diff --git a/src/video_core/host_shaders/CMakeLists.txt b/src/video_core/host_shaders/CMakeLists.txt index a5049c2a7..40676999a 100644 --- a/src/video_core/host_shaders/CMakeLists.txt +++ b/src/video_core/host_shaders/CMakeLists.txt @@ -13,23 +13,23 @@ set(SHADER_FILES texture_filtering/mmpx.frag texture_filtering/x_gradient.frag texture_filtering/y_gradient.frag - antialiasing/opengl_fxaa.frag - antialiasing/opengl_fxaa.vert - antialiasing/opengl_smaa_pass0_pre.frag - antialiasing/opengl_smaa_pass0_pre.vert - antialiasing/opengl_smaa_pass0_post.frag - antialiasing/opengl_smaa_pass0_post.vert - antialiasing/opengl_smaa_pass1_pre.frag - antialiasing/opengl_smaa_pass1_pre.vert - antialiasing/opengl_smaa_pass1_post.frag - antialiasing/opengl_smaa_pass1_post.vert - antialiasing/opengl_smaa_pass2_pre.frag - antialiasing/opengl_smaa_pass2_pre.vert - antialiasing/opengl_smaa_pass2_post.frag - antialiasing/opengl_smaa_pass2_post.vert + antialiasing/OpenGL/opengl_fxaa.frag + antialiasing/OpenGL/opengl_fxaa.vert + antialiasing/OpenGL/opengl_smaa_pass0_pre.frag + antialiasing/OpenGL/opengl_smaa_pass0_pre.vert + antialiasing/OpenGL/opengl_smaa_pass0_post.frag + antialiasing/OpenGL/opengl_smaa_pass0_post.vert + antialiasing/OpenGL/opengl_smaa_pass1_pre.frag + antialiasing/OpenGL/opengl_smaa_pass1_pre.vert + antialiasing/OpenGL/opengl_smaa_pass1_post.frag + antialiasing/OpenGL/opengl_smaa_pass1_post.vert + antialiasing/OpenGL/opengl_smaa_pass2_pre.frag + antialiasing/OpenGL/opengl_smaa_pass2_pre.vert + antialiasing/OpenGL/opengl_smaa_pass2_post.frag + antialiasing/OpenGL/opengl_smaa_pass2_post.vert + antialiasing/OpenGL/opengl_smaa.hlsl scaling/opengl_area_sampling.frag scaling/opengl_area_sampling.vert - antialiasing/smaa.hlsl full_screen_triangle.vert opengl_present.frag opengl_present.vert diff --git a/src/video_core/host_shaders/antialiasing/opengl_fxaa.frag b/src/video_core/host_shaders/antialiasing/OpenGL/opengl_fxaa.frag similarity index 100% rename from src/video_core/host_shaders/antialiasing/opengl_fxaa.frag rename to src/video_core/host_shaders/antialiasing/OpenGL/opengl_fxaa.frag diff --git a/src/video_core/host_shaders/antialiasing/opengl_fxaa.vert b/src/video_core/host_shaders/antialiasing/OpenGL/opengl_fxaa.vert similarity index 100% rename from src/video_core/host_shaders/antialiasing/opengl_fxaa.vert rename to src/video_core/host_shaders/antialiasing/OpenGL/opengl_fxaa.vert diff --git a/src/video_core/host_shaders/antialiasing/smaa.hlsl b/src/video_core/host_shaders/antialiasing/OpenGL/opengl_smaa.hlsl similarity index 100% rename from src/video_core/host_shaders/antialiasing/smaa.hlsl rename to src/video_core/host_shaders/antialiasing/OpenGL/opengl_smaa.hlsl diff --git a/src/video_core/host_shaders/antialiasing/opengl_smaa_pass0_post.frag b/src/video_core/host_shaders/antialiasing/OpenGL/opengl_smaa_pass0_post.frag similarity index 100% rename from src/video_core/host_shaders/antialiasing/opengl_smaa_pass0_post.frag rename to src/video_core/host_shaders/antialiasing/OpenGL/opengl_smaa_pass0_post.frag diff --git a/src/video_core/host_shaders/antialiasing/opengl_smaa_pass0_post.vert b/src/video_core/host_shaders/antialiasing/OpenGL/opengl_smaa_pass0_post.vert similarity index 100% rename from src/video_core/host_shaders/antialiasing/opengl_smaa_pass0_post.vert rename to src/video_core/host_shaders/antialiasing/OpenGL/opengl_smaa_pass0_post.vert diff --git a/src/video_core/host_shaders/antialiasing/opengl_smaa_pass0_pre.frag b/src/video_core/host_shaders/antialiasing/OpenGL/opengl_smaa_pass0_pre.frag similarity index 100% rename from src/video_core/host_shaders/antialiasing/opengl_smaa_pass0_pre.frag rename to src/video_core/host_shaders/antialiasing/OpenGL/opengl_smaa_pass0_pre.frag diff --git a/src/video_core/host_shaders/antialiasing/opengl_smaa_pass0_pre.vert b/src/video_core/host_shaders/antialiasing/OpenGL/opengl_smaa_pass0_pre.vert similarity index 100% rename from src/video_core/host_shaders/antialiasing/opengl_smaa_pass0_pre.vert rename to src/video_core/host_shaders/antialiasing/OpenGL/opengl_smaa_pass0_pre.vert diff --git a/src/video_core/host_shaders/antialiasing/opengl_smaa_pass1_post.frag b/src/video_core/host_shaders/antialiasing/OpenGL/opengl_smaa_pass1_post.frag similarity index 100% rename from src/video_core/host_shaders/antialiasing/opengl_smaa_pass1_post.frag rename to src/video_core/host_shaders/antialiasing/OpenGL/opengl_smaa_pass1_post.frag diff --git a/src/video_core/host_shaders/antialiasing/opengl_smaa_pass1_post.vert b/src/video_core/host_shaders/antialiasing/OpenGL/opengl_smaa_pass1_post.vert similarity index 100% rename from src/video_core/host_shaders/antialiasing/opengl_smaa_pass1_post.vert rename to src/video_core/host_shaders/antialiasing/OpenGL/opengl_smaa_pass1_post.vert diff --git a/src/video_core/host_shaders/antialiasing/opengl_smaa_pass1_pre.frag b/src/video_core/host_shaders/antialiasing/OpenGL/opengl_smaa_pass1_pre.frag similarity index 100% rename from src/video_core/host_shaders/antialiasing/opengl_smaa_pass1_pre.frag rename to src/video_core/host_shaders/antialiasing/OpenGL/opengl_smaa_pass1_pre.frag diff --git