = i.rgb;
+			break;
+		case 1:
+			output.c.rgb = float3(Y, Y, Y);
+			break;
+		case 2:
+			output.c.rgb = float3(Y, Cb, Cr);
+			break;
+		case 3:
+		default:
+			output.c.rgb = float3(i.a, i.a, i.a);
+			break;
+	}
+
+	return output;
+}
+
+float ps_stencil_image_init_0(PS_INPUT input) : SV_Target
+{
+	float c;
+	if ((127.5f / 255.0f) < sample_c(input.t).a) // < 0x80 pass (== 0x80 should not pass)
+		c = float(-1);
+	else
+		c = float(0x7FFFFFFF);
+	return c;
+}
+
+float ps_stencil_image_init_1(PS_INPUT input) : SV_Target
+{
+	float c;
+	if (sample_c(input.t).a < (127.5f / 255.0f)) // >= 0x80 pass
+		c = float(-1);
+	else
+		c = float(0x7FFFFFFF);
+	return c;
+}
+
+float ps_stencil_image_init_2(PS_INPUT input)
+	: SV_Target
+{
+	float c;
+	if ((254.5f / 255.0f) < sample_c(input.t).a) // < 0x80 pass (== 0x80 should not pass)
+		c = float(-1);
+	else
+		c = float(0x7FFFFFFF);
+	return c;
+}
+
+float ps_stencil_image_init_3(PS_INPUT input)
+	: SV_Target
+{
+	float c;
+	if (sample_c(input.t).a < (254.5f / 255.0f)) // >= 0x80 pass
+		c = float(-1);
+	else
+		c = float(0x7FFFFFFF);
+	return c;
+}

bin/resources/shaders/dx11/imgui.fx

@@ -0,0 +1,39 @@
+// SPDX-FileCopyrightText: 2002-2025 PCSX2 Dev Team
+// SPDX-License-Identifier: GPL-3.0+
+
+cbuffer vertexBuffer : register(b0)
+{
+	float4x4 ProjectionMatrix;
+};
+
+struct VS_INPUT
+{
+	float2 pos : POSITION;
+	float4 col : COLOR0;
+	float2 uv : TEXCOORD0;
+};
+
+struct PS_INPUT
+{
+	float4 pos : SV_POSITION;
+	float4 col : COLOR0;
+	float2 uv : TEXCOORD0;
+};
+
+PS_INPUT vs_main(VS_INPUT input)
+{
+	PS_INPUT output;
+	output.pos = mul(ProjectionMatrix, float4(input.pos.xy, 0.f, 1.f));
+	output.col = input.col;
+	output.uv = input.uv;
+	return output;
+}
+
+sampler sampler0 : register(s0);
+Texture2D texture0 : register(t0);
+
+float4 ps_main(PS_INPUT input) : SV_Target
+{
+	float4 out_col = input.col * texture0.Sample(sampler0, input.uv);
+	return out_col;
+}

bin/resources/shaders/dx11/interlace.fx

@@ -0,0 +1,183 @@
+// SPDX-FileCopyrightText: 2002-2025 PCSX2 Dev Team
+// SPDX-License-Identifier: GPL-3.0+
+
+Texture2D Texture;
+SamplerState Sampler;
+
+cbuffer cb0
+{
+	float4 ZrH;
+};
+
+struct PS_INPUT
+{
+	float4 p : SV_Position;
+	float2 t : TEXCOORD0;
+};
+
+
+// Weave shader
+float4 ps_main0(PS_INPUT input) : SV_Target0
+{
+	const int idx   = int(ZrH.x);     // buffer index passed from CPU
+	const int field = idx & 1;        // current field
+	const int vpos  = int(input.p.y); // vertical position of destination texture
+
+	if ((vpos & 1) == field)
+		return Texture.SampleLevel(Sampler, input.t, 0);
+	else
+		discard;
+
+	return float4(0.0f, 0.0f, 0.0f, 0.0f);
+}
+
+
+// Bob shader
+float4 ps_main1(PS_INPUT input) : SV_Target0
+{
+	return Texture.SampleLevel(Sampler, input.t, 0);
+}
+
+
+// Blend shader
+float4 ps_main2(PS_INPUT input) : SV_Target0
+{
+	float2 vstep = float2(0.0f, ZrH.y);
+	float4 c0 = Texture.SampleLevel(Sampler, input.t - vstep, 0);
+	float4 c1 = Texture.SampleLevel(Sampler, input.t, 0);
+	float4 c2 = Texture.SampleLevel(Sampler, input.t + vstep, 0);
+
+	return (c0 + c1 * 2 + c2) / 4;
+}
+
+
+// MAD shader - buffering
+float4 ps_main3(PS_INPUT input) : SV_Target0
+{
+	// We take half the lines from the current frame and stores them in the MAD frame buffer.
+	// the MAD frame buffer is split in 2 consecutive banks of 2 fields each, the fields in each bank
+	// are interleaved (top field at even lines and bottom field at odd lines). 
+	// When the source texture has an odd vres, the first line of bank 1 would be an odd index
+	// causing the wrong lines to be discarded, so a vertical offset (lofs) is added to the vertical
+	// position of the destination texture to force the proper field alignment
+
+	const int    idx    = int(ZrH.x);                                // buffer index passed from CPU
+	const int    bank   = idx >> 1;                                  // current bank
+	const int    field  = idx & 1;                                   // current field
+	const int    vres   = int(ZrH.z) >> 1;                           // vertical resolution of source texture
+	const int    lofs   = ((((vres + 1) >> 1) << 1) - vres) & bank;  // line alignment offset for bank 1
+	const int    vpos   = int(input.p.y) + lofs;                     // vertical position of destination texture
+
+	// if the index of current destination line belongs to the current fiels we update it, otherwise
+	// we leave the old line in the destination buffer
+	if ((vpos & 1) == field)
+		return Texture.SampleLevel(Sampler, input.t, 0);
+	else
+		discard;
+
+	return float4(0.0f, 0.0f, 0.0f, 0.0f);
+}
+
+
+// MAD shader - reconstruction
+float4 ps_main4(PS_INPUT input) : SV_Target0
+{
+	// we use the contents of the MAD frame buffer to reconstruct the missing lines from the current
+	// field.
+
+	const int    idx         = int(ZrH.x);                          // buffer index passed from CPU
+	const int    field       = idx & 1;                             // current field
+	const int    vpos        = int(input.p.y);                      // vertical position of destination texture
+	const float  sensitivity = ZrH.w;                               // passed from CPU, higher values mean more likely to use weave
+	const float3 motion_thr  = float3(1.0, 1.0, 1.0) * sensitivity; //
+	const float2 bofs        = float2(0.0f, 0.5f);                  // position of the bank 1 relative to source texture size
+	const float2 vscale      = float2(1.0f, 0.5f);                  // scaling factor from source to destination texture
+	const float2 lofs        = float2(0.0f, ZrH.y) * vscale;        // distance between two adjacent lines relative to source texture size
+	const float2 iptr        = input.t * vscale;                    // pointer to the current pixel in the source texture
+
+	float2 p_t0; // pointer to current pixel (missing or not) from most recent frame
+	float2 p_t1; // pointer to current pixel (missing or not) from one frame back
+	float2 p_t2; // pointer to current pixel (missing or not) from two frames back
+	float2 p_t3; // pointer to current pixel (missing or not) from three frames back
+
+	switch (idx)
+	{
+		case 1:
+			p_t0 = iptr;
+			p_t1 = iptr;
+			p_t2 = iptr + bofs;
+			p_t3 = iptr + bofs;
+			break;
+		case 2:
+			p_t0 = iptr + bofs;
+			p_t1 = iptr;
+			p_t2 = iptr;
+			p_t3 = iptr + bofs;
+			break;
+		case 3:
+			p_t0 = iptr + bofs;
+			p_t1 = iptr + bofs;
+			p_t2 = iptr;
+			p_t3 = iptr;
+			break;
+		default:
+			p_t0 = iptr;
+			p_t1 = iptr + bofs;
+			p_t2 = iptr + bofs;
+			p_t3 = iptr;
+			break;
+	}
+
+	// calculating motion, only relevant for missing lines where the "center line" is pointed by p_t1
+
+	float4 hn = Texture.SampleLevel(Sampler, p_t0 - lofs, 0); // new high pixel
+	float4 cn = Texture.SampleLevel(Sampler, p_t1, 0);        // new center pixel
+	float4 ln = Texture.SampleLevel(Sampler, p_t0 + lofs, 0); // new low pixel
+
+	float4 ho = Texture.SampleLevel(Sampler, p_t2 - lofs, 0); // old high pixel
+	float4 co = Texture.SampleLevel(Sampler, p_t3, 0);        // old center pixel
+	float4 lo = Texture.SampleLevel(Sampler, p_t2 + lofs, 0); // old low pixel
+
+	float3 mh = hn.rgb - ho.rgb; // high pixel motion
+	float3 mc = cn.rgb - co.rgb; // center pixel motion
+	float3 ml = ln.rgb - lo.rgb; // low pixel motion
+
+	mh = max(mh, -mh) - motion_thr;
+	mc = max(mc, -mc) - motion_thr;
+	ml = max(ml, -ml) - motion_thr;
+
+	#if 1 // use this code to evaluate each color motion separately
+		float mh_max = max(max(mh.x, mh.y), mh.z);
+		float mc_max = max(max(mc.x, mc.y), mc.z);
+		float ml_max = max(max(ml.x, ml.y), ml.z);
+	#else // use this code to evaluate average color motion
+		float mh_max = mh.x + mh.y + mh.z;
+		float mc_max = mc.x + mc.y + mc.z;
+		float ml_max = ml.x + ml.y + ml.z;
+	#endif
+
+	// selecting deinterlacing output
+
+	if ((vpos & 1) == field)
+	{
+		// output coordinate present on current field
+		return Texture.SampleLevel(Sampler, p_t0, 0);
+	}
+	else if ((iptr.y > 0.5f - lofs.y) || (iptr.y < 0.0 + lofs.y))
+	{
+		// top and bottom lines are always weaved
+		return cn;
+	}
+	else
+	{
+		// missing line needs to be reconstructed
+		if (((mh_max > 0.0f) || (ml_max > 0.0f)) || (mc_max > 0.0f))
+			// high motion -> interpolate pixels above and below
+			return (hn + ln) / 2.0f;
+		else
+			// low motion -> weave
+			return cn;
+	}
+
+	return float4(0.0f, 0.0f, 0.0f, 0.0f);
+}

bin/resources/shaders/dx11/merge.fx

@@ -0,0 +1,33 @@
+// SPDX-FileCopyrightText: 2002-2025 PCSX2 Dev Team
+// SPDX-License-Identifier: GPL-3.0+
+
+Texture2D Texture;
+SamplerState Sampler;
+
+cbuffer cb0 : register(b0)
+{
+	float4 BGColor;
+	int EMODA;
+	int EMODC;
+	int cb0_pad[2];
+};
+
+struct PS_INPUT
+{
+	float4 p : SV_Position;
+	float2 t : TEXCOORD0;
+};
+
+float4 ps_main0(PS_INPUT input) : SV_Target0
+{
+	float4 c = Texture.Sample(Sampler, input.t);
+	c.a *= 2.0f;
+	return c;
+}
+
+float4 ps_main1(PS_INPUT input) : SV_Target0
+{
+	float4 c = Texture.Sample(Sampler, input.t);
+	c.a = BGColor.a;
+	return c;
+}