a/src/video_core/host_shaders/antialiasing/opengl_smaa_pass1_pre.vert b/src/video_core/host_shaders/antialiasing/OpenGL/opengl_smaa_pass1_pre.vert similarity index 100% rename from src/video_core/host_shaders/antialiasing/opengl_smaa_pass1_pre.vert rename to src/video_core/host_shaders/antialiasing/OpenGL/opengl_smaa_pass1_pre.vert diff --git a/src/video_core/host_shaders/antialiasing/opengl_smaa_pass2_post.frag b/src/video_core/host_shaders/antialiasing/OpenGL/opengl_smaa_pass2_post.frag similarity index 100% rename from src/video_core/host_shaders/antialiasing/opengl_smaa_pass2_post.frag rename to src/video_core/host_shaders/antialiasing/OpenGL/opengl_smaa_pass2_post.frag diff --git a/src/video_core/host_shaders/antialiasing/opengl_smaa_pass2_post.vert b/src/video_core/host_shaders/antialiasing/OpenGL/opengl_smaa_pass2_post.vert similarity index 100% rename from src/video_core/host_shaders/antialiasing/opengl_smaa_pass2_post.vert rename to src/video_core/host_shaders/antialiasing/OpenGL/opengl_smaa_pass2_post.vert diff --git a/src/video_core/host_shaders/antialiasing/opengl_smaa_pass2_pre.frag b/src/video_core/host_shaders/antialiasing/OpenGL/opengl_smaa_pass2_pre.frag similarity index 100% rename from src/video_core/host_shaders/antialiasing/opengl_smaa_pass2_pre.frag rename to src/video_core/host_shaders/antialiasing/OpenGL/opengl_smaa_pass2_pre.frag diff --git a/src/video_core/host_shaders/antialiasing/opengl_smaa_pass2_pre.vert b/src/video_core/host_shaders/antialiasing/OpenGL/opengl_smaa_pass2_pre.vert similarity index 100% rename from src/video_core/host_shaders/antialiasing/opengl_smaa_pass2_pre.vert rename to src/video_core/host_shaders/antialiasing/OpenGL/opengl_smaa_pass2_pre.vert diff --git a/src/video_core/host_shaders/antialiasing/Vulkan/vulkan_fxaa.frag b/src/video_core/host_shaders/antialiasing/Vulkan/vulkan_fxaa.frag new file mode 100644 index 000000000..ac3d736cd --- /dev/null +++ b/src/video_core/host_shaders/antialiasing/Vulkan/vulkan_fxaa.frag @@ -0,0 +1,256 @@ +/** + * @license + * Copyright (c) 2011 NVIDIA Corporation. All rights reserved. + * + * TO THE MAXIMUM EXTENT PERMITTED BY APPLICABLE LAW, THIS SOFTWARE IS PROVIDED + * *AS IS* AND NVIDIA AND ITS SUPPLIERS DISCLAIM ALL WARRANTIES, EITHER EXPRESS + * OR IMPLIED, INCLUDING, BUT NOT LIMITED TO, NONINFRINGEMENT,IMPLIED WARRANTIES OF + * MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE. IN NO EVENT SHALL NVIDIA + * OR ITS SUPPLIERS BE LIABLE FOR ANY DIRECT, SPECIAL, INCIDENTAL, INDIRECT, OR + * CONSEQUENTIAL DAMAGES WHATSOEVER (INCLUDING, WITHOUT LIMITATION, DAMAGES FOR LOSS + * OF BUSINESS PROFITS, BUSINESS INTERRUPTION, LOSS OF BUSINESS INFORMATION, OR ANY + * OTHER PECUNIARY LOSS) ARISING OUT OF THE USE OF OR INABILITY TO USE THIS SOFTWARE, + * EVEN IF NVIDIA HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH DAMAGES. + */ + +/* +FXAA_PRESET - Choose compile-in knob preset 0-5. +------------------------------------------------------------------------------ +FXAA_EDGE_THRESHOLD - The minimum amount of local contrast required + to apply algorithm. + 1.0/3.0 - too little + 1.0/4.0 - good start + 1.0/8.0 - applies to more edges + 1.0/16.0 - overkill +------------------------------------------------------------------------------ +FXAA_EDGE_THRESHOLD_MIN - Trims the algorithm from processing darks. + Perf optimization. + 1.0/32.0 - visible limit (smaller isn't visible) + 1.0/16.0 - good compromise + 1.0/12.0 - upper limit (seeing artifacts) +------------------------------------------------------------------------------ +FXAA_SEARCH_STEPS - Maximum number of search steps for end of span. +------------------------------------------------------------------------------ +FXAA_SEARCH_THRESHOLD - Controls when to stop searching. + 1.0/4.0 - seems to be the best quality wise +------------------------------------------------------------------------------ +FXAA_SUBPIX_TRIM - Controls sub-pixel aliasing removal. + 1.0/2.0 - low removal + 1.0/3.0 - medium removal + 1.0/4.0 - default removal + 1.0/8.0 - high removal + 0.0 - complete removal +------------------------------------------------------------------------------ +FXAA_SUBPIX_CAP - Insures fine detail is not completely removed. + This is important for the transition of sub-pixel detail, + like fences and wires. + 3.0/4.0 - default (medium amount of filtering) + 7.0/8.0 - high amount of filtering + 1.0 - no capping of sub-pixel aliasing removal +*/ +//? #version 450 + +layout(location = 0) in vec2 frag_tex_coord; +layout(location = 0) out vec4 color; +layout(binding = 0) uniform sampler2D color_texture; + +uniform vec4 i_resolution; +uniform int convert_colors; + +#ifndef FXAA_PRESET + #define FXAA_PRESET 5 +#endif +#if (FXAA_PRESET == 3) + #define FXAA_EDGE_THRESHOLD (1.0/8.0) + #define FXAA_EDGE_THRESHOLD_MIN (1.0/16.0) + #define FXAA_SEARCH_STEPS 16 + #define FXAA_SEARCH_THRESHOLD (1.0/4.0) + #define FXAA_SUBPIX_CAP (3.0/4.0) + #define FXAA_SUBPIX_TRIM (1.0/4.0) +#endif +#if (FXAA_PRESET == 4) + #define FXAA_EDGE_THRESHOLD (1.0/8.0) + #define FXAA_EDGE_THRESHOLD_MIN (1.0/24.0) + #define FXAA_SEARCH_STEPS 24 + #define FXAA_SEARCH_THRESHOLD (1.0/4.0) + #define FXAA_SUBPIX_CAP (3.0/4.0) + #define