bin/resources/shaders/dx11/present.fx

@@ -0,0 +1,480 @@
+// SPDX-FileCopyrightText: 2002-2025 PCSX2 Dev Team
+// SPDX-License-Identifier: GPL-3.0+
+
+struct VS_INPUT
+{
+	float4 p : POSITION;
+	float2 t : TEXCOORD0;
+	float4 c : COLOR;
+};
+
+struct VS_OUTPUT
+{
+	float4 p : SV_Position;
+	float2 t : TEXCOORD0;
+	float4 c : COLOR;
+};
+
+cbuffer cb0 : register(b0)
+{
+	float4 u_source_rect;
+	float4 u_target_rect;
+	float2 u_source_size;
+	float2 u_target_size;
+	float2 u_target_resolution;
+	float2 u_rcp_target_resolution; // 1 / u_target_resolution
+	float2 u_source_resolution;
+	float2 u_rcp_source_resolution; // 1 / u_source_resolution
+	float u_time;
+};
+
+Texture2D Texture;
+SamplerState TextureSampler;
+
+float4 sample_c(float2 uv)
+{
+	return Texture.Sample(TextureSampler, uv);
+}
+
+struct PS_INPUT
+{
+	float4 p : SV_Position;
+	float2 t : TEXCOORD0;
+	float4 c : COLOR;
+};
+
+struct PS_OUTPUT
+{
+	float4 c : SV_Target0;
+};
+
+VS_OUTPUT vs_main(VS_INPUT input)
+{
+	VS_OUTPUT output;
+
+	output.p = input.p;
+	output.t = input.t;
+	output.c = input.c;
+
+	return output;
+}
+
+PS_OUTPUT ps_copy(PS_INPUT input)
+{
+	PS_OUTPUT output;
+
+	output.c = sample_c(input.t);
+
+	return output;
+}
+
+float4 ps_crt(PS_INPUT input, int i)
+{
+	float4 mask[4] =
+		{
+			float4(1, 0, 0, 0),
+			float4(0, 1, 0, 0),
+			float4(0, 0, 1, 0),
+			float4(1, 1, 1, 0)
+		};
+
+	return sample_c(input.t) * saturate(mask[i] + 0.5f);
+}
+
+float4 ps_scanlines(PS_INPUT input, int i)
+{
+	float4 mask[2] =
+		{
+			float4(1, 1, 1, 0),
+			float4(0, 0, 0, 0)
+		};
+
+	return sample_c(input.t) * saturate(mask[i] + 0.5f);
+}
+
+PS_OUTPUT ps_filter_scanlines(PS_INPUT input)
+{
+	PS_OUTPUT output;
+
+	uint4 p = (uint4)input.p;
+
+	output.c = ps_scanlines(input, p.y % 2);
+
+	return output;
+}
+
+PS_OUTPUT ps_filter_diagonal(PS_INPUT input)
+{
+	PS_OUTPUT output;
+
+	uint4 p = (uint4)input.p;
+
+	output.c = ps_crt(input, (p.x + (p.y % 3)) % 3);
+
+	return output;
+}
+
+PS_OUTPUT ps_filter_triangular(PS_INPUT input)
+{
+	PS_OUTPUT output;
+
+	uint4 p = (uint4)input.p;
+
+	// output.c = ps_crt(input, ((p.x + (p.y & 1) * 3) >> 1) % 3);
+	output.c = ps_crt(input, ((p.x + ((p.y >> 1) & 1) * 3) >> 1) % 3);
+
+	return output;
+}
+
+static const float PI = 3.14159265359f;
+PS_OUTPUT ps_filter_complex(PS_INPUT input) // triangular
+{
+	PS_OUTPUT output;
+
+	float2 texdim; 
+	Texture.GetDimensions(texdim.x, texdim.y);
+
+	output.c = (0.9 - 0.4 * cos(2 * PI * input.t.y * texdim.y)) * sample_c(float2(input.t.x, (floor(input.t.y * texdim.y) + 0.5) / texdim.y));
+
+	return output;
+}
+
+//Lottes CRT
+#define MaskingType 4                      //[1|2|3|4] The type of CRT shadow masking used. 1: compressed TV style, 2: Aperture-grille, 3: Stretched VGA style, 4: VGA style.
+#define ScanBrightness -8.00               //[-16.0 to 1.0] The overall brightness of the scanline effect. Lower for darker, higher for brighter.
+#define FilterCRTAmount -3.00              //[-4.0 to 1.0] The amount of filtering used, to replicate the TV CRT look. Lower for less, higher for more.
+#define HorizontalWarp 0.00                //[0.0 to 0.1] The distortion warping effect for the horizontal (x) axis of the screen. Use small increments.
+#define VerticalWarp 0.00                  //[0.0 to 0.1] The distortion warping effect for the verticle (y) axis of the screen. Use small increments.
+#define MaskAmountDark 0.50                //[0.0 to 1.0] The value of the dark masking line effect used. Lower for darker lower end masking, higher for brighter.
+#define MaskAmountLight 1.50               //[0.0 to 2.0] The value of the light masking line effect used. Lower for darker higher end masking, higher for brighter.
+#define BloomPixel -1.50                   //[-2.0 -0.5] Pixel bloom radius. Higher for increased softness of bloom.
+#define BloomScanLine -2.0                 //[-4.0 -1.0] Scanline bloom radius. Higher for increased softness of bloom.
+#define BloomAmount 0.15                   //[0.0 1.0] Bloom intensity. Higher for brighter.
+#define Shape 2.0                          //[0.0 10.0] Kernal filter shape. Lower values will darken image and introduce moire patterns if used with curvature.
+#define UseShadowMask 1                    //[0 or 1] Enables, or disables the use of the CRT shadow mask. 0 is disabled, 1 is enabled.
+
+float ToLinear1(float c)
+{
+	return c <= 0.04045 ? c / 12.92 : pow((abs(c) + 0.055) / 1.055, 2.4);
+}
+
+float3 ToLinear(float3 c)
+{
+	return float3(ToLinear1(c.r), ToLinear1(c.g), ToLinear1(c.b));
+}
+
+float ToSrgb1(float c)
+{
+	return c < 0.0031308 ? c * 12.92 : 1.055 * pow(abs(c), 0.41666) - 0.055;
+}
+
+float3 ToSrgb(float3 c)
+{
+	return float3(ToSrgb1(c.r), ToSrgb1(c.g), ToSrgb1(c.b));
+}
+
+float3 Fetch(float2 pos, float2 off)
+{
+	pos = (floor(pos * u_target_size + off) + float2(0.5, 0.5)) / u_target_size;
+	if (max(abs(pos.x - 0.5), abs(pos.y - 0.5)) > 0.5)
+	{
+		return float3(0.0, 0.0, 0.0);
+	}
+	else
+	{
+		return ToLinear(Texture.Sample(TextureSampler, pos.xy).rgb);
+	}
+}
+
+float2 Dist(float2 pos)
+{
+	pos = pos * float2(640, 480);
+
+	return -((pos - floor(pos)) - float2(0.5, 0.5));
+}
+
+float Gaus(float pos, float scale)
+{
+	return exp2(scale * pos * pos);
+}
+
+float3 Horz3(float2 pos, float off)
+{
+	float3 b = Fetch(pos, float2(-1.0, off));
+	float3 c = Fetch(pos, float2(0.0, off));
+	float3 d = Fetch(pos, float2(1.0, off));
+	float dst = Dist(pos).x;
+
+	// Convert distance to weight.
+	float scale = FilterCRTAmount;
+	float wb = Gaus(dst - 1.0, scale);
+	float wc = Gaus(dst + 0.0, scale);
+	float wd = Gaus(dst + 1.0, scale);
+
+	return (b * wb + c * wc + d * wd) / (wb + wc + wd);
+}
+
+float3 Horz5(float2 pos, float off)
+{
+	float3 a = Fetch(pos, float2(-2.0, off));
+	float3 b = Fetch(pos, float2(-1.0, off));
+	float3 c = Fetch(pos, float2(0.0, off));
+	float3 d = Fetch(pos, float2(1.0, off));
+	float3 e = Fetch(pos, float2(2.0, off));
+	float dst = Dist(pos).x;
+
+	// Convert distance to weight.
+	float scale = FilterCRTAmount;
+
+	float wa = Gaus(dst - 2.0, scale);
+	float wb = Gaus(dst - 1.0, scale);
+	float wc = Gaus(dst + 0.0, scale);
+	float wd = Gaus(dst + 1.0, scale);
+	float we = Gaus(dst + 2.0, scale);
+
+	return (a * wa + b * wb + c * wc + d * wd + e * we) / (wa + wb + wc + wd + we);
+}
+
+float3 Horz7(float2 pos, float off)
+{
+	float3 a = Fetch(pos, float2(-3.0, off));
+	float3 b = Fetch(pos, float2(-2.0, off));
+	float3 c = Fetch(pos, float2(-1.0, off));
+	float3 d = Fetch(pos, float2( 0.0, off));
+	float3 e = Fetch(pos, float2( 1.0, off));
+	float3 f = Fetch(pos, float2( 2.0, off));
+	float3 g = Fetch(pos, float2( 3.0, off));
+
+	float dst = Dist(pos).x;
+	// Convert distance to weight.
+	float scale = BloomPixel;
+	float wa = Gaus(dst - 3.0, scale);
+	float wb = Gaus(dst - 2.0, scale);
+	float wc = Gaus(dst - 1.0, scale);
+	float wd = Gaus(dst + 0.0, scale);
+	float we = Gaus(dst + 1.0, scale);
+	float wf = Gaus(dst + 2.0, scale);
+	float wg = Gaus(dst + 3.0, scale);
+
+	// Return filtered sample.
+	return (a * wa + b * wb + c * wc + d * wd + e * we + f * wf + g * wg) / (wa + wb + wc + wd + we + wf + wg);
+}
+
+// Return scanline weight.
+float Scan(float2 pos, float off)
+{
+	float dst = Dist(pos).y;
+	return Gaus(dst + off, ScanBrightness);
+}
+
+float BloomScan(float2 pos, float off)
+{
+	float dst = Dist(pos).y;
+
+	return Gaus(dst + off, BloomScanLine);
+}
+
+float3 Tri(float2 pos)
+{
+	float3 a = Horz3(pos, -1.0);
+	float3 b = Horz5(pos, 0.0);
+	float3 c = Horz3(pos, 1.0);
+
+	float wa = Scan(pos, -1.0);
+	float wb = Scan(pos, 0.0);
+	float wc = Scan(pos, 1.0);
+
+	return (a * wa) + (b * wb) + (c * wc);
+}
+
+float3 Bloom(float2 pos)
+{
+	float3 a = Horz5(pos,-2.0);
+	float3 b = Horz7(pos,-1.0);
+	float3 c = Horz7(pos, 0.0);
+	float3 d = Horz7(pos, 1.0);
+	float3 e = Horz5(pos, 2.0);
+
+	float wa = BloomScan(pos,-2.0);
+	float wb = BloomScan(pos,-1.0); 
+	float wc = BloomScan(pos, 0.0);
+	float wd = BloomScan(pos, 1.0);
+	float we = BloomScan(pos, 2.0);
+
+	return a * wa + b * wb + c * wc + d * wd + e * we;
+}
+
+float2 Warp(float2 pos)
+{
+	pos = pos * 2.0 - 1.0;
+	pos *= float2(1.0 + (pos.y * pos.y) * HorizontalWarp, 1.0 + (pos.x * pos.x) * VerticalWarp);
+	return pos * 0.5 + 0.5;
+}
+
+float3 Mask(float2 pos)
+{
+#if MaskingType == 1
+	// Very compressed TV style shadow mask.
+	float lines = MaskAmountLight;
+	float odd = 0.0;
+
+	if (frac(pos.x / 6.0) < 0.5)
+	{
+		odd = 1.0;
+	}
+	if (frac((pos.y + odd) / 2.0) < 0.5)
+	{
+		lines = MaskAmountDark;
+	}
+	pos.x = frac(pos.x / 3.0);
+	float3 mask = float3(MaskAmountDark, MaskAmountDark, MaskAmountDark);
+
+	if (pos.x < 0.333)
+	{
+		mask.r = MaskAmountLight;
+	}
+	else if (pos.x < 0.666)
+	{
+		mask.g = MaskAmountLight;
+	}
+	else
+	{
+		mask.b = MaskAmountLight;
+	}
+
+	mask *= lines;
+
+	return mask;
+
+#elif MaskingType == 2
+	// Aperture-grille.
+	pos.x = frac(pos.x / 3.0);
+	float3 mask = float3(MaskAmountDark, MaskAmountDark, MaskAmountDark);
+
+	if (pos.x < 0.333)
+	{
+		mask.r = MaskAmountLight;
+	}
+	else if (pos.x < 0.666)
+	{
+		mask.g = MaskAmountLight;
+	}
+	else
+	{
+		mask.b = MaskAmountLight;
+	}
+
+	return mask;
+
+#elif MaskingType == 3
+	// Stretched VGA style shadow mask (same as prior shaders).
+	pos.x += pos.y * 3.0;
+	float3 mask = float3(MaskAmountDark, MaskAmountDark, MaskAmountDark);
+	pos.x = frac(pos.x / 6.0);
+
+	if (pos.x < 0.333)
+	{
+		mask.r = MaskAmountLight;
+	}
+	else if (pos.x < 0.666)
+	{
+		mask.g = MaskAmountLight;
+	}
+	else
+	{
+		mask.b = MaskAmountLight;
+	}
+
+	return mask;
+
+#else
+	// VGA style shadow mask.
+	pos.xy = floor(pos.xy * float2(1.0, 0.5));
+	pos.x += pos.y * 3.0;
+
+	float3 mask = float3(MaskAmountDark, MaskAmountDark, MaskAmountDark);
+	pos.x = frac(pos.x / 6.0);
+
+	if (pos.x < 0.333)
+	{
+		mask.r = MaskAmountLight;
+	}
+	else if (pos.x < 0.666)
+	{
+		mask.g = MaskAmountLight;
+	}
+	else
+	{
+		mask.b = MaskAmountLight;
+	}
+	return mask;
+#endif
+}
+
+float4 LottesCRTPass(float4 fragcoord)
+{
+	float4 color;
+	fragcoord -= u_target_rect;
+	float2 inSize = u_target_resolution - (2 * u_target_rect.xy);
+
+	float2 pos = Warp(fragcoord.xy / inSize);
+	color.rgb = Tri(pos);
+	color.rgb += Bloom(pos) * BloomAmount;
+#if UseShadowMask
+	color.rgb *= Mask(fragcoord.xy);
+#endif
+	color.rgb = ToSrgb(color.rgb);
+	color.a = 1.0;
+
+	return color;
+}
+
+PS_OUTPUT ps_filter_lottes(PS_INPUT input)
+{
+	PS_OUTPUT output;
+	output.c = LottesCRTPass(input.p);
+
+	return output;
+}
+
+PS_OUTPUT ps_4x_rgss(PS_INPUT input)
+{
+	PS_OUTPUT output;
+
+	float2 dxy = float2(ddx(input.t.x), ddy(input.t.y));
+	float3 color = 0;
+
+	float s = 1.0/8.0;
+	float l = 3.0/8.0;
+
+	color += sample_c(input.t + float2( s, l) * dxy).rgb;
+	color += sample_c(input.t + float2( l,-s) * dxy).rgb;
+	color += sample_c(input.t + float2(-s,-l) * dxy).rgb;
+	color += sample_c(input.t + float2(-l, s) * dxy).rgb;
+
+	output.c = float4(color * 0.25,1);
+	return output;
+}
+
+PS_OUTPUT ps_automagical_supersampling(PS_INPUT input)
+{
+	PS_OUTPUT output;
+
+	float2 ratio = (u_source_size / u_target_size) * 0.5;
+	float2 steps = floor(ratio);
+	float3 col = sample_c(input.t).rgb;
+	float div = 1;
+
+	for (float y = 0; y < steps.y; y++)
+	{
+		for (float x = 0; x < steps.x; x++)
+		{
+			float2 offset = float2(x,y) - ratio * 0.5;
+			col += sample_c(input.t + offset * u_rcp_source_resolution * 2.0).rgb;
+			div++;
+		}
+	}
+
+	output.c = float4(col / div, 1);
+	return output;
+}

bin/resources/shaders/dx11/shadeboost.fx

@@ -0,0 +1,52 @@
+// SPDX-FileCopyrightText: 2002-2025 PCSX2 Dev Team
+// SPDX-License-Identifier: GPL-3.0+
+
+Texture2D Texture;
+SamplerState Sampler;
+
+cbuffer cb0
+{
+	float4 params;
+};
+
+/*
+** Contrast, saturation, brightness
+** Code of this function is from TGM's shader pack
+** http://irrlicht.sourceforge.net/phpBB2/viewtopic.php?t=21057
+*/
+
+// For all settings: 1.0 = 100% 0.5=50% 1.5 = 150% 
+float4 ContrastSaturationBrightness(float4 color) // Ported to HLSL
+{
+	float brt = params.x;
+	float con = params.y;
+	float sat = params.z;
+	
+	// Increase or decrease these values to adjust r, g and b color channels separately
+	const float AvgLumR = 0.5;
+	const float AvgLumG = 0.5;
+	const float AvgLumB = 0.5;
+	
+	const float3 LumCoeff = float3(0.2125, 0.7154, 0.0721);
+	
+	float3 AvgLumin = float3(AvgLumR, AvgLumG, AvgLumB);
+	float3 brtColor = color.rgb * brt;
+	float3 intensity = dot(brtColor, LumCoeff);
+	float3 satColor = lerp(intensity, brtColor, sat);
+	float3 conColor = lerp(AvgLumin, satColor, con);
+
+	color.rgb = conColor;	
+	return color;
+}
+
+struct PS_INPUT
+{
+	float4 p : SV_Position;
+	float2 t : TEXCOORD0;
+};
+
+float4 ps_main(PS_INPUT input) : SV_Target0
+{
+	float4 c = Texture.Sample(Sampler, input.t);
+	return ContrastSaturationBrightness(c);
+}

bin/resources/shaders/opengl/cas.glsl

@@ -0,0 +1,64 @@
+// Based on CAS_Shader.glsl
+// Copyright(c) 2019 Advanced Micro Devices, Inc.All rights reserved.
+// Permission is hereby granted, free of charge, to any person obtaining a copy
+// of this software and associated documentation files(the "Software"), to deal
+// in the Software without restriction, including without limitation the rights
+// to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+// copies of the Software, and to permit persons to whom the Software is
+// furnished to do so, subject to the following conditions :
+// The above copyright notice and this permission notice shall be included in
+// all copies or substantial portions of the Software.
+// THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+// IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+// FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.IN NO EVENT SHALL THE
+// AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+// LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+// OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
+// THE SOFTWARE.
+
+uniform uvec4 const0;
+uniform uvec4 const1;
+uniform ivec2 srcOffset;
+
+layout(binding=0) uniform sampler2D imgSrc;
+layout(binding=0, rgba8) uniform writeonly image2D imgDst;
+
+#define A_GPU 1
+#define A_GLSL 1
+
+#include "ffx_a.h"
+
+AF3 CasLoad(ASU2 p)
+{
+	return texelFetch(imgSrc, srcOffset + ivec2(p), 0).rgb;
+}
+
+// Lets you transform input from the load into a linear color space between 0 and 1. See ffx_cas.h
+// In this case, our input is already linear and between 0 and 1
+void CasInput(inout AF1 r, inout AF1 g, inout AF1 b) {}
+
+#include "ffx_cas.h"
+
+layout(local_size_x=64) in;
+void main()
+{
+	// Do remapping of local xy in workgroup for a more PS-like swizzle pattern.
+	AU2 gxy = ARmp8x8(gl_LocalInvocationID.x)+AU2(gl_WorkGroupID.x<<4u,gl_WorkGroupID.y<<4u);
+
+	// Filter.
+	AF4 c = vec4(0.0f);
+	CasFilter(c.r, c.g, c.b, gxy, const0, const1, CAS_SHARPEN_ONLY);
+	imageStore(imgDst, ASU2(gxy), c);
+	gxy.x += 8u;
+
+	CasFilter(c.r, c.g, c.b, gxy, const0, const1, CAS_SHARPEN_ONLY);
+	imageStore(imgDst, ASU2(gxy), c);
+	gxy.y += 8u;
+
+	CasFilter(c.r, c.g, c.b, gxy, const0, const1, CAS_SHARPEN_ONLY);
+	imageStore(imgDst, ASU2(gxy), c);
+	gxy.x -= 8u;
+
+	CasFilter(c.r, c.g, c.b, gxy, const0, const1, CAS_SHARPEN_ONLY);
+	imageStore(imgDst, ASU2(gxy), c);
+}