FXAA_SUBPIX_TRIM (1.0/4.0) +#endif +#if (FXAA_PRESET == 5) + #define FXAA_EDGE_THRESHOLD (1.0/8.0) + #define FXAA_EDGE_THRESHOLD_MIN (1.0/24.0) + #define FXAA_SEARCH_STEPS 32 + #define FXAA_SEARCH_THRESHOLD (1.0/4.0) + #define FXAA_SUBPIX_CAP (3.0/4.0) + #define FXAA_SUBPIX_TRIM (1.0/4.0) +#endif + +#define FXAA_SUBPIX_TRIM_SCALE (1.0/(1.0 - FXAA_SUBPIX_TRIM)) + +// Return the luma, the estimation of luminance from rgb inputs. +// This approximates luma using one FMA instruction, +// skipping normalization and tossing out blue. +// FxaaLuma() will range 0.0 to 2.963210702. +float FxaaLuma(vec3 rgb) { + return rgb.y * (0.587/0.299) + rgb.x; +} + +vec3 FxaaLerp3(vec3 a, vec3 b, float amountOfA) { + return (vec3(-amountOfA) * b) + ((a * vec3(amountOfA)) + b); +} + +vec4 FxaaTexOff(sampler2D tex, vec2 pos, ivec2 off, vec2 rcpFrame) { + float x = pos.x + float(off.x) * rcpFrame.x; + float y = pos.y + float(off.y) * rcpFrame.y; + return texture(tex, vec2(x, y)); +} + +// pos is the output of FxaaVertexShader interpolated across screen. +// xy -> actual texture position {0.0 to 1.0} +// rcpFrame should be a uniform equal to {1.0/frameWidth, 1.0/frameHeight} +vec3 FxaaPixelShader(vec2 pos, sampler2D tex, vec2 rcpFrame) +{ + vec3 rgbN = FxaaTexOff(tex, pos.xy, ivec2( 0,-1), rcpFrame).xyz; + vec3 rgbW = FxaaTexOff(tex, pos.xy, ivec2(-1, 0), rcpFrame).xyz; + vec3 rgbM = FxaaTexOff(tex, pos.xy, ivec2( 0, 0), rcpFrame).xyz; + vec3 rgbE = FxaaTexOff(tex, pos.xy, ivec2( 1, 0), rcpFrame).xyz; + vec3 rgbS = FxaaTexOff(tex, pos.xy, ivec2( 0, 1), rcpFrame).xyz; + + float lumaN = FxaaLuma(rgbN); + float lumaW = FxaaLuma(rgbW); + float lumaM = FxaaLuma(rgbM); + float lumaE = FxaaLuma(rgbE); + float lumaS = FxaaLuma(rgbS); + float rangeMin = min(lumaM, min(min(lumaN, lumaW), min(lumaS, lumaE))); + float rangeMax = max(lumaM, max(max(lumaN, lumaW), max(lumaS, lumaE))); + + float range = rangeMax - rangeMin; + if(range < max(FXAA_EDGE_THRESHOLD_MIN, rangeMax * FXAA_EDGE_THRESHOLD)) + { + return rgbM; + } + + vec3 rgbL = rgbN + rgbW + rgbM + rgbE + rgbS; + + float lumaL = (lumaN + lumaW + lumaE + lumaS) * 0.25; + float rangeL = abs(lumaL - lumaM); + float blendL = max(0.0, (rangeL / range) - FXAA_SUBPIX_TRIM) * FXAA_SUBPIX_TRIM_SCALE; + blendL = min(FXAA_SUBPIX_CAP, blendL); + + vec3 rgbNW = FxaaTexOff(tex, pos.xy, ivec2(-1,-1), rcpFrame).xyz; + vec3 rgbNE = FxaaTexOff(tex, pos.xy, ivec2( 1,-1), rcpFrame).xyz; + vec3 rgbSW = FxaaTexOff(tex, pos.xy, ivec2(-1, 1), rcpFrame).xyz; + vec3 rgbSE = FxaaTexOff(tex, pos.xy, ivec2( 1, 1), rcpFrame).xyz; + rgbL += (rgbNW + rgbNE + rgbSW + rgbSE); + rgbL *= vec3(1.0/9.0); + + float lumaNW = FxaaLuma(rgbNW); + float lumaNE = FxaaLuma(rgbNE); + float lumaSW = FxaaLuma(rgbSW); + float lumaSE = FxaaLuma(rgbSE); + + float edgeVert = + abs((0.25 * lumaNW) + (-0.5 * lumaN) + (0.25 * lumaNE)) + + abs((0.50 * lumaW ) + (-1.0 * lumaM) + (0.50 * lumaE )) + + abs((0.25 * lumaSW) + (-0.5 * lumaS) + (0.25 * lumaSE)); + float edgeHorz = + abs((0.25 * lumaNW) + (-0.5 * lumaW) + (0.25 * lumaSW)) + + abs((0.50 * lumaN ) + (-1.0 * lumaM) + (0.50 * lumaS )) + + abs((0.25 * lumaNE) + (-0.5 * lumaE) + (0.25 * lumaSE)); + + bool horzSpan = edgeHorz >= edgeVert; + float lengthSign = horzSpan ? -rcpFrame.y : -rcpFrame.x; + + if(!horzSpan) + { + lumaN = lumaW; + lumaS = lumaE; + } + + float gradientN = abs(lumaN - lumaM); + float gradientS = abs(lumaS - lumaM); + lumaN = (lumaN + lumaM) * 0.5; + lumaS = (lumaS + lumaM) * 0.5; + + if (gradientN < gradientS) + { + lumaN = lumaS; + lumaN = lumaS; + gradientN = gradientS; + lengthSign *= -1.0; + } + + vec2 posN; + posN.x = pos.x + (horzSpan ? 0.0 : lengthSign * 0.5); + posN.y = pos.y + (horzSpan ? lengthSign * 0.5 : 0.0); + + gradientN *= FXAA_SEARCH_THRESHOLD; + + vec2 posP = posN; + vec2 offNP = horzSpan ? vec2(rcpFrame.x, 0.0) : vec2(0.0, rcpFrame.y); + float lumaEndN = lumaN; + float lumaEndP = lumaN; + bool doneN = false; + bool doneP = false; + posN += offNP * vec2(-1.0, -1.0); + posP += offNP * vec2( 1.0, 1.0); + + for(int i = 0; i < FXAA_SEARCH_STEPS; i++) { + if(!doneN) + { + lumaEndN = FxaaLuma(texture(tex, posN.xy).xyz); + } + if(!doneP) + { + lumaEndP = FxaaLuma(texture(tex, posP.xy).xyz); + } + + doneN = doneN || (abs(lumaEndN - lumaN) >= gradientN); + doneP = doneP || (abs(lumaEndP - lumaN) >= gradientN); + + if(doneN && doneP) + { + break; + } + if(!doneN) + { + posN -= offNP; + } + if(!doneP) + { + posP += offNP; + } + } + + float dstN = horzSpan ? pos.x - posN.x : pos.y - posN.y; + float dstP = horzSpan ? posP.x - pos.x : posP.y - pos.y; + bool directionN = dstN < dstP; + lumaEndN = directionN ? lumaEndN : lumaEndP; + + if(((lumaM - lumaN) < 0.0) == ((lumaEndN - lumaN) < 0.0)) + { + lengthSign = 0.0; + } + + + float spanLength = (dstP + dstN); + dstN = directionN ? dstN : dstP; + float subPixelOffset = (0.5 + (dstN * (-1.0/spanLength))) * lengthSign; + vec3 rgbF = texture(tex, vec2( + pos.x + (horzSpan ? 