bin/resources/shaders/opengl/convert.glsl

@@ -0,0 +1,604 @@
+// SPDX-FileCopyrightText: 2002-2025 PCSX2 Dev Team
+// SPDX-License-Identifier: GPL-3.0+
+
+//#version 420 // Keep it for editor detection
+
+
+#ifdef VERTEX_SHADER
+
+layout(location = 0) in vec2 POSITION;
+layout(location = 1) in vec2 TEXCOORD0;
+layout(location = 7) in vec4 COLOR;
+
+// FIXME set the interpolation (don't know what dx do)
+// flat means that there is no interpolation. The value given to the fragment shader is based on the provoking vertex conventions.
+//
+// noperspective means that there will be linear interpolation in window-space. This is usually not what you want, but it can have its uses.
+//
+// smooth, the default, means to do perspective-correct interpolation.
+//
+// The centroid qualifier only matters when multisampling. If this qualifier is not present, then the value is interpolated to the pixel's center, anywhere in the pixel, or to one of the pixel's samples. This sample may lie outside of the actual primitive being rendered, since a primitive can cover only part of a pixel's area. The centroid qualifier is used to prevent this; the interpolation point must fall within both the pixel's area and the primitive's area.
+out vec4 PSin_p;
+out vec2 PSin_t;
+out vec4 PSin_c;
+
+void vs_main()
+{
+	PSin_p = vec4(POSITION, 0.5f, 1.0f);
+	PSin_t = TEXCOORD0;
+	PSin_c = COLOR;
+	gl_Position = vec4(POSITION, 0.5f, 1.0f); // NOTE I don't know if it is possible to merge POSITION_OUT and gl_Position
+}
+
+#endif
+
+#ifdef FRAGMENT_SHADER
+
+in vec4 PSin_p;
+in vec2 PSin_t;
+in vec4 PSin_c;
+
+layout(binding = 0) uniform sampler2D TextureSampler;
+
+// Give a different name so I remember there is a special case!
+#if defined(ps_convert_rgba8_16bits) || defined(ps_convert_float32_32bits)
+layout(location = 0) out uint SV_Target1;
+#else
+layout(location = 0) out vec4 SV_Target0;
+#endif
+
+vec4 sample_c()
+{
+	return texture(TextureSampler, PSin_t);
+}
+
+#ifdef ps_copy
+void ps_copy()
+{
+	SV_Target0 = sample_c();
+}
+#endif
+
+#ifdef ps_depth_copy
+void ps_depth_copy()
+{
+  gl_FragDepth = sample_c().r;
+}
+#endif
+
+#ifdef ps_downsample_copy
+uniform ivec2 ClampMin;
+uniform int DownsampleFactor;
+uniform float Weight;
+
+void ps_downsample_copy()
+{
+	ivec2 coord = max(ivec2(gl_FragCoord.xy) * DownsampleFactor, ClampMin);
+	vec4 result = vec4(0);
+	for (int yoff = 0; yoff < DownsampleFactor; yoff++)
+	{
+		for (int xoff = 0; xoff < DownsampleFactor; xoff++)
+			result += texelFetch(TextureSampler, coord + ivec2(xoff, yoff), 0);
+	}
+	SV_Target0 = result / Weight;
+}
+#endif
+
+#ifdef ps_convert_rgba8_16bits
+// Need to be careful with precision here, it can break games like Spider-Man 3 and Dogs Life
+void ps_convert_rgba8_16bits()
+{
+	highp uvec4 i = uvec4(sample_c() * vec4(255.5f, 255.5f, 255.5f, 255.5f));
+
+	SV_Target1 = ((i.x & 0x00F8u) >> 3) | ((i.y & 0x00F8u) << 2) | ((i.z & 0x00f8u) << 7) | ((i.w & 0x80u) << 8);
+}
+#endif
+
+#ifdef ps_convert_float32_32bits
+void ps_convert_float32_32bits()
+{
+	// Convert a GL_FLOAT32 depth texture into a 32 bits UINT texture
+	SV_Target1 = uint(exp2(32.0f) * sample_c().r);
+}
+#endif
+
+#ifdef ps_convert_float32_rgba8
+void ps_convert_float32_rgba8()
+{
+	// Convert a GL_FLOAT32 depth texture into a RGBA color texture
+	uint d = uint(sample_c().r * exp2(32.0f));
+	SV_Target0 = vec4(uvec4((d & 0xFFu), ((d >> 8) & 0xFFu), ((d >> 16) & 0xFFu), (d >> 24))) / vec4(255.0);
+}
+#endif
+
+#ifdef ps_convert_float16_rgb5a1
+void ps_convert_float16_rgb5a1()
+{
+	// Convert a GL_FLOAT32 (only 16 lsb) depth into a RGB5A1 color texture
+	uint d = uint(sample_c().r * exp2(32.0f));
+	SV_Target0 = vec4(uvec4(d << 3, d >> 2, d >> 7, d >> 8) & uvec4(0xf8, 0xf8, 0xf8, 0x80)) / 255.0f;
+}
+#endif
+
+float rgba8_to_depth32(vec4 unorm)
+{
+	uvec4 c = uvec4(unorm * vec4(255.5f));
+	return float(c.r | (c.g << 8) | (c.b << 16) | (c.a << 24)) * exp2(-32.0f);
+}
+
+float rgba8_to_depth24(vec4 unorm)
+{
+	uvec3 c = uvec3(unorm.rgb * vec3(255.5f));
+	return float(c.r | (c.g << 8) | (c.b << 16)) * exp2(-32.0f);
+}
+
+float rgba8_to_depth16(vec4 unorm)
+{
+	uvec2 c = uvec2(unorm.rg * vec2(255.5f));
+	return float(c.r | (c.g << 8)) * exp2(-32.0f);
+}
+
+float rgb5a1_to_depth16(vec4 unorm)
+{
+	uvec4 c = uvec4(unorm * vec4(255.5f));
+	return float(((c.r & 0xF8u) >> 3) | ((c.g & 0xF8u) << 2) | ((c.b & 0xF8u) << 7) | ((c.a & 0x80u) << 8)) * exp2(-32.0f);
+}
+
+#ifdef ps_convert_float32_float24
+void ps_convert_float32_float24()
+{
+	// Truncates depth value to 24bits
+	uint d = uint(sample_c().r * exp2(32.0f)) & 0xFFFFFFu;
+	gl_FragDepth = float(d) * exp2(-32.0f);
+}
+#endif
+
+#ifdef ps_convert_rgba8_float32
+void ps_convert_rgba8_float32()
+{
+	// Convert an RGBA texture into a float depth texture
+	gl_FragDepth = rgba8_to_depth32(sample_c());
+}
+#endif
+
+#ifdef ps_convert_rgba8_float24
+void ps_convert_rgba8_float24()
+{
+	// Same as above but without the alpha channel (24 bits Z)
+
+	// Convert an RGBA texture into a float depth texture
+	gl_FragDepth = rgba8_to_depth24(sample_c());
+}
+#endif
+
+#ifdef ps_convert_rgba8_float16
+void ps_convert_rgba8_float16()
+{
+	// Same as above but without the A/B channels (16 bits Z)
+
+	// Convert an RGBA texture into a float depth texture
+	gl_FragDepth = rgba8_to_depth16(sample_c());
+}
+#endif
+
+#ifdef ps_convert_rgb5a1_float16
+void ps_convert_rgb5a1_float16()
+{
+	// Convert an RGB5A1 (saved as RGBA8) color to a 16 bit Z
+	gl_FragDepth = rgb5a1_to_depth16(sample_c());
+}
+#endif
+
+#define SAMPLE_RGBA_DEPTH_BILN(CONVERT_FN) \
+	ivec2 dims = textureSize(TextureSampler, 0); \
+	vec2 top_left_f = PSin_t * vec2(dims) - 0.5f; \
+	ivec2 top_left = ivec2(floor(top_left_f)); \
+	ivec4 coords = clamp(ivec4(top_left, top_left + 1), ivec4(0), dims.xyxy - 1); \
+	vec2 mix_vals = fract(top_left_f); \
+	float depthTL = CONVERT_FN(texelFetch(TextureSampler, coords.xy, 0)); \
+	float depthTR = CONVERT_FN(texelFetch(TextureSampler, coords.zy, 0)); \
+	float depthBL = CONVERT_FN(texelFetch(TextureSampler, coords.xw, 0)); \
+	float depthBR = CONVERT_FN(texelFetch(TextureSampler, coords.zw, 0)); \
+	gl_FragDepth = mix(mix(depthTL, depthTR, mix_vals.x), mix(depthBL, depthBR, mix_vals.x), mix_vals.y);
+
+#ifdef ps_convert_rgba8_float32_biln
+void ps_convert_rgba8_float32_biln()
+{
+	// Convert an RGBA texture into a float depth texture
+	SAMPLE_RGBA_DEPTH_BILN(rgba8_to_depth32);
+}
+#endif
+
+#ifdef ps_convert_rgba8_float24_biln
+void ps_convert_rgba8_float24_biln()
+{
+	// Same as above but without the alpha channel (24 bits Z)
+
+	// Convert an RGBA texture into a float depth texture
+	SAMPLE_RGBA_DEPTH_BILN(rgba8_to_depth24);
+}
+#endif
+
+#ifdef ps_convert_rgba8_float16_biln
+void ps_convert_rgba8_float16_biln()
+{
+	// Same as above but without the A/B channels (16 bits Z)
+
+	// Convert an RGBA texture into a float depth texture
+	SAMPLE_RGBA_DEPTH_BILN(rgba8_to_depth16);
+}
+#endif
+
+#ifdef ps_convert_rgb5a1_float16_biln
+void ps_convert_rgb5a1_float16_biln()
+{
+	// Convert an RGB5A1 (saved as RGBA8) color to a 16 bit Z
+	SAMPLE_RGBA_DEPTH_BILN(rgb5a1_to_depth16);
+}
+#endif
+
+#ifdef ps_convert_rgb5a1_8i
+uniform uint SBW;
+uniform uint DBW;
+uniform uint PSM;
+uniform float ScaleFactor;
+
+void ps_convert_rgb5a1_8i()
+{
+	// Convert a RGB5A1 texture into a 8 bits packed texture
+	// Input column: 16x2 RGB5A1 pixels
+	// 0: 16 RGBA
+	// 1: 16 RGBA
+	// Output column: 16x4 Index pixels
+	// 0: 16 R5G2
+	// 1: 16 R5G2
+	// 2: 16 G2B5A1
+	// 3: 16 G2B5A1
+	
+	uvec2 pos = uvec2(gl_FragCoord.xy);
+
+	// Collapse separate R G B A areas into their base pixel
+	uvec2 column = (pos & ~uvec2(0u, 3u)) / uvec2(1,2);
+	uvec2 subcolumn = (pos & uvec2(0u, 1u));
+	column.x -= (column.x / 128) * 64;
+	column.y += (column.y / 32) * 32;
+	
+	// Deal with swizzling differences
+	if ((PSM & 0x8) != 0) // PSMCT16S
+	{
+		if ((pos.x & 32) != 0)
+		{
+			column.y += 32; // 4 columns high times 4 to get bottom 4 blocks
+			column.x &= ~32;
+		}
+		
+		if ((pos.x & 64) != 0)
+		{
+			column.x -= 32;
+		}
+		
+		if (((pos.x & 16) != 0) != ((pos.y & 16) != 0))
+		{
+			column.x ^= 16; 
+			column.y ^= 8;
+		}
+		
+		if ((PSM & 0x30) != 0) // PSMZ16S - Untested but hopefully ok if anything uses it.
+		{
+			column.x ^= 32;
+			column.y ^= 16;
+		}
+	}
+	else // PSMCT16
+	{
+		if ((pos.y & 32) != 0)
+		{
+			column.y -= 16;
+			column.x += 32;
+		}
+		
+		if ((pos.x & 96) != 0)
+		{
+			uint multi = (pos.x & 96) / 32;
+			column.y += 16 * multi; // 4 columns high times 4 to get bottom 4 blocks
+			column.x -= (pos.x & 96);
+		}
+		
+		if (((pos.x & 16) != 0) != ((pos.y & 16) != 0))
+		{
+			column.x ^= 16; 
+			column.y ^= 8;
+		}
+		
+		if ((PSM & 0x30) != 0) // PSMZ16 - Untested but hopefully ok if anything uses it.
+		{
+			column.x ^= 32;
+			column.y ^= 32;
+		}
+	}
+	uvec2 coord = column | subcolumn;
+	
+	// Compensate for potentially differing page pitch.
+	uvec2 block_xy = coord / uvec2(64u, 64u);
+	uint block_num = (block_xy.y * (DBW / 128u)) + block_xy.x;
+	uvec2 block_offset = uvec2((block_num % (SBW / 64u)) * 64u, (block_num / (SBW / 64u)) * 64u);
+	coord = (coord % uvec2(64u, 64u)) + block_offset;
+
+	// Apply offset to cols 1 and 2
+	uint is_col23 = pos.y & 4u;
+	uint is_col13 = pos.y & 2u;
+	uint is_col12 = is_col23 ^ (is_col13 << 1);
+	coord.x ^= is_col12; // If cols 1 or 2, flip bit 3 of x
+
+	if (floor(ScaleFactor) != ScaleFactor)
+		coord = uvec2(vec2(coord) * ScaleFactor);
+	else
+		coord *= uvec2(ScaleFactor);
+
+	vec4 pixel = texelFetch(TextureSampler, ivec2(coord), 0);
+	
+	uvec4 denorm_c = uvec4(pixel * 255.5f);
+	if ((pos.y & 2u) == 0u)
+	{
+		uint red = (denorm_c.r >> 3) & 0x1Fu;
+		uint green = (denorm_c.g >> 3) & 0x1Fu;
+		float sel0 = float(((green << 5) | red) & 0xFF) / 255.0f;
+		
+		SV_Target0 = vec4(sel0);
+	}
+	else
+	{
+		uint green = (denorm_c.g >> 3) & 0x1Fu;
+		uint blue = (denorm_c.b >> 3) & 0x1Fu;
+		uint alpha = denorm_c.a & 0x80u;
+		float sel0 = float((alpha | (blue << 2) | (green >> 3)) & 0xFF) / 255.0f;
+
+		SV_Target0 = vec4(sel0);
+	}
+}
+#endif
+
+#ifdef ps_convert_rgba_8i
+uniform uint SBW;
+uniform uint DBW;
+uniform uint PSM;
+uniform float ScaleFactor;
+
+void ps_convert_rgba_8i()
+{
+	// Convert a RGBA texture into a 8 bits packed texture
+	// Input column: 8x2 RGBA pixels
+	// 0: 8 RGBA
+	// 1: 8 RGBA
+	// Output column: 16x4 Index pixels
+	// 0: 8 R | 8 B
+	// 1: 8 R | 8 B
+	// 2: 8 G | 8 A
+	// 3: 8 G | 8 A
+	uvec2 pos = uvec2(gl_FragCoord.xy);
+
+	// Collapse separate R G B A areas into their base pixel
+	uvec2 block = (pos & ~uvec2(15u, 3u)) >> 1;
+	uvec2 subblock = pos & uvec2(7u, 1u);
+	uvec2 coord = block | subblock;
+
+	// Compensate for potentially differing page pitch.
+	uvec2 block_xy = coord / uvec2(64u, 32u);
+	uint block_num = (block_xy.y * (DBW / 128u)) + block_xy.x;
+	uvec2 block_offset = uvec2((block_num % (SBW / 64u)) * 64u, (block_num / (SBW / 64u)) * 32u);
+	coord = (coord % uvec2(64u, 32u)) + block_offset;
+
+	// Apply offset to cols 1 and 2
+	uint is_col23 = pos.y & 4u;
+	uint is_col13 = pos.y & 2u;
+	uint is_col12 = is_col23 ^ (is_col13 << 1);
+	coord.x ^= is_col12; // If cols 1 or 2, flip bit 3 of x
+
+	if (floor(ScaleFactor) != ScaleFactor)
+		coord = uvec2(vec2(coord) * ScaleFactor);
+	else
+		coord *= uvec2(ScaleFactor);
+
+	vec4 pixel = texelFetch(TextureSampler, ivec2(coord), 0);
+	vec2  sel0 = (pos.y & 2u) == 0u ? pixel.rb : pixel.ga;
+	float sel1 = (pos.x & 8u) == 0u ? sel0.x : sel0.y;
+	SV_Target0 = vec4(sel1);
+}
+#endif
+
+#ifdef ps_filter_transparency
+void ps_filter_transparency()
+{
+	vec4 c = sample_c();
+	SV_Target0 = vec4(c.rgb, 1.0);
+}
+#endif
+
+// Used for DATE (stencil)
+// DATM == 1
+#ifdef ps_datm1
+void ps_datm1()
+{
+	if(sample_c().a < (127.5f / 255.0f)) // >= 0x80 pass
+		discard;
+}
+#endif
+
+// Used for DATE (stencil)
+// DATM == 0
+#ifdef ps_datm0
+void ps_datm0()
+{
+	if((127.5f / 255.0f) < sample_c().a) // < 0x80 pass (== 0x80 should not pass)
+		discard;
+}
+#endif
+
+// Used for DATE (stencil)
+// DATM == 1
+#ifdef ps_datm1_rta_correction
+void ps_datm1_rta_correction()
+{
+	if(sample_c().a < (254.5f / 255.0f)) // >= 0x80 pass
+		discard;
+}
+#endif
+
+// Used for DATE (stencil)
+// DATM == 0
+#ifdef ps_datm0_rta_correction
+void ps_datm0_rta_correction()
+{
+	if((254.5f / 255.0f) < sample_c().a) // < 0x80 pass (== 0x80 should not pass)
+		discard;
+}
+#endif
+
+#ifdef ps_rta_correction
+void ps_rta_correction()
+{
+	vec4 value = sample_c();
+	SV_Target0 = vec4(value.rgb, value.a / (128.25f / 255.0f));
+}
+#endif
+
+#ifdef ps_rta_decorrection
+void ps_rta_decorrection()
+{
+	vec4 value = sample_c();
+	SV_Target0 = vec4(value.rgb, value.a * (128.25f / 255.0f));
+}
+#endif
+
+#ifdef ps_colclip_init
+void ps_colclip_init()
+{
+	vec4 value = sample_c();
+	SV_Target0 = vec4(round(value.rgb * 255.0f) / 65535.0f, value.a);
+}
+#endif
+
+#ifdef ps_colclip_resolve
+void ps_colclip_resolve()
+{
+	vec4 value = sample_c();
+	SV_Target0 = vec4(vec3(uvec3(value.rgb * 65535.0f) & 255u) / 255.0f, value.a);
+}
+#endif
+
+#ifdef ps_convert_clut_4
+uniform uvec3 offset;
+uniform float scale;
+
+void ps_convert_clut_4()
+{
+	// CLUT4 is easy, just two rows of 8x8.
+	uint index = uint(gl_FragCoord.x) + offset.z;
+	uvec2 pos = uvec2(index % 8u, index / 8u);
+
+	ivec2 final = ivec2(floor(vec2(offset.xy + pos) * vec2(scale)));
+	SV_Target0 = texelFetch(TextureSampler, final, 0);
+}
+#endif
+
+#ifdef ps_convert_clut_8
+uniform uvec3 offset;
+uniform float scale;
+
+void ps_convert_clut_8()
+{
+	uint index = min(uint(gl_FragCoord.x) + offset.z, 255u);
+
+	// CLUT is arranged into 8 groups of 16x2, with the top-right and bottom-left quadrants swapped.
+	// This can probably be done better..
+	uint subgroup = (index / 8u) % 4u;
+	uvec2 pos;
+	pos.x = (index % 8u) + ((subgroup >= 2u) ? 8u : 0u);
+	pos.y = ((index / 32u) * 2u) + (subgroup % 2u);
+
+	ivec2 final = ivec2(floor(vec2(offset.xy + pos) * vec2(scale)));
+	SV_Target0 = texelFetch(TextureSampler, final, 0);
+}
+#endif
+
+#ifdef ps_yuv
+uniform ivec2 EMOD;
+
+void ps_yuv()
+{
+	vec4 i = sample_c();
+	vec4 o = vec4(0.0f);
+
+	mat3 rgb2yuv; // Value from GS manual
+	rgb2yuv[0] = vec3(0.587, -0.311, -0.419);
+	rgb2yuv[1] = vec3(0.114, 0.500, -0.081);
+	rgb2yuv[2] = vec3(0.299, -0.169, 0.500);
+
+	vec3 yuv = rgb2yuv * i.gbr;
+
+	float Y = float(0xDB)/255.0f * yuv.x + float(0x10)/255.0f;
+	float Cr = float(0xE0)/255.0f * yuv.y + float(0x80)/255.0f;
+	float Cb = float(0xE0)/255.0f * yuv.z + float(0x80)/255.0f;
+
+	switch(EMOD.x)
+	{
+		case 0:
+			o.a = i.a;
+			break;
+		case 1:
+			o.a = Y;
+			break;
+		case 2:
+			o.a = Y/2.0f;
+			break;
+		case 3:
+			o.a = 0.0f;
+			break;
+	}
+
+	switch(EMOD.y)
+	{
+		case 0:
+			o.rgb = i.rgb;
+			break;
+		case 1:
+			o.rgb = vec3(Y);
+			break;
+		case 2:
+			o.rgb = vec3(Y, Cb, Cr);
+			break;
+		case 3:
+			o.rgb = vec3(i.a);
+			break;
+	}
+
+	SV_Target0 = o;
+}
+#endif
+
+#if defined(ps_stencil_image_init_0) || defined(ps_stencil_image_init_1) || defined(ps_stencil_image_init_2) || defined(ps_stencil_image_init_3)
+
+void main()
+{
+	SV_Target0 = vec4(0x7FFFFFFF);
+
+	#ifdef ps_stencil_image_init_0
+		if((127.5f / 255.0f) < sample_c().a) // < 0x80 pass (== 0x80 should not pass)
+			SV_Target0 = vec4(-1);
+	#endif
+	#ifdef ps_stencil_image_init_1
+		if(sample_c().a < (127.5f / 255.0f)) // >= 0x80 pass
+			SV_Target0 = vec4(-1);
+	#endif
+	#ifdef ps_stencil_image_init_2
+		if((254.5f / 255.0f) < sample_c().a) // < 0x80 pass (== 0x80 should not pass)
+			SV_Target0 = vec4(-1);
+	#endif
+	#ifdef ps_stencil_image_init_3
+		if(sample_c().a < (254.5f / 255.0f)) // >= 0x80 pass
+			SV_Target0 = vec4(-1);
+	#endif
+}
+#endif
+
+#endif

bin/resources/shaders/opengl/imgui.glsl

@@ -0,0 +1,38 @@
+// SPDX-FileCopyrightText: 2002-2025 PCSX2 Dev Team
+// SPDX-License-Identifier: GPL-3.0+
+
+#ifdef VERTEX_SHADER
+
+layout(location = 0) in vec2 Position;
+layout(location = 1) in vec2 UV;
+layout(location = 2) in vec4 Color;
+
+uniform mat4 ProjMtx;
+
+out vec2 Frag_UV;
+out vec4 Frag_Color;
+
+void vs_main()
+{
+	Frag_UV = UV;
+	Frag_Color = Color;
+	gl_Position = ProjMtx * vec4(Position.xy, 0.0, 1.0);
+}
+
+#endif
+
+#ifdef FRAGMENT_SHADER
+
+layout(binding = 0) uniform sampler2D Texture;
+
+in vec2 Frag_UV;
+in vec4 Frag_Color;
+
+layout(location = 0) out vec4 Out_Color;
+
+void ps_main()
+{
+	Out_Color = Frag_Color * texture(Texture, Frag_UV.st);
+}
+
+#endif

bin/resources/shaders/opengl/interlace.glsl

@@ -0,0 +1,181 @@
+// SPDX-FileCopyrightText: 2002-2025 PCSX2 Dev Team
+// SPDX-License-Identifier: GPL-3.0+
+
+//#version 420 // Keep it for editor detection
+
+#ifdef FRAGMENT_SHADER
+
+in vec4 PSin_p;
+in vec2 PSin_t;
+in vec4 PSin_c;
+
+uniform vec4 ZrH;
+
+layout(binding = 0) uniform sampler2D TextureSampler;
+
+layout(location = 0) out vec4 SV_Target0;
+
+
+// Weave shader
+void ps_main0()
+{
+	int idx   = int(ZrH.x);          // buffer index passed from CPU
+	int field = idx & 1;             // current field
+	int vpos  = int(gl_FragCoord.y); // vertical position of destination texture
+
+	if ((vpos & 1) == field)
+		SV_Target0 = textureLod(TextureSampler, PSin_t, 0);
+	else
+		discard;
+}
+
+
+// Bob shader
+void ps_main1()
+{
+	SV_Target0 = textureLod(TextureSampler, PSin_t, 0);
+}
+
+
+// Blend shader
+void ps_main2()
+{
+	vec2 vstep = vec2(0.0f, ZrH.y);
+	vec4 c0 = textureLod(TextureSampler, PSin_t - vstep, 0);
+	vec4 c1 = textureLod(TextureSampler, PSin_t, 0);
+	vec4 c2 = textureLod(TextureSampler, PSin_t + vstep, 0);
+
+	SV_Target0 = (c0 + c1 * 2.0f + c2) / 4.0f;
+}
+
+
+// MAD shader - buffering
+void ps_main3()
+{
+	// We take half the lines from the current frame and stores them in the MAD frame buffer.
+	// the MAD frame buffer is split in 2 consecutive banks of 2 fields each, the fields in each bank
+	// are interleaved (top field at even lines and bottom field at odd lines).
+	// When the source texture has an odd vres, the first line of bank 1 would be an odd index
+	// causing the wrong lines to be discarded, so a vertical offset (lofs) is added to the vertical
+	// position of the destination texture to force the proper field alignment
+
+	int  idx    = int(ZrH.x);                                // buffer index passed from CPU
+	int  bank   = idx >> 1;                                  // current bank
+	int  field  = idx & 1;                                   // current field
+	int  vres   = int(ZrH.z) >> 1;                           // vertical resolution of source texture
+	int  lofs   = ((((vres + 1) >> 1) << 1) - vres) & bank;  // line alignment offset for bank 1
+	int  vpos   = int(gl_FragCoord.y) + lofs;                // vertical position of destination texture
+
+	// if the index of current destination line belongs to the current fiels we update it, otherwise
+	// we leave the old line in the destination buffer
+	if ((vpos & 1) == field)
+		SV_Target0 = textureLod(TextureSampler, PSin_t, 0);
+	else
+		discard;
+}
+
+
+// MAD shader - reconstruction
+void ps_main4()
+{
+	// we use the contents of the MAD frame buffer to reconstruct the missing lines from the current field.
+
+	int   idx          = int(ZrH.x);                         // buffer index passed from CPU
+	int   field        = idx & 1;                            // current field
+	int   vpos         = int(gl_FragCoord.y);                // vertical position of destination texture
+	float sensitivity  = ZrH.w;                              // passed from CPU, higher values mean more likely to use weave
+	vec3  motion_thr   = vec3(1.0, 1.0, 1.0) * sensitivity;  //
+	vec2  bofs         = vec2(0.0f, 0.5f);                   // position of the bank 1 relative to source texture size
+	vec2  vscale       = vec2(1.0f, 0.5f);                   // scaling factor from source to destination texture
+	vec2  lofs         = vec2(0.0f, ZrH.y) * vscale;         // distance between two adjacent lines relative to source texture size
+	vec2  iptr         = PSin_t * vscale;                    // pointer to the current pixel in the source texture
+
+	vec2 p_t0; // pointer to current pixel (missing or not) from most recent frame
+	vec2 p_t1; // pointer to current pixel (missing or not) from one frame back
+	vec2 p_t2; // pointer to current pixel (missing or not) from two frames back
+	vec2 p_t3; // pointer to current pixel (missing or not) from three frames back
+
+	switch (idx)
+	{
+		case 1:
+			p_t0 = iptr;
+			p_t1 = iptr;
+			p_t2 = iptr + bofs;
+			p_t3 = iptr + bofs;
+			break;
+		case 2:
+			p_t0 = iptr + bofs;
+			p_t1 = iptr;
+			p_t2 = iptr;
+			p_t3 = iptr + bofs;
+			break;
+		case 3:
+			p_t0 = iptr + bofs;
+			p_t1 = iptr + bofs;
+			p_t2 = iptr;
+			p_t3 = iptr;
+			break;
+		default:
+			p_t0 = iptr;
+			p_t1 = iptr + bofs;
+			p_t2 = iptr + bofs;
+			p_t3 = iptr;
+			break;
+	}
+
+
+	// calculating motion, only relevant for missing lines where the "center line" is pointed
+	// by p_t1
+
+	vec4 hn = textureLod(TextureSampler, p_t0 - lofs, 0); // new high pixel
+	vec4 cn = textureLod(TextureSampler, p_t1, 0);        // new center pixel
+	vec4 ln = textureLod(TextureSampler, p_t0 + lofs, 0); // new low pixel
+
+	vec4 ho = textureLod(TextureSampler, p_t2 - lofs, 0); // old high pixel
+	vec4 co = textureLod(TextureSampler, p_t3, 0);        // old center pixel
+	vec4 lo = textureLod(TextureSampler, p_t2 + lofs, 0); // old low pixel
+
+	vec3 mh = hn.rgb - ho.rgb; // high pixel motion
+	vec3 mc = cn.rgb - co.rgb; // center pixel motion
+	vec3 ml = ln.rgb - lo.rgb; // low pixel motion
+
+	mh = max(mh, -mh) - motion_thr;
+	mc = max(mc, -mc) - motion_thr;
+	ml = max(ml, -ml) - motion_thr;
+
+	#if 1 // use this code to evaluate each color motion separately
+		float mh_max = max(max(mh.x, mh.y), mh.z);
+		float mc_max = max(max(mc.x, mc.y), mc.z);
+		float ml_max = max(max(ml.x, ml.y), ml.z);
+	#else // use this code to evaluate average color motion
+		float mh_max = mh.x + mh.y + mh.z;
+		float mc_max = mc.x + mc.y + mc.z;
+		float ml_max = ml.x + ml.y + ml.z;
+	#endif
+
+
+	// selecting deinterlacing output
+
+	if ((vpos & 1) == field)
+	{
+		// output coordinate present on current field
+		SV_Target0 = textureLod(TextureSampler, p_t0, 0);
+	}
+	else if ((iptr.y > 0.5f - lofs.y) || (iptr.y < 0.0 + lofs.y))
+	{
+		// top and bottom lines are always weaved
+		SV_Target0 = cn;
+	}
+	else
+	{
+		// missing line needs to be reconstructed
+		if(((mh_max > 0.0f) || (ml_max > 0.0f)) || (mc_max > 0.0f))
+			// high motion -> interpolate pixels above and below
+			SV_Target0 = (hn + ln) / 2.0f;
+		else
+			// low motion -> weave
+			SV_Target0 = cn;
+	}
+}
+
+#endif

bin/resources/shaders/opengl/merge.glsl

@@ -0,0 +1,33 @@
+// SPDX-FileCopyrightText: 2002-2025 PCSX2 Dev Team
+// SPDX-License-Identifier: GPL-3.0+
+
+//#version 420 // Keep it for editor detection
+
+#ifdef FRAGMENT_SHADER
+
+in vec4 PSin_p;
+in vec2 PSin_t;
+in vec4 PSin_c;
+
+uniform vec4 BGColor;
+
+layout(binding = 0) uniform sampler2D TextureSampler;
+
+layout(location = 0) out vec4 SV_Target0;
+
+void ps_main0()
+{
+	vec4 c = texture(TextureSampler, PSin_t);
+	// Note: clamping will be done by fixed unit
+	c.a *= 2.0f;
+	SV_Target0 = c;
+}
+
+void ps_main1()
+{
+	vec4 c = texture(TextureSampler, PSin_t);
+	c.a = BGColor.a;
+	SV_Target0 = c;
+}
+
+#endif

bin/resources/shaders/opengl/present.glsl

@@ -0,0 +1,476 @@
+// SPDX-FileCopyrightText: 2002-2025 PCSX2 Dev Team
+// SPDX-License-Identifier: GPL-3.0+
+
+//#version 420 // Keep it for editor detection
+
+
+#ifdef VERTEX_SHADER
+
+layout(location = 0) in vec2 POSITION;
+layout(location = 1) in vec2 TEXCOORD0;
+layout(location = 7) in vec4 COLOR;
+
+// FIXME set the interpolation (don't know what dx do)
+// flat means that there is no interpolation. The value given to the fragment shader is based on the provoking vertex conventions.
+//
+// noperspective means that there will be linear interpolation in window-space. This is usually not what you want, but it can have its uses.
+//
+// smooth, the default, means to do perspective-correct interpolation.
+//
+// The centroid qualifier only matters when multisampling. If this qualifier is not present, then the value is interpolated to the pixel's center, anywhere in the pixel, or to one of the pixel's samples. This sample may lie outside of the actual primitive being rendered, since a primitive can cover only part of a pixel's area. The centroid qualifier is used to prevent this; the interpolation point must fall within both the pixel's area and the primitive's area.
+out vec4 PSin_p;
+out vec2 PSin_t;
+out vec4 PSin_c;
+
+void vs_main()
+{
+	PSin_p = vec4(POSITION, 0.5f, 1.0f);
+	PSin_t = TEXCOORD0;
+	PSin_c = COLOR;
+	gl_Position = vec4(POSITION, 0.5f, 1.0f); // NOTE I don't know if it is possible to merge POSITION_OUT and gl_Position
+}
+
+#endif
+
+#ifdef FRAGMENT_SHADER
+
+uniform vec4 u_source_rect;
+uniform vec4 u_target_rect;
+uniform vec2 u_source_size;
+uniform vec2 u_target_size;
+uniform vec2 u_target_resolution;
+uniform vec2 u_rcp_target_resolution; // 1 / u_target_resolution
+uniform vec2 u_source_resolution;
+uniform vec2 u_rcp_source_resolution; // 1 / u_source_resolution
+uniform float u_time;
+
+in vec4 PSin_p;
+in vec2 PSin_t;
+in vec4 PSin_c;
+
+layout(binding = 0) uniform sampler2D TextureSampler;
+
+layout(location = 0) out vec4 SV_Target0;
+
+vec4 sample_c()
+{
+	return texture(TextureSampler, PSin_t);
+}
+
+vec4 sample_c(vec2 uv)
+{
+	return texture(TextureSampler, uv);
+}
+
+vec4 ps_crt(uint i)
+{
+	vec4 mask[4] = vec4[4](
+		vec4(1, 0, 0, 0),
+		vec4(0, 1, 0, 0),
+		vec4(0, 0, 1, 0),
+		vec4(1, 1, 1, 0));
+	return sample_c() * clamp((mask[i] + 0.5f), 0.0f, 1.0f);
+}
+
+#ifdef ps_copy
+void ps_copy()
+{
+	SV_Target0 = sample_c();
+}
+#endif
+
+#ifdef ps_filter_scanlines
+vec4 ps_scanlines(uint i)
+{
+	vec4 mask[2] = vec4[2]
+	(
+		vec4(1, 1, 1, 0),
+		vec4(0, 0, 0, 0)
+	);
+
+	return sample_c() * clamp((mask[i] + 0.5f), 0.0f, 1.0f);
+}
+
+void ps_filter_scanlines() // scanlines
+{
+	highp uvec4 p = uvec4(gl_FragCoord);
+
+	vec4 c = ps_scanlines(p.y % 2u);
+
+	SV_Target0 = c;
+}
+#endif
+
+#ifdef ps_filter_diagonal
+void ps_filter_diagonal() // diagonal
+{
+	highp uvec4 p = uvec4(gl_FragCoord);
+
+	vec4 c = ps_crt((p.x + (p.y % 3u)) % 3u);
+
+	SV_Target0 = c;
+}
+#endif
+
+#ifdef ps_filter_triangular
+void ps_filter_triangular() // triangular
+{
+	highp uvec4 p = uvec4(gl_FragCoord);
+
+	vec4 c = ps_crt(((p.x + ((p.y >> 1u) & 1u) * 3u) >> 1u) % 3u);
+
+	SV_Target0 = c;
+}
+#endif
+
+#ifdef ps_filter_complex
+void ps_filter_complex()
+{
+	const float PI = 3.14159265359f;
+	vec2 texdim = vec2(textureSize(TextureSampler, 0));
+	float factor = (0.9f - 0.4f * cos(2.0f * PI * PSin_t.y * texdim.y));
+	vec4 c = factor * texture(TextureSampler, vec2(PSin_t.x, (floor(PSin_t.y * texdim.y) + 0.5f) / texdim.y));
+
+	SV_Target0 = c;
+}
+#endif
+
+#ifdef ps_filter_lottes
+
+#define MaskingType 4                      //[1|2|3|4] The type of CRT shadow masking used. 1: compressed TV style, 2: Aperture-grille, 3: Stretched VGA style, 4: VGA style.
+#define ScanBrightness -8.00               //[-16.0 to 1.0] The overall brightness of the scanline effect. Lower for darker, higher for brighter.
+#define FilterCRTAmount -3.00              //[-4.0 to 1.0] The amount of filtering used, to replicate the TV CRT look. Lower for less, higher for more.
+#define HorizontalWarp 0.00                //[0.0 to 0.1] The distortion warping effect for the horizontal (x) axis of the screen. Use small increments.
+#define VerticalWarp 0.00                  //[0.0 to 0.1] The distortion warping effect for the verticle (y) axis of the screen. Use small increments.
+#define MaskAmountDark 0.50                //[0.0 to 1.0] The value of the dark masking line effect used. Lower for darker lower end masking, higher for brighter.
+#define MaskAmountLight 1.50               //[0.0 to 2.0] The value of the light masking line effect used. Lower for darker higher end masking, higher for brighter.
+#define BloomPixel -1.50                   //[-2.0 -0.5] Pixel bloom radius. Higher for increased softness of bloom.
+#define BloomScanLine -2.0                 //[-4.0 -1.0] Scanline bloom radius. Higher for increased softness of bloom.
+#define BloomAmount 0.15                   //[0.0 1.0] Bloom intensity. Higher for brighter.
+#define Shape 2.0                          //[0.0 10.0] Kernal filter shape. Lower values will darken image and introduce moire patterns if used with curvature.
+#define UseShadowMask 1                    //[0 or 1] Enables, or disables the use of the CRT shadow mask. 0 is disabled, 1 is enabled.
+
+float ToLinear1(float c)
+{
+	return c <= 0.04045 ? c / 12.92 : pow((c + 0.055) / 1.055, 2.4);
+}
+
+vec3 ToLinear(vec3 c)
+{
+	return vec3(ToLinear1(c.r), ToLinear1(c.g), ToLinear1(c.b));
+}
+
+float ToSrgb1(float c)
+{
+	return c < 0.0031308 ? c * 12.92 : 1.055 * pow(c, 0.41666) - 0.055;
+}
+
+vec3 ToSrgb(vec3 c)
+{
+	return vec3(ToSrgb1(c.r), ToSrgb1(c.g), ToSrgb1(c.b));
+}
+
+vec3 Fetch(vec2 pos, vec2 off)
+{
+	pos = (floor(pos * u_target_size + off) + vec2(0.5, 0.5)) / u_target_size;
+	if (max(abs(pos.x - 0.5), abs(pos.y - 0.5)) > 0.5)
+	{
+		return vec3(0.0, 0.0, 0.0);
+	}
+	else
+	{
+		return ToLinear(texture(TextureSampler, pos.xy).rgb);
+	}
+}
+
+vec2 Dist(vec2 pos)
+{
+	pos = pos * vec2(640, 480);
+
+	return -((pos - floor(pos)) - vec2(0.5, 0.5));
+}
+
+float Gaus(float pos, float scale)
+{
+	return exp2(scale * pow(abs(pos), Shape));
+}
+
+vec3 Horz3(vec2 pos, float off)
+{
+	vec3 b = Fetch(pos, vec2(-1.0, off));
+	vec3 c = Fetch(pos, vec2(0.0, off));
+	vec3 d = Fetch(pos, vec2(1.0, off));
+	float dst = Dist(pos).x;
+
+	// Convert distance to weight.
+	float scale = FilterCRTAmount;
+	float wb = Gaus(dst - 1.0, scale);
+	float wc = Gaus(dst + 0.0, scale);
+	float wd = Gaus(dst + 1.0, scale);
+
+	return (b * wb + c * wc + d * wd) / (wb + wc + wd);
+}
+
+vec3 Horz5(vec2 pos, float off)
+{
+	vec3 a = Fetch(pos, vec2(-2.0, off));
+	vec3 b = Fetch(pos, vec2(-1.0, off));
+	vec3 c = Fetch(pos, vec2(0.0, off));
+	vec3 d = Fetch(pos, vec2(1.0, off));
+	vec3 e = Fetch(pos, vec2(2.0, off));
+	float dst = Dist(pos).x;
+
+	// Convert distance to weight.
+	float scale = FilterCRTAmount;
+
+	float wa = Gaus(dst - 2.0, scale);
+	float wb = Gaus(dst - 1.0, scale);
+	float wc = Gaus(dst + 0.0, scale);
+	float wd = Gaus(dst + 1.0, scale);
+	float we = Gaus(dst + 2.0, scale);
+
+	return (a * wa + b * wb + c * wc + d * wd + e * we) / (wa + wb + wc + wd + we);
+}
+
+vec3 Horz7(vec2 pos, float off)
+{
+	vec3 a = Fetch(pos, vec2(-3.0, off));
+	vec3 b = Fetch(pos, vec2(-2.0, off));
+	vec3 c = Fetch(pos, vec2(-1.0, off));
+	vec3 d = Fetch(pos, vec2( 0.0, off));
+	vec3 e = Fetch(pos, vec2( 1.0, off));
+	vec3 f = Fetch(pos, vec2( 2.0, off));
+	vec3 g = Fetch(pos, vec2( 3.0, off));
+
+	float dst = Dist(pos).x;
+	// Convert distance to weight.
+	float scale = BloomPixel;
+	float wa = Gaus(dst - 3.0, scale);
+	float wb = Gaus(dst - 2.0, scale);
+	float wc = Gaus(dst - 1.0, scale);
+	float wd = Gaus(dst + 0.0, scale);
+	float we = Gaus(dst + 1.0, scale);
+	float wf = Gaus(dst + 2.0, scale);
+	float wg = Gaus(dst + 3.0, scale);
+
+	// Return filtered sample.
+	return (a * wa + b * wb + c * wc + d * wd + e * we + f * wf + g * wg) / (wa + wb + wc + wd + we + wf + wg);
+}
+
+// Return scanline weight.
+float Scan(vec2 pos, float off)
+{
+	float dst = Dist(pos).y;
+	return Gaus(dst + off, ScanBrightness);
+}
+
+float BloomScan(vec2 pos, float off)
+{
+	float dst = Dist(pos).y;
+
+	return Gaus(dst + off, BloomScanLine);
+}
+
+vec3 Tri(vec2 pos)
+{
+	vec3 a = Horz3(pos, -1.0);
+	vec3 b = Horz5(pos, 0.0);
+	vec3 c = Horz3(pos, 1.0);
+
+	float wa = Scan(pos, -1.0);
+	float wb = Scan(pos, 0.0);
+	float wc = Scan(pos, 1.0);
+
+	return (a * wa) + (b * wb) + (c * wc);
+}
+
+vec3 Bloom(vec2 pos)
+{
+	vec3 a = Horz5(pos,-2.0);
+	vec3 b = Horz7(pos,-1.0);
+	vec3 c = Horz7(pos, 0.0);
+	vec3 d = Horz7(pos, 1.0);
+	vec3 e = Horz5(pos, 2.0);
+
+	float wa = BloomScan(pos,-2.0);
+	float wb = BloomScan(pos,-1.0); 
+	float wc = BloomScan(pos, 0.0);
+	float wd = BloomScan(pos, 1.0);
+	float we = BloomScan(pos, 2.0);
+
+	return a * wa + b * wb + c * wc + d * wd + e * we;
+}
+
+vec2 Warp(vec2 pos)
+{
+	pos = pos * 2.0 - 1.0;
+	pos *= vec2(1.0 + (pos.y * pos.y) * HorizontalWarp, 1.0 + (pos.x * pos.x) * VerticalWarp);
+	return pos * 0.5 + 0.5;
+}
+
+vec3 Mask(vec2 pos)
+{
+#if MaskingType == 1
+	// Very compressed TV style shadow mask.
+	float lines = MaskAmountLight;
+	float odd = 0.0;
+
+	if (fract(pos.x / 6.0) < 0.5)
+	{
+		odd = 1.0;
+	}
+	if (fract((pos.y + odd) / 2.0) < 0.5)
+	{
+		lines = MaskAmountDark;
+	}
+	pos.x = fract(pos.x / 3.0);
+	vec3 mask = vec3(MaskAmountDark, MaskAmountDark, MaskAmountDark);
+
+	if (pos.x < 0.333)
+	{
+		mask.r = MaskAmountLight;
+	}
+	else if (pos.x < 0.666)
+	{
+		mask.g = MaskAmountLight;
+	}
+	else
+	{
+		mask.b = MaskAmountLight;
+	}
+
+	mask *= lines;
+
+	return mask;
+
+#elif MaskingType == 2
+	// Aperture-grille.
+	pos.x = fract(pos.x / 3.0);
+	vec3 mask = vec3(MaskAmountDark, MaskAmountDark, MaskAmountDark);
+
+	if (pos.x < 0.333)
+	{
+		mask.r = MaskAmountLight;
+	}
+	else if (pos.x < 0.666)
+	{
+		mask.g = MaskAmountLight;
+	}
+	else
+	{
+		mask.b = MaskAmountLight;
+	}
+
+	return mask;
+
+#elif MaskingType == 3
+	// Stretched VGA style shadow mask (same as prior shaders).
+	pos.x += pos.y * 3.0;
+	vec3 mask = vec3(MaskAmountDark, MaskAmountDark, MaskAmountDark);
+	pos.x = fract(pos.x / 6.0);
+
+	if (pos.x < 0.333)
+	{
+		mask.r = MaskAmountLight;
+	}
+	else if (pos.x < 0.666)
+	{
+		mask.g = MaskAmountLight;
+	}
+	else
+	{
+		mask.b = MaskAmountLight;
+	}
+
+	return mask;
+
+#else
+	// VGA style shadow mask.
+	pos.xy = floor(pos.xy * vec2(1.0, 0.5));
+	pos.x += pos.y * 3.0;
+
+	vec3 mask = vec3(MaskAmountDark, MaskAmountDark, MaskAmountDark);
+	pos.x = fract(pos.x / 6.0);
+
+	if (pos.x < 0.333)
+	{
+		mask.r = MaskAmountLight;
+	}
+	else if (pos.x < 0.666)
+	{
+		mask.g = MaskAmountLight;
+	}
+	else
+	{
+		mask.b = MaskAmountLight;
+	}
+	return mask;
+#endif
+}
+
+vec4 LottesCRTPass()
+{
+	//flipped y axis in opengl
+	vec2 fragcoord = vec2(gl_FragCoord.x, u_target_resolution.y - gl_FragCoord.y) - u_target_rect.xy;
+	vec4 color;
+	vec2 inSize = u_target_resolution - (2.0 * u_target_rect.xy);
+
+	vec2 pos = Warp(fragcoord.xy / inSize);
+	color.rgb = Tri(pos);
+	color.rgb += Bloom(pos) * BloomAmount;
+#if UseShadowMask
+	color.rgb *= Mask(fragcoord.xy);
+#endif
+	color.rgb = ToSrgb(color.rgb);
+
+	return color;
+}
+
+void ps_filter_lottes()
+{
+	SV_Target0 = LottesCRTPass();
+}
+
+#endif
+
+#ifdef ps_4x_rgss
+void ps_4x_rgss()
+{
+	vec2 dxy = vec2(dFdx(PSin_t.x), dFdy(PSin_t.y));
+	vec3 color = vec3(0);
+
+	float s = 1.0/8.0;
+	float l = 3.0/8.0;
+
+	color += sample_c(PSin_t + vec2( s, l) * dxy).rgb;
+	color += sample_c(PSin_t + vec2( l,-s) * dxy).rgb;
+	color += sample_c(PSin_t + vec2(-s,-l) * dxy).rgb;
+	color += sample_c(PSin_t + vec2(-l, s) * dxy).rgb;
+
+	SV_Target0 = vec4(color * 0.25,1);
+}
+#endif
+
+#ifdef ps_automagical_supersampling
+void ps_automagical_supersampling()
+{
+	vec2 ratio = (u_source_size / u_target_size) * 0.5;
+	vec2 steps = floor(ratio);
+	vec3 col = sample_c(PSin_t).rgb;
+	float div = 1;
+
+	for (float y = 0; y < steps.y; y++)
+	{
+		for (float x = 0; x < steps.x; x++)
+		{
+			vec2 offset = vec2(x,y) - ratio * 0.5;
+			col += sample_c(PSin_t + offset * u_rcp_source_resolution * 2.0).rgb;
+			div++;
+		}
+	}
+
+	SV_Target0 = vec4(col / div, 1);
+}
+#endif
+
+#endif