0.0 : subPixelOffset), + pos.y + (horzSpan ? subPixelOffset : 0.0))).xyz; + return FxaaLerp3(rgbL, rgbF, blendL); +} + +vec3 sRGBToLinear(vec3 c) { + return mix(c / 12.92, pow((c + 0.055) / 1.055, vec3(2.4)), step(0.04045, c)); +} + +void main() +{ + vec4 pixel = vec4(FxaaPixelShader(frag_tex_coord, color_texture, vec2(i_resolution.z, i_resolution.w)), 1.0) * 1.0; + if (convert_colors == 1){ + pixel = vec4(sRGBToLinear(pixel.rgb), pixel.a); + } + color = pixel; +} diff --git a/src/video_core/host_shaders/antialiasing/Vulkan/vulkan_fxaa.vert b/src/video_core/host_shaders/antialiasing/Vulkan/vulkan_fxaa.vert new file mode 100644 index 000000000..793dcc580 --- /dev/null +++ b/src/video_core/host_shaders/antialiasing/Vulkan/vulkan_fxaa.vert @@ -0,0 +1,13 @@ +//? #version 450 +layout(location = 0) in vec2 vert_position; +layout(location = 1) in vec2 vert_tex_coord; +layout(location = 0) out vec2 frag_tex_coord; + +uniform vec4 i_resolution; + +void main() +{ + gl_Position = vec4(vert_position, 0.0, 1.0); + frag_tex_coord = vert_tex_coord; +} + diff --git a/src/video_core/host_shaders/antialiasing/Vulkan/vulkan_smaa.hlsl b/src/video_core/host_shaders/antialiasing/Vulkan/vulkan_smaa.hlsl new file mode 100644 index 000000000..05e2b5eed --- /dev/null +++ b/src/video_core/host_shaders/antialiasing/Vulkan/vulkan_smaa.hlsl @@ -0,0 +1,1361 @@ +/** + * Copyright (C) 2013 Jorge Jimenez (jorge@iryoku.com) + * Copyright (C) 2013 Jose I. Echevarria (joseignacioechevarria@gmail.com) + * Copyright (C) 2013 Belen Masia (bmasia@unizar.es) + * Copyright (C) 2013 Fernando Navarro (fernandn@microsoft.com) + * Copyright (C) 2013 Diego Gutierrez (diegog@unizar.es) + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * 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. As clarification, there + * is no requirement that the copyright notice and permission be included in + * binary distributions 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. + */ + + +/** + * _______ ___ ___ ___ ___ + * / || \/ | / \ / \ + * | (---- | \ / | / ^ \ / ^ \ + * \ \ | |\/| | / /_\ \ / /_\ \ + * ----) | | | | | / _____ \ / _____ \ + * |_______/ |__| |__| /__/ \__\ /__/ \__\ + * + * E N H A N C E D + * S U B P I X E L M O R P H O L O G I C A L A N T I A L I A S I N G + * + * http://www.iryoku.com/smaa/ + * + * Hi, welcome aboard! + * + * Here you'll find instructions to get the shader up and running as fast as + * possible. + * + * IMPORTANTE NOTICE: when updating, remember to update both this file and the + * precomputed textures! They may change from version to version. + * + * The shader has three passes, chained together as follows: + * + * |input|------------------· + * v | + * [ SMAA*EdgeDetection ] | + * v | + * |edgesTex| | + * v | + * [ SMAABlendingWeightCalculation ] | + * v | + * |blendTex| | + * v | + * [ SMAANeighborhoodBlending ] <------· + * v + * |output| + * + * Note that each [pass] has its own vertex and pixel shader. Remember to use + * oversized triangles instead of quads to avoid overshading along the + * diagonal. + * + * You've three edge detection methods to choose from: luma, color or depth. + * They represent different quality/performance and anti-aliasing/sharpness + * tradeoffs, so our recommendation is for you to choose the one that best + * suits your particular scenario: + * + * - Depth edge detection is usually the fastest but it may miss some edges. + * + * - Luma edge detection is usually more expensive than depth edge detection, + * but catches visible edges that depth edge detection can miss. + * + * - Color edge detection is usually the most expensive one but catches + * chroma-only edges. + * + * For quickstarters: just use luma edge detection. + * + * The general advice is to not rush the integration process and ensure each + * step is done correctly (don't try to integrate SMAA T2x with predicated edge + * detection from the start!). Ok then, let's go! + * + * 1. The first step is to create two RGBA temporal render targets for holding + * |edgesTex| and |blendTex|. + * + * In DX10 or DX11, you can use a RG render target for the edges texture. + * In the case of NVIDIA GPUs, using RG render targets seems to actually be + * slower. + * + * On the Xbox 360, you can use the same render target for resolving both + * |edgesTex| and |blendTex|, as they aren't needed simultaneously. + * + * 2. Both temporal render targets |edgesTex| and |blendTex| must be cleared + * each frame. Do not forget to clear the alpha channel! + * + * 3. The next step is loading the two supporting precalculated textures, + * 'areaTex' and 'searchTex'. You'll find them in the 'Textures' folder as + * C++ headers, and also as regular DDS files. They'll be needed for the + * 'SMAABlendingWeightCalculation' pass. + * + * If you use the C++ headers, be sure to load them in the format specified + * inside of them. + * + * You can also compress 'areaTex' and 'searchTex' using BC5 and BC4 + * respectively, if you have that option in your content processor pipeline. + * When compressing then, you get a non-perceptible quality decrease, and a + * marginal performance increase. + * + * 4. All samplers must be set to linear filtering and clamp. + * + * After you get the technique working, remember that 64-bit inputs have + * half-rate linear filtering on GCN. + * + * If SMAA is applied to 64-bit color buffers, switching to point filtering + * when accesing them will