bin/resources/shaders/opengl/shadeboost.glsl

@@ -0,0 +1,60 @@
+// SPDX-FileCopyrightText: 2002-2025 PCSX2 Dev Team
+// SPDX-License-Identifier: GPL-3.0+
+
+//#version 420 // Keep it for editor detection
+
+/*
+** Contrast, saturation, brightness
+** Code of this function is from TGM's shader pack
+** http://irrlicht.sourceforge.net/phpBB2/viewtopic.php?t=21057
+** TGM's author comment about the license (included in the previous link)
+** "do with it, what you want! its total free!
+** (but would be nice, if you say that you used my shaders  :wink: ) but not necessary"
+*/
+
+#ifdef FRAGMENT_SHADER
+
+uniform vec4 params;
+
+in vec4 PSin_p;
+in vec2 PSin_t;
+in vec4 PSin_c;
+
+layout(binding = 0) uniform sampler2D TextureSampler;
+
+layout(location = 0) out vec4 SV_Target0;
+
+// For all settings: 1.0 = 100% 0.5=50% 1.5 = 150%
+vec4 ContrastSaturationBrightness(vec4 color)
+{
+	float brt = params.x;
+	float con = params.y;
+	float sat = params.z;
+
+	// Increase or decrease these values to adjust r, g and b color channels separately
+	const float AvgLumR = 0.5;
+	const float AvgLumG = 0.5;
+	const float AvgLumB = 0.5;
+
+	const vec3 LumCoeff = vec3(0.2125, 0.7154, 0.0721);
+
+	vec3 AvgLumin = vec3(AvgLumR, AvgLumG, AvgLumB);
+	vec3 brtColor = color.rgb * brt;
+	float dot_intensity = dot(brtColor, LumCoeff);
+	vec3 intensity = vec3(dot_intensity, dot_intensity, dot_intensity);
+	vec3 satColor = mix(intensity, brtColor, sat);
+	vec3 conColor = mix(AvgLumin, satColor, con);
+
+	color.rgb = conColor;
+	return color;
+}
+
+
+void ps_main()
+{
+	vec4 c = texture(TextureSampler, PSin_t);
+	SV_Target0 = ContrastSaturationBrightness(c);
+}
+
+
+#endif

bin/resources/shaders/opengl/tfx_vgs.glsl

@@ -0,0 +1,225 @@
+// SPDX-FileCopyrightText: 2002-2025 PCSX2 Dev Team
+// SPDX-License-Identifier: GPL-3.0+
+
+//#version 420 // Keep it for text editor detection
+
+layout(std140, binding = 1) uniform cb20
+{
+	vec2  VertexScale;
+	vec2  VertexOffset;
+
+	vec2  TextureScale;
+	vec2  TextureOffset;
+
+	vec2  PointSize;
+	uint  MaxDepth;
+	uint  pad_cb20;
+};
+
+#ifdef VERTEX_SHADER
+
+out SHADER
+{
+	vec4 t_float;
+	vec4 t_int;
+	#if VS_IIP != 0
+		vec4 c;
+	#else
+		flat vec4 c;
+	#endif
+} VSout;
+
+const float exp_min32 = exp2(-32.0f);
+
+#if VS_EXPAND == 0
+
+layout(location = 0) in vec2  i_st;
+layout(location = 2) in vec4  i_c;
+layout(location = 3) in float i_q;
+layout(location = 4) in uvec2 i_p;
+layout(location = 5) in uint  i_z;
+layout(location = 6) in uvec2 i_uv;
+layout(location = 7) in vec4  i_f;
+
+void texture_coord()
+{
+	vec2 uv = vec2(i_uv) - TextureOffset;
+	vec2 st = i_st - TextureOffset;
+
+	// Float coordinate
+	VSout.t_float.xy = st;
+	VSout.t_float.w  = i_q;
+
+	// Integer coordinate => normalized
+	VSout.t_int.xy = uv * TextureScale;
+#if VS_FST
+	// Integer coordinate => integral
+	VSout.t_int.zw = uv;
+#else
+	// Some games uses float coordinate for post-processing effect
+	VSout.t_int.zw = st / TextureScale;
+#endif
+}
+
+void vs_main()
+{
+	// Clamp to max depth, gs doesn't wrap
+	highp uint z = min(i_z, MaxDepth);
+
+	// pos -= 0.05 (1/320 pixel) helps avoiding rounding problems (integral part of pos is usually 5 digits, 0.05 is about as low as we can go)
+	// example: ceil(afterseveralvertextransformations(y = 133)) => 134 => line 133 stays empty
+	// input granularity is 1/16 pixel, anything smaller than that won't step drawing up/left by one pixel
+	// example: 133.0625 (133 + 1/16) should start from line 134, ceil(133.0625 - 0.05) still above 133
+	gl_Position.xy = vec2(i_p) - vec2(0.05f, 0.05f);
+	gl_Position.xy = gl_Position.xy * VertexScale - VertexOffset;
+	gl_Position.z = float(z) * exp_min32;
+	gl_Position.w = 1.0f;
+
+	texture_coord();
+
+	VSout.c = i_c;
+	VSout.t_float.z = i_f.x; // pack for with texture
+
+	#if VS_POINT_SIZE
+		gl_PointSize = PointSize.x;
+	#endif
+}
+
+#else // VS_EXPAND
+
+struct RawVertex
+{
+	vec2 ST;
+	uint RGBA;
+	float Q;
+	uint XY;
+	uint Z;
+	uint UV;
+	uint FOG;
+};
+
+layout(std140, binding = 2) readonly buffer VertexBuffer {
+	RawVertex vertex_buffer[];
+};
+
+struct ProcessedVertex
+{
+	vec4 p;
+	vec4 t_float;
+	vec4 t_int;
+	vec4 c;
+};
+
+ProcessedVertex load_vertex(uint index)
+{
+#if defined(GL_ARB_shader_draw_parameters) && GL_ARB_shader_draw_parameters
+	RawVertex rvtx = vertex_buffer[index + gl_BaseVertexARB];
+#else
+	RawVertex rvtx = vertex_buffer[index];
+#endif
+
+	vec2 i_st = rvtx.ST;
+	vec4 i_c = vec4(uvec4(bitfieldExtract(rvtx.RGBA, 0, 8), bitfieldExtract(rvtx.RGBA, 8, 8),
+	                      bitfieldExtract(rvtx.RGBA, 16, 8), bitfieldExtract(rvtx.RGBA, 24, 8)));
+	float i_q = rvtx.Q;
+	uvec2 i_p = uvec2(bitfieldExtract(rvtx.XY, 0, 16), bitfieldExtract(rvtx.XY, 16, 16));
+	uint i_z = rvtx.Z;
+	uvec2 i_uv = uvec2(bitfieldExtract(rvtx.UV, 0, 16), bitfieldExtract(rvtx.UV, 16, 16));
+	vec4 i_f = unpackUnorm4x8(rvtx.FOG);
+
+	ProcessedVertex vtx;
+
+	uint z = min(i_z, MaxDepth);
+	vtx.p.xy = vec2(i_p) - vec2(0.05f, 0.05f);
+	vtx.p.xy = vtx.p.xy * VertexScale - VertexOffset;
+	vtx.p.z = float(z) * exp_min32;
+	vtx.p.w = 1.0f;
+
+	vec2 uv = vec2(i_uv) - TextureOffset;
+	vec2 st = i_st - TextureOffset;
+
+	vtx.t_float.xy = st;
+	vtx.t_float.w  = i_q;
+
+	vtx.t_int.xy = uv * TextureScale;
+#if VS_FST
+	vtx.t_int.zw = uv;
+#else
+	vtx.t_int.zw = st / TextureScale;
+#endif
+
+	vtx.c = i_c;
+	vtx.t_float.z = i_f.x;
+
+	return vtx;
+}
+
+void main()
+{
+	ProcessedVertex vtx;
+
+#if defined(GL_ARB_shader_draw_parameters) && GL_ARB_shader_draw_parameters
+	uint vid = uint(gl_VertexID - gl_BaseVertexARB);
+#else
+	uint vid = uint(gl_VertexID);
+#endif
+
+#if VS_EXPAND == 1 // Point
+
+	vtx = load_vertex(vid >> 2);
+
+	vtx.p.x += ((vid & 1u) != 0u) ? PointSize.x : 0.0f;
+	vtx.p.y += ((vid & 2u) != 0u) ? PointSize.y : 0.0f;
+
+#elif VS_EXPAND == 2 // Line
+
+	uint vid_base = vid >> 2;
+	bool is_bottom = (vid & 2u) != 0u;
+	bool is_right = (vid & 1u) != 0u;
+	uint vid_other = is_bottom ? vid_base - 1 : vid_base + 1;
+	vtx = load_vertex(vid_base);
+	ProcessedVertex other = load_vertex(vid_other);
+
+	vec2 line_vector = normalize(vtx.p.xy - other.p.xy);
+	vec2 line_normal = vec2(line_vector.y, -line_vector.x);
+	vec2 line_width = (line_normal * PointSize) / 2;
+	// line_normal is inverted for bottom point
+	vec2 offset = ((uint(is_bottom) ^ uint(is_right)) != 0u) ? line_width : -line_width;
+	vtx.p.xy += offset;
+
+	// Lines will be run as (0 1 2) (1 2 3)
+	// This means that both triangles will have a point based off the top line point as their first point
+	// So we don't have to do anything for !IIP
+
+#elif VS_EXPAND == 3 // Sprite
+
+	// Sprite points are always in pairs
+	uint vid_base = vid >> 1;
+	uint vid_lt = vid_base & ~1u;
+	uint vid_rb = vid_base | 1u;
+
+	ProcessedVertex lt = load_vertex(vid_lt);
+	ProcessedVertex rb = load_vertex(vid_rb);
+	vtx = rb;
+
+	bool is_right = ((vid & 1u) != 0u);
+	vtx.p.x = is_right ? lt.p.x : vtx.p.x;
+	vtx.t_float.x = is_right ? lt.t_float.x : vtx.t_float.x;
+	vtx.t_int.xz = is_right ? lt.t_int.xz : vtx.t_int.xz;
+
+	bool is_bottom = ((vid & 2u) != 0u);
+	vtx.p.y = is_bottom ? lt.p.y : vtx.p.y;
+	vtx.t_float.y = is_bottom ? lt.t_float.y : vtx.t_float.y;
+	vtx.t_int.yw = is_bottom ? lt.t_int.yw : vtx.t_int.yw;
+
+#endif
+
+	gl_Position = vtx.p;
+	VSout.t_float = vtx.t_float;
+	VSout.t_int = vtx.t_int;
+	VSout.c = vtx.c;
+}
+
+#endif // VS_EXPAND
+
+#endif // VERTEX_SHADER