increase the performance. Search for + * 'SMAASamplePoint' to see which textures may benefit from point + * filtering, and where (which is basically the color input in the edge + * detection and resolve passes). + * + * 5. All texture reads and buffer writes must be non-sRGB, with the exception + * of the input read and the output write in + * 'SMAANeighborhoodBlending' (and only in this pass!). If sRGB reads in + * this last pass are not possible, the technique will work anyway, but + * will perform antialiasing in gamma space. + * + * IMPORTANT: for best results the input read for the color/luma edge + * detection should *NOT* be sRGB. + * + * 6. Before including SMAA.h you'll have to setup the render target metrics, + * the target and any optional configuration defines. Optionally you can + * use a preset. + * + * You have the following targets available: + * SMAA_HLSL_3 + * SMAA_HLSL_4 + * SMAA_HLSL_4_1 + * SMAA_GLSL_3 * + * SMAA_GLSL_4 * + * + * * (See SMAA_INCLUDE_VS and SMAA_INCLUDE_PS below). + * + * And four presets: + * SMAA_PRESET_LOW (%60 of the quality) + * SMAA_PRESET_MEDIUM (%80 of the quality) + * SMAA_PRESET_HIGH (%95 of the quality) + * SMAA_PRESET_ULTRA (%99 of the quality) + * + * For example: + * #define SMAA_RT_METRICS float4(1.0 / 1280.0, 1.0 / 720.0, 1280.0, 720.0) + * #define SMAA_HLSL_4 + * #define SMAA_PRESET_HIGH + * #include "SMAA.h" + * + * Note that SMAA_RT_METRICS doesn't need to be a macro, it can be a + * uniform variable. The code is designed to minimize the impact of not + * using a constant value, but it is still better to hardcode it. + * + * Depending on how you encoded 'areaTex' and 'searchTex', you may have to + * add (and customize) the following defines before including SMAA.h: + * #define SMAA_AREATEX_SELECT(sample) sample.rg + * #define SMAA_SEARCHTEX_SELECT(sample) sample.r + * + * If your engine is already using porting macros, you can define + * SMAA_CUSTOM_SL, and define the porting functions by yourself. + * + * 7. Then, you'll have to setup the passes as indicated in the scheme above. + * You can take a look into SMAA.fx, to see how we did it for our demo. + * Checkout the function wrappers, you may want to copy-paste them! + * + * 8. It's recommended to validate the produced |edgesTex| and |blendTex|. + * You can use a screenshot from your engine to compare the |edgesTex| + * and |blendTex| produced inside of the engine with the results obtained + * with the reference demo. + * + * 9. After you get the last pass to work, it's time to optimize. You'll have + * to initialize a stencil buffer in the first pass (discard is already in + * the code), then mask execution by using it the second pass. The last + * pass should be executed in all pixels. + * + * + * After this point you can choose to enable predicated thresholding, + * temporal supersampling and motion blur integration: + * + * a) If you want to use predicated thresholding, take a look into + * SMAA_PREDICATION; you'll need to pass an extra texture in the edge + * detection pass. + * + * b) If you want to enable temporal supersampling (SMAA T2x): + * + * 1. The first step is to render using subpixel jitters. I won't go into + * detail, but it's as simple as moving each vertex position in the + * vertex shader, you can check how we do it in our DX10 demo. + * + * 2. Then, you must setup the temporal resolve. You may want to take a look + * into SMAAResolve for resolving 2x modes. After you get it working, you'll + * probably see ghosting everywhere. But fear not, you can enable the + * CryENGINE temporal reprojection by setting the SMAA_REPROJECTION macro. + * Check out SMAA_DECODE_VELOCITY if your velocity buffer is encoded. + * + * 3. The next step is to apply SMAA to each subpixel jittered frame, just as + * done for 1x. + * + * 4. At this point you should already have something usable, but for best + * results the proper area textures must be set depending on current jitter. + * For this, the parameter 'subsampleIndices' of + * 'SMAABlendingWeightCalculationPS' must be set as follows, for our T2x + * mode: + * + * @SUBSAMPLE_INDICES + * + * | S# | Camera Jitter | subsampleIndices | + * +----+------------------+---------------------+ + * | 0 | ( 0.25, -0.25) | float4(1, 1, 1, 0) | + * | 1 | (-0.25, 0.25) | float4(2, 2, 2, 0) | + * + * These jitter positions assume a bottom-to-top y axis. S# stands for the + * sample number. + * + * More information about temporal supersampling here: + * http://iryoku.com/aacourse/downloads/13-Anti-Aliasing-Methods-in-CryENGINE-3.pdf + * + * c) If you want to enable spatial multisampling (SMAA S2x): + * + * 1. The scene must be rendered using MSAA 2x. The MSAA 2x buffer must be + * created with: + * - DX10: see below (*) + * - DX10.1: D3D10_STANDARD_MULTISAMPLE_PATTERN or + * - DX11: D3D11_STANDARD_MULTISAMPLE_PATTERN + * + * This allows to ensure that the subsample order matches the table in + * @SUBSAMPLE_INDICES. + * + * (*) In the case of DX10, we refer the reader to: + * - SMAA::detectMSAAOrder and + * - SMAA::msaaReorder + * + * These functions allow to match the standard multisample patterns by + * detecting the subsample order for a specific