bin/resources/shaders/vulkan/cas.glsl

@@ -0,0 +1,71 @@
+#version 460 core
+#extension GL_EXT_samplerless_texture_functions : require
+
+// Based on CAS_Shader.glsl
+// Copyright(c) 2019 Advanced Micro Devices, Inc.All rights reserved.
+// Permission is hereby granted, free of charge, to any person obtaining a copy
+// of this software and associated documentation files(the "Software"), to deal
+// in the Software without restriction, including without limitation the rights
+// to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+// copies of the Software, and to permit persons to whom the Software is
+// furnished to do so, subject to the following conditions :
+// The above copyright notice and this permission notice shall be included in
+// all copies or substantial portions of the Software.
+// THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+// IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+// FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.IN NO EVENT SHALL THE
+// AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+// LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+// OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
+// THE SOFTWARE.
+
+layout(push_constant) uniform const_buffer
+{
+    uvec4 const0;
+    uvec4 const1;
+    ivec2 srcOffset;
+};
+
+layout(set=0, binding=0) uniform texture2D imgSrc;
+layout(set=0, binding=1, rgba8) uniform writeonly image2D imgDst;
+layout(constant_id=0) const int sharpenOnly = 0;
+
+#define A_GPU 1
+#define A_GLSL 1
+
+#include "ffx_a.h"
+
+AF3 CasLoad(ASU2 p)
+{
+    return texelFetch(imgSrc, srcOffset + ivec2(p), 0).rgb;
+}
+
+// Lets you transform input from the load into a linear color space between 0 and 1. See ffx_cas.h
+// In this case, our input is already linear and between 0 and 1
+void CasInput(inout AF1 r, inout AF1 g, inout AF1 b) {}
+
+#include "ffx_cas.h"
+
+layout(local_size_x=64) in;
+void main()
+{
+    // Do remapping of local xy in workgroup for a more PS-like swizzle pattern.
+    AU2 gxy = ARmp8x8(gl_LocalInvocationID.x)+AU2(gl_WorkGroupID.x<<4u,gl_WorkGroupID.y<<4u);
+
+    // Filter.
+    AF4 c = vec4(0.0f);
+    CasFilter(c.r, c.g, c.b, gxy, const0, const1, sharpenOnly != 0);
+    imageStore(imgDst, ASU2(gxy), c);
+    gxy.x += 8u;
+
+    CasFilter(c.r, c.g, c.b, gxy, const0, const1, sharpenOnly != 0);
+    imageStore(imgDst, ASU2(gxy), c);
+    gxy.y += 8u;
+
+    CasFilter(c.r, c.g, c.b, gxy, const0, const1, sharpenOnly != 0);
+    imageStore(imgDst, ASU2(gxy), c);
+    gxy.x -= 8u;
+
+    CasFilter(c.r, c.g, c.b, gxy, const0, const1, sharpenOnly != 0);
+    imageStore(imgDst, ASU2(gxy), c);
+}

bin/resources/shaders/vulkan/convert.glsl

@@ -0,0 +1,618 @@
+// SPDX-FileCopyrightText: 2002-2025 PCSX2 Dev Team
+// SPDX-License-Identifier: GPL-3.0+
+
+#ifdef VERTEX_SHADER
+
+layout(location = 0) in vec4 a_pos;
+layout(location = 1) in vec2 a_tex;
+
+layout(location = 0) out vec2 v_tex;
+
+void main()
+{
+	gl_Position = vec4(a_pos.x, -a_pos.y, a_pos.z, a_pos.w);
+	v_tex = a_tex;
+}
+
+#endif
+
+#ifdef FRAGMENT_SHADER
+
+layout(location = 0) in vec2 v_tex;
+
+#if defined(ps_convert_rgba8_16bits) || defined(ps_convert_float32_32bits)
+layout(location = 0) out uint o_col0;
+#elif !defined(ps_datm1) && \
+	!defined(ps_datm0) && \
+	!defined(ps_datm1_rta_correction) && \
+	!defined(ps_datm0_rta_correction) && \
+	!defined(ps_convert_rgba8_float32) && \
+	!defined(ps_convert_rgba8_float24) && \
+	!defined(ps_convert_rgba8_float16) && \
+	!defined(ps_convert_rgb5a1_float16) && \
+	!defined(ps_convert_rgba8_float32_biln) && \
+	!defined(ps_convert_rgba8_float24_biln) && \
+	!defined(ps_convert_rgba8_float16_biln) && \
+	!defined(ps_convert_rgb5a1_float16_biln) && \
+	!defined(ps_depth_copy)
+layout(location = 0) out vec4 o_col0;
+#endif
+
+layout(set = 0, binding = 0) uniform sampler2D samp0;
+
+vec4 sample_c(vec2 uv)
+{
+	return texture(samp0, uv);
+}
+
+#ifdef ps_copy
+void ps_copy()
+{
+	o_col0 = sample_c(v_tex);
+}
+#endif
+
+#ifdef ps_depth_copy
+void ps_depth_copy()
+{
+  gl_FragDepth = sample_c(v_tex).r;
+}
+#endif
+
+#ifdef ps_downsample_copy
+layout(push_constant) uniform cb10
+{
+	ivec2 ClampMin;
+	int DownsampleFactor;
+	int pad0;
+	float Weight;
+	vec3 pad1;
+};
+void ps_downsample_copy()
+{
+	ivec2 coord = max(ivec2(gl_FragCoord.xy) * DownsampleFactor, ClampMin);
+	vec4 result = vec4(0);
+	for (int yoff = 0; yoff < DownsampleFactor; yoff++)
+	{
+		for (int xoff = 0; xoff < DownsampleFactor; xoff++)
+			result += texelFetch(samp0, coord + ivec2(xoff, yoff), 0);
+	}
+	o_col0 = result / Weight;
+}
+#endif
+
+#ifdef ps_filter_transparency
+void ps_filter_transparency()
+{
+	vec4 c = sample_c(v_tex);
+	o_col0 = vec4(c.rgb, 1.0);
+}
+#endif
+
+#ifdef ps_convert_rgba8_16bits
+// Need to be careful with precision here, it can break games like Spider-Man 3 and Dogs Life
+void ps_convert_rgba8_16bits()
+{
+	uvec4 i = uvec4(sample_c(v_tex) * vec4(255.5f, 255.5f, 255.5f, 255.5f));
+
+	o_col0 = ((i.x & 0x00F8u) >> 3) | ((i.y & 0x00F8u) << 2) | ((i.z & 0x00f8u) << 7) | ((i.w & 0x80u) << 8);
+}
+#endif
+
+#ifdef ps_datm1
+void ps_datm1()
+{
+	if(sample_c(v_tex).a < (127.5f / 255.0f)) // >= 0x80 pass
+		discard;
+}
+#endif
+
+#ifdef ps_datm0
+void ps_datm0()
+{
+	if((127.5f / 255.0f) < sample_c(v_tex).a) // < 0x80 pass (== 0x80 should not pass)
+		discard;
+}
+#endif
+
+#ifdef ps_datm1_rta_correction
+void ps_datm1_rta_correction()
+{
+	if(sample_c(v_tex).a < (254.5f / 255.0f)) // >= 0x80 pass
+		discard;
+}
+#endif
+
+#ifdef ps_datm0_rta_correction
+void ps_datm0_rta_correction()
+{
+	if((254.5f / 255.0f) < sample_c(v_tex).a) // < 0x80 pass (== 0x80 should not pass)
+		discard;
+}
+#endif
+
+#ifdef ps_rta_correction
+void ps_rta_correction()
+{
+	vec4 value = sample_c(v_tex);
+	o_col0 = vec4(value.rgb, value.a / (128.25f / 255.0f));
+}
+#endif
+
+#ifdef ps_rta_decorrection
+void ps_rta_decorrection()
+{
+	vec4 value = sample_c(v_tex);
+	o_col0 = vec4(value.rgb, value.a * (128.25f / 255.0f));
+}
+#endif
+
+#ifdef ps_colclip_init
+void ps_colclip_init()
+{
+	vec4 value = sample_c(v_tex);
+	o_col0 = vec4(roundEven(value.rgb * 255.0f) / 65535.0f, value.a);
+}
+#endif
+
+#ifdef ps_colclip_resolve
+void ps_colclip_resolve()
+{
+	vec4 value = sample_c(v_tex);
+	o_col0 = vec4(vec3(uvec3(value.rgb * 65535.5f) & 255u) / 255.0f, value.a);
+}
+#endif
+
+#ifdef ps_convert_float32_32bits
+void ps_convert_float32_32bits()
+{
+	// Convert a vec32 depth texture into a 32 bits UINT texture
+	o_col0 = uint(exp2(32.0f) * sample_c(v_tex).r);
+}
+#endif
+
+#ifdef ps_convert_float32_rgba8
+void ps_convert_float32_rgba8()
+{
+	// Convert a vec32 depth texture into a RGBA color texture
+	uint d = uint(sample_c(v_tex).r * exp2(32.0f));
+	o_col0 = vec4(uvec4((d & 0xFFu), ((d >> 8) & 0xFFu), ((d >> 16) & 0xFFu), (d >> 24))) / vec4(255.0);
+}
+#endif
+
+#ifdef ps_convert_float16_rgb5a1
+void ps_convert_float16_rgb5a1()
+{
+	// Convert a vec32 (only 16 lsb) depth into a RGB5A1 color texture
+	uint d = uint(sample_c(v_tex).r * exp2(32.0f));
+	o_col0 = vec4(uvec4(d << 3, d >> 2, d >> 7, d >> 8) & uvec4(0xf8, 0xf8, 0xf8, 0x80)) / 255.0f;
+}
+#endif
+
+float rgba8_to_depth32(vec4 unorm)
+{
+	uvec4 c = uvec4(unorm * vec4(255.5f));
+	return float(c.r | (c.g << 8) | (c.b << 16) | (c.a << 24)) * exp2(-32.0f);
+}
+
+float rgba8_to_depth24(vec4 unorm)
+{
+	uvec3 c = uvec3(unorm.rgb * vec3(255.5f));
+	return float(c.r | (c.g << 8) | (c.b << 16)) * exp2(-32.0f);
+}
+
+float rgba8_to_depth16(vec4 unorm)
+{
+	uvec2 c = uvec2(unorm.rg * vec2(255.5f));
+	return float(c.r | (c.g << 8)) * exp2(-32.0f);
+}
+
+float rgb5a1_to_depth16(vec4 unorm)
+{
+	uvec4 c = uvec4(unorm * vec4(255.5f));
+	return float(((c.r & 0xF8u) >> 3) | ((c.g & 0xF8u) << 2) | ((c.b & 0xF8u) << 7) | ((c.a & 0x80u) << 8)) * exp2(-32.0f);
+}
+
+#ifdef ps_convert_float32_float24
+void ps_convert_float32_float24()
+{
+	// Truncates depth value to 24bits
+	uint d = uint(sample_c(v_tex).r * exp2(32.0f)) & 0xFFFFFFu;
+	gl_FragDepth = float(d) * exp2(-32.0f);
+}
+#endif
+
+#ifdef ps_convert_rgba8_float32
+void ps_convert_rgba8_float32()
+{
+	// Convert an RGBA texture into a float depth texture
+	gl_FragDepth = rgba8_to_depth32(sample_c(v_tex));
+}
+#endif
+
+#ifdef ps_convert_rgba8_float24
+void ps_convert_rgba8_float24()
+{
+	// Same as above but without the alpha channel (24 bits Z)
+
+	// Convert an RGBA texture into a float depth texture
+	gl_FragDepth = rgba8_to_depth24(sample_c(v_tex));
+}
+#endif
+
+#ifdef ps_convert_rgba8_float16
+void ps_convert_rgba8_float16()
+{
+	// Same as above but without the A/B channels (16 bits Z)
+
+	// Convert an RGBA texture into a float depth texture
+	gl_FragDepth = rgba8_to_depth16(sample_c(v_tex));
+}
+#endif
+
+#ifdef ps_convert_rgb5a1_float16
+void ps_convert_rgb5a1_float16()
+{
+	// Convert an RGB5A1 (saved as RGBA8) color to a 16 bit Z
+	gl_FragDepth = rgb5a1_to_depth16(sample_c(v_tex));
+}
+#endif
+
+#define SAMPLE_RGBA_DEPTH_BILN(CONVERT_FN) \
+	ivec2 dims = textureSize(samp0, 0); \
+	vec2 top_left_f = v_tex * vec2(dims) - 0.5f; \
+	ivec2 top_left = ivec2(floor(top_left_f)); \
+	ivec4 coords = clamp(ivec4(top_left, top_left + 1), ivec4(0), dims.xyxy - 1); \
+	vec2 mix_vals = fract(top_left_f); \
+	float depthTL = CONVERT_FN(texelFetch(samp0, coords.xy, 0)); \
+	float depthTR = CONVERT_FN(texelFetch(samp0, coords.zy, 0)); \
+	float depthBL = CONVERT_FN(texelFetch(samp0, coords.xw, 0)); \
+	float depthBR = CONVERT_FN(texelFetch(samp0, coords.zw, 0)); \
+	gl_FragDepth = mix(mix(depthTL, depthTR, mix_vals.x), mix(depthBL, depthBR, mix_vals.x), mix_vals.y);
+
+#ifdef ps_convert_rgba8_float32_biln
+void ps_convert_rgba8_float32_biln()
+{
+	// Convert an RGBA texture into a float depth texture
+	SAMPLE_RGBA_DEPTH_BILN(rgba8_to_depth32);
+}
+#endif
+
+#ifdef ps_convert_rgba8_float24_biln
+void ps_convert_rgba8_float24_biln()
+{
+	// Same as above but without the alpha channel (24 bits Z)
+
+	// Convert an RGBA texture into a float depth texture
+	SAMPLE_RGBA_DEPTH_BILN(rgba8_to_depth24);
+}
+#endif
+
+#ifdef ps_convert_rgba8_float16_biln
+void ps_convert_rgba8_float16_biln()
+{
+	// Same as above but without the A/B channels (16 bits Z)
+
+	// Convert an RGBA texture into a float depth texture
+	SAMPLE_RGBA_DEPTH_BILN(rgba8_to_depth16);
+}
+#endif
+
+#ifdef ps_convert_rgb5a1_float16_biln
+void ps_convert_rgb5a1_float16_biln()
+{
+	// Convert an RGB5A1 (saved as RGBA8) color to a 16 bit Z
+	SAMPLE_RGBA_DEPTH_BILN(rgb5a1_to_depth16);
+}
+#endif
+
+#ifdef ps_convert_rgb5a1_8i
+layout(push_constant) uniform cb10
+{
+	uint SBW;
+	uint DBW;
+	uint PSM;
+	float cb_pad1;
+	float ScaleFactor;
+	vec3 cb_pad2;
+};
+
+void ps_convert_rgb5a1_8i()
+{
+	// Convert a RGB5A1 texture into a 8 bits packed texture
+	// Input column: 16x2 RGB5A1 pixels
+	// 0: 16 RGBA
+	// 1: 16 RGBA
+	// Output column: 16x4 Index pixels
+	// 0: 16 R5G2
+	// 1: 16 R5G2
+	// 2: 16 G2B5A1
+	// 3: 16 G2B5A1
+	
+	uvec2 pos = uvec2(gl_FragCoord.xy);
+
+	// Collapse separate R G B A areas into their base pixel
+	uvec2 column = (pos & ~uvec2(0u, 3u)) / uvec2(1,2);
+	uvec2 subcolumn = (pos & uvec2(0u, 1u));
+	column.x -= (column.x / 128) * 64;
+	column.y += (column.y / 32) * 32;
+	
+	// Deal with swizzling differences
+	if ((PSM & 0x8) != 0) // PSMCT16S
+	{
+		if ((pos.x & 32) != 0)
+		{
+			column.y += 32; // 4 columns high times 4 to get bottom 4 blocks
+			column.x &= ~32;
+		}
+		
+		if ((pos.x & 64) != 0)
+		{
+			column.x -= 32;
+		}
+		
+		if (((pos.x & 16) != 0) != ((pos.y & 16) != 0))
+		{
+			column.x ^= 16; 
+			column.y ^= 8;
+		}
+		
+		if ((PSM & 0x30) != 0) // PSMZ16S - Untested but hopefully ok if anything uses it.
+		{
+			column.x ^= 32;
+			column.y ^= 16;
+		}
+	}
+	else // PSMCT16
+	{
+		if ((pos.y & 32) != 0)
+		{
+			column.y -= 16;
+			column.x += 32;
+		}
+		
+		if ((pos.x & 96) != 0)
+		{
+			uint multi = (pos.x & 96) / 32;
+			column.y += 16 * multi; // 4 columns high times 4 to get bottom 4 blocks
+			column.x -= (pos.x & 96);
+		}
+		
+		if (((pos.x & 16) != 0) != ((pos.y & 16) != 0))
+		{
+			column.x ^= 16; 
+			column.y ^= 8;
+		}
+		
+		if ((PSM & 0x30) != 0) // PSMZ16 - Untested but hopefully ok if anything uses it.
+		{
+			column.x ^= 32;
+			column.y ^= 32;
+		}
+	}
+	uvec2 coord = column | subcolumn;
+	
+	// Compensate for potentially differing page pitch.
+	uvec2 block_xy = coord / uvec2(64u, 64u);
+	uint block_num = (block_xy.y * (DBW / 128u)) + block_xy.x;
+	uvec2 block_offset = uvec2((block_num % (SBW / 64u)) * 64u, (block_num / (SBW / 64u)) * 64u);
+	coord = (coord % uvec2(64u, 64u)) + block_offset;
+
+	// Apply offset to cols 1 and 2
+	uint is_col23 = pos.y & 4u;
+	uint is_col13 = pos.y & 2u;
+	uint is_col12 = is_col23 ^ (is_col13 << 1);
+	coord.x ^= is_col12; // If cols 1 or 2, flip bit 3 of x
+
+	if (floor(ScaleFactor) != ScaleFactor)
+		coord = uvec2(vec2(coord) * ScaleFactor);
+	else
+		coord *= uvec2(ScaleFactor);
+
+	vec4 pixel = texelFetch(samp0, ivec2(coord), 0);
+	uvec4 denorm_c = uvec4(pixel * 255.5f);
+	if ((pos.y & 2u) == 0u)
+	{
+		uint red = (denorm_c.r >> 3) & 0x1Fu;
+		uint green = (denorm_c.g >> 3) & 0x1Fu;
+		float sel0 = float(((green << 5) | red) & 0xFF) / 255.0f;
+		
+		o_col0 = vec4(sel0);
+	}
+	else
+	{
+		uint green = (denorm_c.g >> 3) & 0x1Fu;
+		uint blue = (denorm_c.b >> 3) & 0x1Fu;
+		uint alpha = denorm_c.a & 0x80u;
+		float sel0 = float((alpha | (blue << 2) | (green >> 3)) & 0xFF) / 255.0f;
+
+		o_col0 = vec4(sel0);
+	}
+}
+#endif
+
+#ifdef ps_convert_rgba_8i
+layout(push_constant) uniform cb10
+{
+	uint SBW;
+	uint DBW;
+	uint PSM;
+	float cb_pad1;
+	float ScaleFactor;
+	vec3 cb_pad2;
+};
+
+void ps_convert_rgba_8i()
+{
+	// Convert a RGBA texture into a 8 bits packed texture
+	// Input column: 8x2 RGBA pixels
+	// 0: 8 RGBA
+	// 1: 8 RGBA
+	// Output column: 16x4 Index pixels
+	// 0: 8 R | 8 B
+	// 1: 8 R | 8 B
+	// 2: 8 G | 8 A
+	// 3: 8 G | 8 A
+	uvec2 pos = uvec2(gl_FragCoord.xy);
+
+	// Collapse separate R G B A areas into their base pixel
+	uvec2 block = (pos & ~uvec2(15u, 3u)) >> 1;
+	uvec2 subblock = pos & uvec2(7u, 1u);
+	uvec2 coord = block | subblock;
+
+	// Compensate for potentially differing page pitch.
+	uvec2 block_xy = coord / uvec2(64u, 32u);
+	uint block_num = (block_xy.y * (DBW / 128u)) + block_xy.x;
+	uvec2 block_offset = uvec2((block_num % (SBW / 64u)) * 64u, (block_num / (SBW / 64u)) * 32u);
+	coord = (coord % uvec2(64u, 32u)) + block_offset;
+
+	// Apply offset to cols 1 and 2
+	uint is_col23 = pos.y & 4u;
+	uint is_col13 = pos.y & 2u;
+	uint is_col12 = is_col23 ^ (is_col13 << 1);
+	coord.x ^= is_col12; // If cols 1 or 2, flip bit 3 of x
+
+	if (floor(ScaleFactor) != ScaleFactor)
+		coord = uvec2(vec2(coord) * ScaleFactor);
+	else
+		coord *= uvec2(ScaleFactor);
+
+	vec4 pixel = texelFetch(samp0, ivec2(coord), 0);
+	vec2  sel0 = (pos.y & 2u) == 0u ? pixel.rb : pixel.ga;
+	float sel1 = (pos.x & 8u) == 0u ? sel0.x : sel0.y;
+	o_col0 = vec4(sel1); // Divide by something here?
+}
+#endif
+
+#ifdef ps_convert_clut_4
+layout(push_constant) uniform cb10
+{
+	uvec2 offset;
+	uint doffset;
+	uint cb_pad1;
+	float scale;
+	vec3 cb_pad2;
+};
+
+void ps_convert_clut_4()
+{
+	// CLUT4 is easy, just two rows of 8x8.
+	uint index = uint(gl_FragCoord.x) + doffset;
+	uvec2 pos = uvec2(index % 8u, index / 8u);
+
+	ivec2 final = ivec2(floor(vec2(offset + pos) * vec2(scale)));
+	o_col0 = texelFetch(samp0, final, 0);
+}
+#endif
+
+#ifdef ps_convert_clut_8
+layout(push_constant) uniform cb10
+{
+	uvec2 offset;
+	uint doffset;
+	uint cb_pad1;
+	float scale;
+	vec3 cb_pad2;
+};
+
+void ps_convert_clut_8()
+{
+	uint index = min(uint(gl_FragCoord.x) + doffset, 255u);
+
+	// CLUT is arranged into 8 groups of 16x2, with the top-right and bottom-left quadrants swapped.
+	// This can probably be done better..
+	uint subgroup = (index / 8u) % 4u;
+	uvec2 pos;
+	pos.x = (index % 8u) + ((subgroup >= 2u) ? 8u : 0u);
+	pos.y = ((index / 32u) * 2u) + (subgroup % 2u);
+
+	ivec2 final = ivec2(floor(vec2(offset + pos) * vec2(scale)));
+	o_col0 = texelFetch(samp0, final, 0);
+}
+#endif
+
+#ifdef ps_yuv
+layout(push_constant) uniform cb10
+{
+	int EMODA;
+	int EMODC;
+};
+
+void ps_yuv()
+{
+	vec4 i = sample_c(v_tex);
+	vec4 o = vec4(0.0f);
+
+	mat3 rgb2yuv;
+	rgb2yuv[0] = vec3(0.587, -0.311, -0.419);
+	rgb2yuv[1] = vec3(0.114, 0.500, -0.081);
+	rgb2yuv[2] = vec3(0.299, -0.169, 0.500);
+
+	vec3 yuv = rgb2yuv * i.gbr;
+
+	float Y = float(0xDB)/255.0f * yuv.x + float(0x10)/255.0f;
+	float Cr = float(0xE0)/255.0f * yuv.y + float(0x80)/255.0f;
+	float Cb = float(0xE0)/255.0f * yuv.z + float(0x80)/255.0f;
+
+	switch(EMODA)
+	{
+		case 0:
+			o.a = i.a;
+			break;
+		case 1:
+			o.a = Y;
+			break;
+		case 2:
+			o.a = Y/2.0f;
+			break;
+		case 3:
+			o.a = 0.0f;
+			break;
+	}
+
+	switch(EMODC)
+	{
+		case 0:
+			o.rgb = i.rgb;
+			break;
+		case 1:
+			o.rgb = vec3(Y);
+			break;
+		case 2:
+			o.rgb = vec3(Y, Cb, Cr);
+			break;
+		case 3:
+			o.rgb = vec3(i.a);
+			break;
+	}
+
+	o_col0 = o;
+}
+#endif
+
+#if defined(ps_stencil_image_init_0) || defined(ps_stencil_image_init_1) || defined(ps_stencil_image_init_2) || defined(ps_stencil_image_init_3)
+
+void main()
+{
+	o_col0 = vec4(0x7FFFFFFF);
+
+	#ifdef ps_stencil_image_init_0
+		if((127.5f / 255.0f) < sample_c(v_tex).a) // < 0x80 pass (== 0x80 should not pass)
+			o_col0 = vec4(-1);
+	#endif
+	#ifdef ps_stencil_image_init_1
+		if(sample_c(v_tex).a < (127.5f / 255.0f)) // >= 0x80 pass
+			o_col0 = vec4(-1);
+	#endif
+	#ifdef ps_stencil_image_init_2
+		if((254.5f / 255.0f) < sample_c(v_tex).a) // < 0x80 pass (== 0x80 should not pass)
+			o_col0 = vec4(-1);
+	#endif
+	#ifdef ps_stencil_image_init_3
+		if(sample_c(v_tex).a < (254.5f / 255.0f)) // >= 0x80 pass
+			o_col0 = vec4(-1);
+	#endif
+}
+#endif
+
+#endif