GPU, and reordering + * them appropriately. + * + * 2. A shader must be run to output each subsample into a separate buffer + * (DX10 is required). You can use SMAASeparate for this purpose, or just do + * it in an existing pass (for example, in the tone mapping pass, which has + * the advantage of feeding tone mapped subsamples to SMAA, which will yield + * better results). + * + * 3. The full SMAA 1x pipeline must be run for each separated buffer, storing + * the results in the final buffer. The second run should alpha blend with + * the existing final buffer using a blending factor of 0.5. + * 'subsampleIndices' must be adjusted as in the SMAA T2x case (see point + * b). + * + * d) If you want to enable temporal supersampling on top of SMAA S2x + * (which actually is SMAA 4x): + * + * 1. SMAA 4x consists on temporally jittering SMAA S2x, so the first step is + * to calculate SMAA S2x for current frame. In this case, 'subsampleIndices' + * must be set as follows: + * + * | F# | S# | Camera Jitter | Net Jitter | subsampleIndices | + * +----+----+--------------------+-------------------+----------------------+ + * | 0 | 0 | ( 0.125, 0.125) | ( 0.375, -0.125) | float4(5, 3, 1, 3) | + * | 0 | 1 | ( 0.125, 0.125) | (-0.125, 0.375) | float4(4, 6, 2, 3) | + * +----+----+--------------------+-------------------+----------------------+ + * | 1 | 2 | (-0.125, -0.125) | ( 0.125, -0.375) | float4(3, 5, 1, 4) | + * | 1 | 3 | (-0.125, -0.125) | (-0.375, 0.125) | float4(6, 4, 2, 4) | + * + * These jitter positions assume a bottom-to-top y axis. F# stands for the + * frame number. S# stands for the sample number. + * + * 2. After calculating SMAA S2x for current frame (with the new subsample + * indices), previous frame must be reprojected as in SMAA T2x mode (see + * point b). + * + * e) If motion blur is used, you may want to do the edge detection pass + * together with motion blur. This has two advantages: + * + * 1. Pixels under heavy motion can be omitted from the edge detection process. + * For these pixels we can just store "no edge", as motion blur will take + * care of them. + * 2. The center pixel tap is reused. + * + * Note that in this case depth testing should be used instead of stenciling, + * as we have to write all the pixels in the motion blur pass. + * + * That's it! + */ + +//----------------------------------------------------------------------------- +// SMAA Presets + +/** + * Note that if you use one of these presets, the following configuration + * macros will be ignored if set in the "Configurable Defines" section. + */ + +#if defined(SMAA_PRESET_LOW) +#define SMAA_THRESHOLD 0.15 +#define SMAA_MAX_SEARCH_STEPS 4 +#define SMAA_DISABLE_DIAG_DETECTION +#define SMAA_DISABLE_CORNER_DETECTION +#elif defined(SMAA_PRESET_MEDIUM) +#define SMAA_THRESHOLD 0.1 +#define SMAA_MAX_SEARCH_STEPS 8 +#define SMAA_DISABLE_DIAG_DETECTION +#define SMAA_DISABLE_CORNER_DETECTION +#elif defined(SMAA_PRESET_HIGH) +#define SMAA_THRESHOLD 0.1 +#define SMAA_MAX_SEARCH_STEPS 16 +#define SMAA_MAX_SEARCH_STEPS_DIAG 8 +#define SMAA_CORNER_ROUNDING 25 +#elif defined(SMAA_PRESET_ULTRA) +#define SMAA_THRESHOLD 0.05 +#define SMAA_MAX_SEARCH_STEPS 32 +#define SMAA_MAX_SEARCH_STEPS_DIAG 16 +#define SMAA_CORNER_ROUNDING 25 +#endif + +//----------------------------------------------------------------------------- +// Configurable Defines + +/** + * SMAA_THRESHOLD specifies the threshold or sensitivity to edges. + * Lowering this value you will be able to detect more edges at the expense of + * performance. + * + * Range: [0, 0.5] + * 0.1 is a reasonable value, and allows to catch most visible edges. + * 0.05 is a rather overkill value, that allows to catch 'em all. + * + * If temporal supersampling is used, 0.2 could be a reasonable value, as low + * contrast edges are properly filtered by just 2x. + */ +#ifndef SMAA_THRESHOLD +#define SMAA_THRESHOLD 0.1 +#endif + +/** + * SMAA_DEPTH_THRESHOLD specifies the threshold for depth edge detection. + * + * Range: depends on the depth range of the scene. + */ +#ifndef SMAA_DEPTH_THRESHOLD +#define SMAA_DEPTH_THRESHOLD (0.1 * SMAA_THRESHOLD) +#endif + +/** + * SMAA_MAX_SEARCH_STEPS specifies the maximum steps performed in the + * horizontal/vertical pattern searches, at each side of the pixel. + * + * In number of pixels, it's actually the double. So the maximum line length + * perfectly handled by, for example 16, is 64 (by perfectly, we meant that + * longer lines won't look as good, but still antialiased). + * + * Range: [0, 112] + */ +#ifndef SMAA_MAX_SEARCH_STEPS +#define SMAA_MAX_SEARCH_STEPS 16 +#endif + +/** + * SMAA_MAX_SEARCH_STEPS_DIAG specifies the maximum steps performed in the + * diagonal pattern searches, at each side of the pixel. In this case we jump + * one pixel at time, instead of two. + * + * Range: [0, 20] + * + * On high-end machines it is cheap (between a 0.8x and 0.9x slower for 16 + * steps), but it can have a significant impact on older machines. + * + * Define SMAA_DISABLE_DIAG_DETECTION to disable diagonal processing. + */ +#ifndef SMAA_MAX_SEARCH_STEPS_DIAG +#define SMAA_MAX_SEARCH_STEPS_DIAG 8 +#endif + +/** + * SMAA_CORNER_ROUNDING specifies how much sharp corners will be rounded. + * + * Range: [0, 100] + * + * Define SMAA_DISABLE_CORNER_DETECTION to disable corner processing. + */ +#ifndef SMAA_CORNER_ROUNDING +#define SMAA_CORNER_ROUNDING 25 +#endif + +/** + * If there is an neighbor edge that has SMAA_LOCAL_CONTRAST_FACTOR times + * bigger contrast than current edge, current edge will be discarded. + * + * This allows to eliminate spurious crossing edges, and is based on the fact + * that, if there is too much contrast in a direction, that will hide + * perceptually contrast in the other neighbors. + */ +#ifndef SMAA_LOCAL_CONTRAST_ADAPTATION_FACTOR +#define SMAA_LOCAL_CONTRAST_ADAPTATION_FACTOR 2.0 +#endif + +/** + * Predicated thresholding allows to better preserve texture details and to + * improve performance, by decreasing the number of detected edges using an + * additional buffer like the light accumulation buffer, object ids or even the + * depth buffer (the depth buffer usage may be limited to indoor or short range + * scenes). + * + * It locally decreases the luma or color threshold if an edge is found in an + * additional buffer (so the global threshold can be higher). + * + * This method was developed by Playstation EDGE MLAA team, and used in + * Killzone 3, by using the light accumulation buffer. More information here: + * http://iryoku.com/aacourse/downloads/06-MLAA-on-PS3.pptx + */ +#ifndef SMAA_PREDICATION +#define SMAA_PREDICATION 0 +#endif + +/** + * Threshold to be used in the additional predication buffer. + * + * Range: depends on the input, so you'll have to find the magic number that + * works for you. + */ +#ifndef SMAA_PREDICATION_THRESHOLD +#define SMAA_PREDICATION_THRESHOLD 0.01 +#endif + +/** + * How much to scale the global threshold used for luma or color edge + * detection when using predication. + * + * Range: [1, 5] + */ +#ifndef SMAA_PREDICATION_SCALE +#define SMAA_PREDICATION_SCALE 2.0 +#endif + +/** + * How much to locally decrease the threshold. + * + * Range: [0, 1] + */ +#ifndef SMAA_PREDICATION_STRENGTH +#define SMAA_PREDICATION_STRENGTH 0.4 +#endif + +/** + * Temporal reprojection allows to remove ghosting artifacts when using + * temporal supersampling. We use the CryEngine 3 method which also introduces + * velocity weighting. This feature is of extreme importance for totally + * removing ghosting. More information here: + * http://iryoku.com/aacourse/downloads/13-Anti-Aliasing-Methods-in-CryENGINE-3.pdf + * + * Note that you'll need to setup a velocity buffer for enabling reprojection. + * For static geometry, saving the previous depth buffer is a viable + * alternative. + */ +#ifndef SMAA_REPROJECTION +#define SMAA_REPROJECTION 0 +#endif + +/** + * SMAA_REPROJECTION_WEIGHT_SCALE controls the velocity weighting. It allows to + * remove ghosting trails behind the moving object, which are not removed by + * just using reprojection. Using low values will exhibit ghosting, while using + * high values will disable temporal supersampling under motion. + * + * Behind the scenes, velocity weighting removes temporal supersampling when + * the velocity of the subsamples differs (meaning they are different objects). + * + * Range: [0, 80] + */ +#ifndef SMAA_REPROJECTION_WEIGHT_SCALE +#define SMAA_REPROJECTION_WEIGHT_SCALE 30.0 +#endif + +/** + * On some compilers, discard cannot be used in vertex shaders. Thus, they need + * to be compiled separately. + */ +#ifndef SMAA_INCLUDE_VS +#define SMAA_INCLUDE_VS 1 +#endif +#ifndef SMAA_INCLUDE_PS +#define SMAA_INCLUDE_PS 1 +#endif + +//----------------------------------------------------------------------------- +// Texture Access Defines + +#ifndef SMAA_AREATEX_SELECT +#if defined(SMAA_HLSL_3) +#define SMAA_AREATEX_SELECT(sample) sample.ra +#else +#define SMAA_AREATEX_SELECT(sample) sample.rg +#endif +#endif + +#ifndef SMAA_SEARCHTEX_SELECT +#define SMAA_SEARCHTEX_SELECT(sample) sample.r +#endif + +#ifndef SMAA_DECODE_VELOCITY +#define SMAA_DECODE_VELOCITY(sample) sample.rg +#endif + +//----------------------------------------------------------------------------- +// Non-Configurable Defines + +#define SMAA_AREATEX_MAX_DISTANCE 16 +#define SMAA_AREATEX_MAX_DISTANCE_DIAG 20 +#define SMAA_AREATEX_PIXEL_SIZE (1.0 / float2(160.0, 560.0)) +#define SMAA_AREATEX_SUBTEX_SIZE (1.0 / 7.0) +#define SMAA_SEARCHTEX_SIZE float2(66.0, 33.0) +#define SMAA_SEARCHTEX_PACKED_SIZE float2(64.0, 16.0) +#define SMAA_CORNER_ROUNDING_NORM (float(SMAA_CORNER_ROUNDING) / 100.0) + +//----------------------------------------------------------------------------- +// Porting Functions + +#if defined(SMAA_HLSL_3) +#define SMAATexture2D(tex) sampler2D tex +#define SMAATexturePass2D(tex) tex +#define SMAASampleLevelZero(tex, coord) tex2Dlod(tex, float4(coord, 0.0, 0.0)) +#define SMAASampleLevelZeroPoint(tex, coord) tex2Dlod(tex, float4(coord, 0.0, 0.0)) +#define SMAASampleLevelZeroOffset(tex, coord, offset) tex2Dlod(tex, float4(coord + offset * SMAA_RT_METRICS.xy, 0.0, 0.0)) +#define SMAASample(tex, coord) tex2D(tex, coord) +#define SMAASamplePoint(tex, coord) tex2D(tex, coord) +#define SMAASampleOffset(tex, coord, offset) tex2D(tex, coord + offset * SMAA_RT_METRICS.xy) +#define