bin/resources/shaders/vulkan/imgui.glsl

@@ -0,0 +1,42 @@
+// SPDX-FileCopyrightText: 2002-2025 PCSX2 Dev Team
+// SPDX-License-Identifier: GPL-3.0+
+
+#ifdef VERTEX_SHADER
+
+layout(location = 0) in vec2 Position;
+layout(location = 1) in vec2 UV;
+layout(location = 2) in vec4 Color;
+
+layout(push_constant) uniform PushConstants
+{
+	vec2 uScale;
+	vec2 uTranslate;
+};
+
+layout(location = 0) out vec2 Frag_UV;
+layout(location = 1) out vec4 Frag_Color;
+
+void vs_main()
+{
+	Frag_UV = UV;
+	Frag_Color = Color;
+	gl_Position = vec4(Position * uScale + uTranslate, 0.0f, 1.0f);
+}
+
+#endif
+
+#ifdef FRAGMENT_SHADER
+
+layout(binding = 0) uniform sampler2D Texture;
+
+layout(location = 0) in vec2 Frag_UV;
+layout(location = 1) in vec4 Frag_Color;
+
+layout(location = 0) out vec4 Out_Color;
+
+void ps_main()
+{
+	Out_Color = Frag_Color * texture(Texture, Frag_UV.st);
+}
+
+#endif

bin/resources/shaders/vulkan/interlace.glsl

@@ -0,0 +1,203 @@
+// SPDX-FileCopyrightText: 2002-2025 PCSX2 Dev Team
+// SPDX-License-Identifier: GPL-3.0+
+
+#ifdef VERTEX_SHADER
+
+layout(location = 0) in vec4 a_pos;
+layout(location = 1) in vec2 a_tex;
+
+layout(location = 0) out vec2 v_tex;
+
+void main()
+{
+	gl_Position = vec4(a_pos.x, -a_pos.y, a_pos.z, a_pos.w);
+	v_tex = a_tex;
+}
+
+#endif
+
+#ifdef FRAGMENT_SHADER
+
+layout(location = 0) in vec2 v_tex;
+layout(location = 0) out vec4 o_col0;
+
+layout(push_constant) uniform cb0
+{
+	vec4 ZrH;
+};
+
+layout(set = 0, binding = 0) uniform sampler2D samp0;
+
+
+// Weave shader
+#ifdef ps_main0
+void ps_main0()
+{
+	const int idx   = int(ZrH.x);          // buffer index passed from CPU
+	const int field = idx & 1;             // current field
+	const int vpos  = int(gl_FragCoord.y); // vertical position of destination texture
+
+	if ((vpos & 1) == field)
+		o_col0 = textureLod(samp0, v_tex, 0);
+	else
+		discard;
+}
+#endif
+
+
+// Bob shader
+#ifdef ps_main1
+void ps_main1()
+{
+	o_col0 = textureLod(samp0, v_tex, 0);
+}
+#endif
+
+
+// Blend shader
+#ifdef ps_main2
+void ps_main2()
+{
+	vec2 vstep = vec2(0.0f, ZrH.y);
+	vec4 c0 = textureLod(samp0, v_tex - vstep, 0);
+	vec4 c1 = textureLod(samp0, v_tex, 0);
+	vec4 c2 = textureLod(samp0, v_tex + vstep, 0);
+
+	o_col0 = (c0 + c1 * 2.0f + c2) / 4.0f;
+}
+#endif
+
+
+// MAD shader - buffering
+#ifdef ps_main3
+void ps_main3()
+{
+	// We take half the lines from the current frame and stores them in the MAD frame buffer.
+	// the MAD frame buffer is split in 2 consecutive banks of 2 fields each, the fields in each bank
+	// are interleaved (top field at even lines and bottom field at odd lines).
+	// When the source texture has an odd vres, the first line of bank 1 would be an odd index
+	// causing the wrong lines to be discarded, so a vertical offset (lofs) is added to the vertical
+	// position of the destination texture to force the proper field alignment
+
+	const int  idx    = int(ZrH.x);                               // buffer index passed from CPU
+	const int  bank   = idx >> 1;                                 // current bank
+	const int  field  = idx & 1;                                  // current field
+	const int  vres   = int(ZrH.z) >> 1;                          // vertical resolution of source texture
+	const int  lofs   = ((((vres + 1) >> 1) << 1) - vres) & bank; // line alignment offset for bank 1
+	const int  vpos   = int(gl_FragCoord.y) + lofs;               // vertical position of destination texture
+
+	// if the index of current destination line belongs to the current fiels we update it, otherwise
+	// we leave the old line in the destination buffer
+	if ((vpos & 1) == field)
+		o_col0 = textureLod(samp0, v_tex, 0);
+	else
+		discard;
+}
+#endif
+
+
+// MAD shader - reconstruction
+#ifdef ps_main4
+void ps_main4()
+{
+	// we use the contents of the MAD frame buffer to reconstruct the missing lines from the current
+	// field.
+
+	const int   idx          = int(ZrH.x);                         // buffer index passed from CPU
+	const int   bank         = idx >> 1;                           // current bank
+	const int   field        = idx & 1;                            // current field
+	const int   vpos         = int(gl_FragCoord.y);                // vertical position of destination texture
+	const float sensitivity  = ZrH.w;                              // passed from CPU, higher values mean more likely to use weave
+	const vec3  motion_thr   = vec3(1.0, 1.0, 1.0) * sensitivity;  //
+	const vec2  bofs         = vec2(0.0f, 0.5f);                   // position of the bank 1 relative to source texture size
+	const vec2  vscale       = vec2(1.0f, 0.5f);                   // scaling factor from source to destination texture
+	const vec2  lofs         = vec2(0.0f, ZrH.y) * vscale;         // distance between two adjacent lines relative to source texture size
+	const vec2  iptr         = v_tex * vscale;                     // pointer to the current pixel in the source texture
+
+	vec2 p_t0; // pointer to current pixel (missing or not) from most recent frame
+	vec2 p_t1; // pointer to current pixel (missing or not) from one frame back
+	vec2 p_t2; // pointer to current pixel (missing or not) from two frames back
+	vec2 p_t3; // pointer to current pixel (missing or not) from three frames back
+
+	switch (idx)
+	{
+		case 1:
+			p_t0 = iptr;
+			p_t1 = iptr;
+			p_t2 = iptr + bofs;
+			p_t3 = iptr + bofs;
+			break;
+		case 2:
+			p_t0 = iptr + bofs;
+			p_t1 = iptr;
+			p_t2 = iptr;
+			p_t3 = iptr + bofs;
+			break;
+		case 3:
+			p_t0 = iptr + bofs;
+			p_t1 = iptr + bofs;
+			p_t2 = iptr;
+			p_t3 = iptr;
+			break;
+		default:
+			p_t0 = iptr;
+			p_t1 = iptr + bofs;
+			p_t2 = iptr + bofs;
+			p_t3 = iptr;
+			break;
+	}
+
+	// calculating motion, only relevant for missing lines where the "center line" is pointed by p_t1
+
+	vec4 hn = textureLod(samp0, p_t0 - lofs, 0); // new high pixel
+	vec4 cn = textureLod(samp0, p_t1, 0);        // new center pixel
+	vec4 ln = textureLod(samp0, p_t0 + lofs, 0); // new low pixel
+
+	vec4 ho = textureLod(samp0, p_t2 - lofs, 0); // old high pixel
+	vec4 co = textureLod(samp0, p_t3, 0);        // old center pixel
+	vec4 lo = textureLod(samp0, p_t2 + lofs, 0); // old low pixel
+
+	vec3 mh = hn.rgb - ho.rgb; // high pixel motion
+	vec3 mc = cn.rgb - co.rgb; // center pixel motion
+	vec3 ml = ln.rgb - lo.rgb; // low pixel motion
+
+	mh = max(mh, -mh) - motion_thr;
+	mc = max(mc, -mc) - motion_thr;
+	ml = max(ml, -ml) - motion_thr;
+
+	#if 1 // use this code to evaluate each color motion separately
+		float mh_max = max(max(mh.x, mh.y), mh.z);
+		float mc_max = max(max(mc.x, mc.y), mc.z);
+		float ml_max = max(max(ml.x, ml.y), ml.z);
+	#else // use this code to evaluate average color motion
+		float mh_max = mh.x + mh.y + mh.z;
+		float mc_max = mc.x + mc.y + mc.z;
+		float ml_max = ml.x + ml.y + ml.z;
+	#endif
+
+	// selecting deinterlacing output
+
+	if ((vpos & 1) == field) // output coordinate present on current field
+	{
+		// output coordinate present on current field
+		o_col0 = textureLod(samp0, p_t0, 0);
+	}
+	else if ((iptr.y > 0.5f - lofs.y) || (iptr.y < 0.0 + lofs.y))
+	{
+		// top and bottom lines are always weaved
+		o_col0 = cn;
+	}
+	else
+	{
+		// missing line needs to be reconstructed
+		if(((mh_max > 0.0f) || (ml_max > 0.0f)) || (mc_max > 0.0f))
+			// high motion -> interpolate pixels above and below
+			o_col0 = (hn + ln) / 2.0f;
+		else
+			// low motion -> weave
+			o_col0 = cn;
+	}
+}
+#endif
+
+#endif

bin/resources/shaders/vulkan/merge.glsl

@@ -0,0 +1,46 @@
+// SPDX-FileCopyrightText: 2002-2025 PCSX2 Dev Team
+// SPDX-License-Identifier: GPL-3.0+
+
+#ifdef VERTEX_SHADER
+
+layout(location = 0) in vec4 a_pos;
+layout(location = 1) in vec2 a_tex;
+
+layout(location = 0) out vec2 v_tex;
+
+void main()
+{
+	gl_Position = vec4(a_pos.x, -a_pos.y, a_pos.z, a_pos.w);
+	v_tex = a_tex;
+}
+
+#endif
+
+#ifdef FRAGMENT_SHADER
+
+layout(location = 0) in vec2 v_tex;
+layout(location = 0) out vec4 o_col0;
+
+layout(push_constant) uniform cb10
+{
+		vec4 BGColor;
+};
+
+layout(set = 0, binding = 0) uniform sampler2D samp0;
+
+void ps_main0()
+{
+		vec4 c = texture(samp0, v_tex);
+		// Note: clamping will be done by fixed unit
+		c.a *= 2.0f;
+		o_col0 = c;
+}
+
+void ps_main1()
+{
+		vec4 c = texture(samp0, v_tex);
+		c.a = BGColor.a;
+		o_col0 = c;
+}
+
+#endif

bin/resources/shaders/vulkan/present.glsl

@@ -0,0 +1,446 @@
+// SPDX-FileCopyrightText: 2002-2025 PCSX2 Dev Team
+// SPDX-License-Identifier: GPL-3.0+
+
+#ifdef VERTEX_SHADER
+
+layout(location = 0) in vec4 a_pos;
+layout(location = 1) in vec2 a_tex;
+
+layout(location = 0) out vec2 v_tex;
+
+void main()
+{
+	gl_Position = vec4(a_pos.x, -a_pos.y, a_pos.z, a_pos.w);
+	v_tex = a_tex;
+}
+
+#endif
+
+#ifdef FRAGMENT_SHADER
+
+layout(push_constant) uniform cb10
+{
+	vec4 u_source_rect;
+	vec4 u_target_rect;
+	vec2 u_source_size;
+	vec2 u_target_size;
+	vec2 u_target_resolution;
+	vec2 u_rcp_target_resolution; // 1 / u_target_resolution
+	vec2 u_source_resolution;
+	vec2 u_rcp_source_resolution; // 1 / u_source_resolution
+	float u_time;
+};
+
+layout(location = 0) in vec2 v_tex;
+
+layout(location = 0) out vec4 o_col0;
+
+layout(set = 0, binding = 0) uniform sampler2D samp0;
+
+vec4 sample_c(vec2 uv)
+{
+	return texture(samp0, uv);
+}
+
+vec4 ps_crt(uint i)
+{
+	vec4 mask[4] = vec4[4](
+		vec4(1, 0, 0, 0),
+		vec4(0, 1, 0, 0),
+		vec4(0, 0, 1, 0),
+		vec4(1, 1, 1, 0));
+	return sample_c(v_tex) * clamp((mask[i] + 0.5f), 0.0f, 1.0f);
+}
+
+vec4 ps_scanlines(uint i)
+{
+	vec4 mask[2] =
+		{
+			vec4(1, 1, 1, 0),
+			vec4(0, 0, 0, 0)};
+
+	return sample_c(v_tex) * clamp((mask[i] + 0.5f), 0.0f, 1.0f);
+}
+
+#ifdef ps_copy
+void ps_copy()
+{
+	o_col0 = sample_c(v_tex);
+}
+#endif
+
+#ifdef ps_filter_scanlines
+void ps_filter_scanlines() // scanlines
+{
+	uvec4 p = uvec4(gl_FragCoord);
+
+	o_col0 = ps_scanlines(p.y % 2);
+}
+#endif
+
+#ifdef ps_filter_diagonal
+void ps_filter_diagonal() // diagonal
+{
+	uvec4 p = uvec4(gl_FragCoord);
+	o_col0 = ps_crt((p.x + (p.y % 3)) % 3);
+}
+#endif
+
+#ifdef ps_filter_triangular
+void ps_filter_triangular() // triangular
+{
+	uvec4 p = uvec4(gl_FragCoord);
+
+	// output.c = ps_crt(input, ((p.x + (p.y & 1) * 3) >> 1) % 3);
+	o_col0 = ps_crt(((p.x + ((p.y >> 1) & 1) * 3) >> 1) % 3);
+}
+#endif
+
+#ifdef ps_filter_complex
+void ps_filter_complex() // triangular
+{
+	const float PI = 3.14159265359f;
+	vec2 texdim = vec2(textureSize(samp0, 0));
+
+	o_col0 = (0.9 - 0.4 * cos(2 * PI * v_tex.y * texdim.y)) * sample_c(vec2(v_tex.x, (floor(v_tex.y * texdim.y) + 0.5) / texdim.y));
+}
+#endif
+
+#ifdef ps_filter_lottes
+
+#define MaskingType 4                      //[1|2|3|4] The type of CRT shadow masking used. 1: compressed TV style, 2: Aperture-grille, 3: Stretched VGA style, 4: VGA style.
+#define ScanBrightness -8.00               //[-16.0 to 1.0] The overall brightness of the scanline effect. Lower for darker, higher for brighter.
+#define FilterCRTAmount -3.00              //[-4.0 to 1.0] The amount of filtering used, to replicate the TV CRT look. Lower for less, higher for more.
+#define HorizontalWarp 0.00                //[0.0 to 0.1] The distortion warping effect for the horizontal (x) axis of the screen. Use small increments.
+#define VerticalWarp 0.00                  //[0.0 to 0.1] The distortion warping effect for the verticle (y) axis of the screen. Use small increments.
+#define MaskAmountDark 0.50                //[0.0 to 1.0] The value of the dark masking line effect used. Lower for darker lower end masking, higher for brighter.
+#define MaskAmountLight 1.50               //[0.0 to 2.0] The value of the light masking line effect used. Lower for darker higher end masking, higher for brighter.
+#define BloomPixel -1.50                   //[-2.0 -0.5] Pixel bloom radius. Higher for increased softness of bloom.
+#define BloomScanLine -2.0                 //[-4.0 -1.0] Scanline bloom radius. Higher for increased softness of bloom.
+#define BloomAmount 0.15                   //[0.0 1.0] Bloom intensity. Higher for brighter.
+#define Shape 2.0                          //[0.0 10.0] Kernal filter shape. Lower values will darken image and introduce moire patterns if used with curvature.
+#define UseShadowMask 1                    //[0 or 1] Enables, or disables the use of the CRT shadow mask. 0 is disabled, 1 is enabled.
+
+float ToLinear1(float c)
+{
+	return c <= 0.04045 ? c / 12.92 : pow((c + 0.055) / 1.055, 2.4);
+}
+
+vec3 ToLinear(vec3 c)
+{
+	return vec3(ToLinear1(c.r), ToLinear1(c.g), ToLinear1(c.b));
+}
+
+float ToSrgb1(float c)
+{
+	return c < 0.0031308 ? c * 12.92 : 1.055 * pow(c, 0.41666) - 0.055;
+}
+
+vec3 ToSrgb(vec3 c)
+{
+	return vec3(ToSrgb1(c.r), ToSrgb1(c.g), ToSrgb1(c.b));
+}
+
+vec3 Fetch(vec2 pos, vec2 off)
+{
+	pos = (floor(pos * u_target_size + off) + vec2(0.5, 0.5)) / u_target_size;
+	if (max(abs(pos.x - 0.5), abs(pos.y - 0.5)) > 0.5)
+	{
+		return vec3(0.0, 0.0, 0.0);
+	}
+	else
+	{
+		return ToLinear(texture(samp0, pos.xy).rgb);
+	}
+}
+
+vec2 Dist(vec2 pos)
+{
+	pos = pos * vec2(640, 480);
+
+	return -((pos - floor(pos)) - vec2(0.5, 0.5));
+}
+
+float Gaus(float pos, float scale)
+{
+	return exp2(scale * pow(abs(pos), Shape));
+}
+
+vec3 Horz3(vec2 pos, float off)
+{
+	vec3 b = Fetch(pos, vec2(-1.0, off));
+	vec3 c = Fetch(pos, vec2(0.0, off));
+	vec3 d = Fetch(pos, vec2(1.0, off));
+	float dst = Dist(pos).x;
+
+	// Convert distance to weight.
+	float scale = FilterCRTAmount;
+	float wb = Gaus(dst - 1.0, scale);
+	float wc = Gaus(dst + 0.0, scale);
+	float wd = Gaus(dst + 1.0, scale);
+
+	return (b * wb + c * wc + d * wd) / (wb + wc + wd);
+}
+
+vec3 Horz5(vec2 pos, float off)
+{
+	vec3 a = Fetch(pos, vec2(-2.0, off));
+	vec3 b = Fetch(pos, vec2(-1.0, off));
+	vec3 c = Fetch(pos, vec2(0.0, off));
+	vec3 d = Fetch(pos, vec2(1.0, off));
+	vec3 e = Fetch(pos, vec2(2.0, off));
+	float dst = Dist(pos).x;
+
+	// Convert distance to weight.
+	float scale = FilterCRTAmount;
+
+	float wa = Gaus(dst - 2.0, scale);
+	float wb = Gaus(dst - 1.0, scale);
+	float wc = Gaus(dst + 0.0, scale);
+	float wd = Gaus(dst + 1.0, scale);
+	float we = Gaus(dst + 2.0, scale);
+
+	return (a * wa + b * wb + c * wc + d * wd + e * we) / (wa + wb + wc + wd + we);
+}
+
+vec3 Horz7(vec2 pos, float off)
+{
+	vec3 a = Fetch(pos, vec2(-3.0, off));
+	vec3 b = Fetch(pos, vec2(-2.0, off));
+	vec3 c = Fetch(pos, vec2(-1.0, off));
+	vec3 d = Fetch(pos, vec2( 0.0, off));
+	vec3 e = Fetch(pos, vec2( 1.0, off));
+	vec3 f = Fetch(pos, vec2( 2.0, off));
+	vec3 g = Fetch(pos, vec2( 3.0, off));
+
+	float dst = Dist(pos).x;
+	// Convert distance to weight.
+	float scale = BloomPixel;
+	float wa = Gaus(dst - 3.0, scale);
+	float wb = Gaus(dst - 2.0, scale);
+	float wc = Gaus(dst - 1.0, scale);
+	float wd = Gaus(dst + 0.0, scale);
+	float we = Gaus(dst + 1.0, scale);
+	float wf = Gaus(dst + 2.0, scale);
+	float wg = Gaus(dst + 3.0, scale);
+
+	// Return filtered sample.
+	return (a * wa + b * wb + c * wc + d * wd + e * we + f * wf + g * wg) / (wa + wb + wc + wd + we + wf + wg);
+}
+
+// Return scanline weight.
+float Scan(vec2 pos, float off)
+{
+	float dst = Dist(pos).y;
+	return Gaus(dst + off, ScanBrightness);
+}
+
+float BloomScan(vec2 pos, float off)
+{
+	float dst = Dist(pos).y;
+
+	return Gaus(dst + off, BloomScanLine);
+}
+
+vec3 Tri(vec2 pos)
+{
+	vec3 a = Horz3(pos, -1.0);
+	vec3 b = Horz5(pos, 0.0);
+	vec3 c = Horz3(pos, 1.0);
+
+	float wa = Scan(pos, -1.0);
+	float wb = Scan(pos, 0.0);
+	float wc = Scan(pos, 1.0);
+
+	return (a * wa) + (b * wb) + (c * wc);
+}
+
+vec3 Bloom(vec2 pos)
+{
+	vec3 a = Horz5(pos,-2.0);
+	vec3 b = Horz7(pos,-1.0);
+	vec3 c = Horz7(pos, 0.0);
+	vec3 d = Horz7(pos, 1.0);
+	vec3 e = Horz5(pos, 2.0);
+
+	float wa = BloomScan(pos,-2.0);
+	float wb = BloomScan(pos,-1.0); 
+	float wc = BloomScan(pos, 0.0);
+	float wd = BloomScan(pos, 1.0);
+	float we = BloomScan(pos, 2.0);
+
+	return a * wa + b * wb + c * wc + d * wd + e * we;
+}
+
+vec2 Warp(vec2 pos)
+{
+	pos = pos * 2.0 - 1.0;
+	pos *= vec2(1.0 + (pos.y * pos.y) * HorizontalWarp, 1.0 + (pos.x * pos.x) * VerticalWarp);
+	return pos * 0.5 + 0.5;
+}
+
+vec3 Mask(vec2 pos)
+{
+#if MaskingType == 1
+	// Very compressed TV style shadow mask.
+	float lines = MaskAmountLight;
+	float odd = 0.0;
+
+	if (fract(pos.x / 6.0) < 0.5)
+	{
+		odd = 1.0;
+	}
+	if (fract((pos.y + odd) / 2.0) < 0.5)
+	{
+		lines = MaskAmountDark;
+	}
+	pos.x = fract(pos.x / 3.0);
+	vec3 mask = vec3(MaskAmountDark, MaskAmountDark, MaskAmountDark);
+
+	if (pos.x < 0.333)
+	{
+		mask.r = MaskAmountLight;
+	}
+	else if (pos.x < 0.666)
+	{
+		mask.g = MaskAmountLight;
+	}
+	else
+	{
+		mask.b = MaskAmountLight;
+	}
+
+	mask *= lines;
+
+	return mask;
+
+#elif MaskingType == 2
+	// Aperture-grille.
+	pos.x = fract(pos.x / 3.0);
+	vec3 mask = vec3(MaskAmountDark, MaskAmountDark, MaskAmountDark);
+
+	if (pos.x < 0.333)
+	{
+		mask.r = MaskAmountLight;
+	}
+	else if (pos.x < 0.666)
+	{
+		mask.g = MaskAmountLight;
+	}
+	else
+	{
+		mask.b = MaskAmountLight;
+	}
+
+	return mask;
+
+#elif MaskingType == 3
+	// Stretched VGA style shadow mask (same as prior shaders).
+	pos.x += pos.y * 3.0;
+	vec3 mask = vec3(MaskAmountDark, MaskAmountDark, MaskAmountDark);
+	pos.x = fract(pos.x / 6.0);
+
+	if (pos.x < 0.333)
+	{
+		mask.r = MaskAmountLight;
+	}
+	else if (pos.x < 0.666)
+	{
+		mask.g = MaskAmountLight;
+	}
+	else
+	{
+		mask.b = MaskAmountLight;
+	}
+
+	return mask;
+
+#else
+	// VGA style shadow mask.
+	pos.xy = floor(pos.xy * vec2(1.0, 0.5));
+	pos.x += pos.y * 3.0;
+
+	vec3 mask = vec3(MaskAmountDark, MaskAmountDark, MaskAmountDark);
+	pos.x = fract(pos.x / 6.0);
+
+	if (pos.x < 0.333)
+	{
+		mask.r = MaskAmountLight;
+	}
+	else if (pos.x < 0.666)
+	{
+		mask.g = MaskAmountLight;
+	}
+	else
+	{
+		mask.b = MaskAmountLight;
+	}
+	return mask;
+#endif
+}
+
+vec4 LottesCRTPass()
+{
+	vec4 color;
+	vec4 fragcoord = gl_FragCoord - u_target_rect;
+	vec2 inSize = u_target_resolution - (2 * u_target_rect.xy);
+
+	vec2 pos = Warp(fragcoord.xy / inSize);
+	color.rgb = Tri(pos);
+	color.rgb += Bloom(pos) * BloomAmount;
+#if UseShadowMask
+	color.rgb *= Mask(fragcoord.xy);
+#endif
+	color.rgb = ToSrgb(color.rgb);
+
+	return color;
+}
+
+void ps_filter_lottes()
+{
+	o_col0 = LottesCRTPass();
+}
+
+#endif
+
+#ifdef ps_4x_rgss
+void ps_4x_rgss()
+{
+	vec2 dxy = vec2(dFdx(v_tex.x), dFdy(v_tex.y));
+	vec3 color = vec3(0);
+
+	float s = 1.0/8.0;
+	float l = 3.0/8.0;
+
+	color += sample_c(v_tex + vec2( s, l) * dxy).rgb;
+	color += sample_c(v_tex + vec2( l,-s) * dxy).rgb;
+	color += sample_c(v_tex + vec2(-s,-l) * dxy).rgb;
+	color += sample_c(v_tex + vec2(-l, s) * dxy).rgb;
+
+	o_col0 = vec4(color * 0.25,1);
+}
+#endif
+
+#ifdef ps_automagical_supersampling
+void ps_automagical_supersampling()
+{
+	vec2 ratio = (u_source_size / u_target_size) * 0.5;
+	vec2 steps = floor(ratio);
+	vec3 col = sample_c(v_tex).rgb;
+	float div = 1;
+
+	for (float y = 0; y < steps.y; y++)
+	{
+		for (float x = 0; x < steps.x; x++)
+		{
+			vec2 offset = vec2(x,y) - ratio * 0.5;
+			col += sample_c(v_tex + offset * u_rcp_source_resolution * 2.0).rgb;
+			div++;
+		}
+	}
+
+	o_col0 = vec4(col / div, 1);
+}
+#endif
+
+#endif

bin/resources/shaders/vulkan/shadeboost.glsl

@@ -0,0 +1,74 @@
+// SPDX-FileCopyrightText: 2002-2025 PCSX2 Dev Team
+// SPDX-License-Identifier: GPL-3.0+
+
+//#version 420 // Keep it for editor detection
+
+#ifdef VERTEX_SHADER
+
+layout(location = 0) in vec4 a_pos;
+layout(location = 1) in vec2 a_tex;
+
+layout(location = 0) out vec2 v_tex;
+
+void main()
+{
+	gl_Position = vec4(a_pos.x, -a_pos.y, a_pos.z, a_pos.w);
+	v_tex = a_tex;
+}
+
+#endif
+
+/*
+** Contrast, saturation, brightness
+** Code of this function is from TGM's shader pack
+** http://irrlicht.sourceforge.net/phpBB2/viewtopic.php?t=21057
+** TGM's author comment about the license (included in the previous link)
+** "do with it, what you want! its total free!
+** (but would be nice, if you say that you used my shaders  :wink: ) but not necessary"
+*/
+
+#ifdef FRAGMENT_SHADER
+
+layout(push_constant) uniform cb0
+{
+	vec4 params;
+};
+
+layout(set = 0, binding = 0) uniform sampler2D samp0;
+layout(location = 0) in vec2 v_tex;
+layout(location = 0) out vec4 o_col0;
+
+// For all settings: 1.0 = 100% 0.5=50% 1.5 = 150%
+vec4 ContrastSaturationBrightness(vec4 color)
+{
+    float brt = params.x;
+    float con = params.y;
+    float sat = params.z;
+
+    // Increase or decrease these values to adjust r, g and b color channels separately
+    const float AvgLumR = 0.5;
+    const float AvgLumG = 0.5;
+    const float AvgLumB = 0.5;
+
+    const vec3 LumCoeff = vec3(0.2125, 0.7154, 0.0721);
+
+    vec3 AvgLumin = vec3(AvgLumR, AvgLumG, AvgLumB);
+    vec3 brtColor = color.rgb * brt;
+    float dot_intensity = dot(brtColor, LumCoeff);
+    vec3 intensity = vec3(dot_intensity, dot_intensity, dot_intensity);
+    vec3 satColor = mix(intensity, brtColor, sat);
+    vec3 conColor = mix(AvgLumin, satColor, con);
+
+    color.rgb = conColor;
+    return color;
+}
+
+
+void main()
+{
+    vec4 c = texture(samp0, v_tex);
+    o_col0 = ContrastSaturationBrightness(c);
+}
+
+
+#endif