SMAA_FLATTEN [flatten] +#define SMAA_BRANCH [branch] +#endif +#if defined(SMAA_HLSL_4) || defined(SMAA_HLSL_4_1) +SamplerState LinearSampler { Filter = MIN_MAG_LINEAR_MIP_POINT; AddressU = Clamp; AddressV = Clamp; }; +SamplerState PointSampler { Filter = MIN_MAG_MIP_POINT; AddressU = Clamp; AddressV = Clamp; }; +#define SMAATexture2D(tex) Texture2D tex +#define SMAATexturePass2D(tex) tex +#define SMAASampleLevelZero(tex, coord) tex.SampleLevel(LinearSampler, coord, 0) +#define SMAASampleLevelZeroPoint(tex, coord) tex.SampleLevel(PointSampler, coord, 0) +#define SMAASampleLevelZeroOffset(tex, coord, offset) tex.SampleLevel(LinearSampler, coord, 0, offset) +#define SMAASample(tex, coord) tex.Sample(LinearSampler, coord) +#define SMAASamplePoint(tex, coord) tex.Sample(PointSampler, coord) +#define SMAASampleOffset(tex, coord, offset) tex.Sample(LinearSampler, coord, offset) +#define SMAA_FLATTEN [flatten] +#define SMAA_BRANCH [branch] +#define SMAATexture2DMS2(tex) Texture2DMS tex +#define SMAALoad(tex, pos, sample) tex.Load(pos, sample) +#if defined(SMAA_HLSL_4_1) +#define SMAAGather(tex, coord) tex.Gather(LinearSampler, coord, 0) +#endif +#endif +#if defined(SMAA_GLSL_3) || defined(SMAA_GLSL_4) +#define SMAATexture2D(tex) sampler2D tex +#define SMAATexturePass2D(tex) tex +#define SMAASampleLevelZero(tex, coord) textureLod(tex, coord, 0.0) +#define SMAASampleLevelZeroPoint(tex, coord) textureLod(tex, coord, 0.0) +#define SMAASampleLevelZeroOffset(tex, coord, offset) textureLodOffset(tex, coord, 0.0, offset) +#define SMAASample(tex, coord) texture(tex, coord) +#define SMAASamplePoint(tex, coord) texture(tex, coord) +#define SMAASampleOffset(tex, coord, offset) texture(tex, coord, offset) +#define SMAA_FLATTEN +#define SMAA_BRANCH +#define lerp(a, b, t) mix(a, b, t) +#define saturate(a) clamp(a, 0.0, 1.0) +#if defined(SMAA_GLSL_4) +#define mad(a, b, c) fma(a, b, c) +#define SMAAGather(tex, coord) textureGather(tex, coord) +#else +#define mad(a, b, c) (a * b + c) +#endif +#define float2 vec2 +#define float3 vec3 +#define float4 vec4 +#define int2 ivec2 +#define int3 ivec3 +#define int4 ivec4 +#define bool2 bvec2 +#define bool3 bvec3 +#define bool4 bvec4 +#endif + +#if !defined(SMAA_HLSL_3) && !defined(SMAA_HLSL_4) && !defined(SMAA_HLSL_4_1) && !defined(SMAA_GLSL_3) && !defined(SMAA_GLSL_4) && !defined(SMAA_CUSTOM_SL) +#error you must define the shading language: SMAA_HLSL_*, SMAA_GLSL_* or SMAA_CUSTOM_SL +#endif + +//----------------------------------------------------------------------------- +// Misc functions + +/** + * Gathers current pixel, and the top-left neighbors. + */ +float3 SMAAGatherNeighbours(float2 texcoord, + float4 offset[3], + SMAATexture2D(tex)) { + #ifdef SMAAGather + return SMAAGather(tex, texcoord + SMAA_RT_METRICS.xy * float2(-0.5, -0.5)).grb; + #else + float P = SMAASamplePoint(tex, texcoord).r; + float Pleft = SMAASamplePoint(tex, offset[0].xy).r; + float Ptop = SMAASamplePoint(tex, offset[0].zw).r; + return float3(P, Pleft, Ptop); + #endif +} + +/** + * Adjusts the threshold by means of predication. + */ +float2 SMAACalculatePredicatedThreshold(float2 texcoord, + float4 offset[3], + SMAATexture2D(predicationTex)) { + float3 neighbours = SMAAGatherNeighbours(texcoord, offset, SMAATexturePass2D(predicationTex)); + float2 delta = abs(neighbours.xx - neighbours.yz); + float2 edges = step(SMAA_PREDICATION_THRESHOLD, delta); + return SMAA_PREDICATION_SCALE * SMAA_THRESHOLD * (1.0 - SMAA_PREDICATION_STRENGTH * edges); +} + +/** + * Conditional move: + */ +void SMAAMovc(bool2 cond, inout float2 variable, float2 value) { + SMAA_FLATTEN if (cond.x) variable.x = value.x; + SMAA_FLATTEN if (cond.y) variable.y = value.y; +} + +void SMAAMovc(bool4 cond, inout float4 variable, float4 value) { + SMAAMovc(cond.xy, variable.xy, value.xy); + SMAAMovc(cond.zw, variable.zw, value.zw); +} + + +#if SMAA_INCLUDE_VS +//----------------------------------------------------------------------------- +// Vertex Shaders + +/** + * Edge Detection Vertex Shader + */ +void SMAAEdgeDetectionVS(float2 texcoord, + out float4 offset[3]) { + offset[0] = mad(SMAA_RT_METRICS.xyxy, float4(-1.0, 0.0, 0.0, -1.0), texcoord.xyxy); + offset[1] = mad(SMAA_RT_METRICS.xyxy, float4( 1.0, 0.0, 0.0, 1.0), texcoord.xyxy); + offset[2] = mad(SMAA_RT_METRICS.xyxy, float4(-2.0, 0.0, 0.0, -2.0), texcoord.xyxy); +} + +/** + * Blend Weight Calculation Vertex Shader + */ +void SMAABlendingWeightCalculationVS(float2 texcoord, + out float2 pixcoord, + out float4 offset[3]) { + pixcoord = texcoord * SMAA_RT_METRICS.zw; + + // We will use these offsets for the searches later on (see @PSEUDO_GATHER4): + offset[0] = mad(SMAA_RT_METRICS.xyxy, float4(-0.25, -0.125, 1.25, -0.125), texcoord.xyxy); + offset[1] = mad(SMAA_RT_METRICS.xyxy, float4(-0.125, -0.25, -0.125, 1.25), texcoord.xyxy); + + // And these for the searches, they indicate the ends of the loops: + offset[2] = mad(SMAA_RT_METRICS.xxyy, + float4(-2.0, 2.0, -2.0, 2.0) * float(SMAA_MAX_SEARCH_STEPS), + float4(offset[0].xz, offset[1].yw)); +} + +/** + * Neighborhood Blending Vertex Shader + */ +void SMAANeighborhoodBlendingVS(float2 texcoord, + out float4 offset) { + offset = mad(SMAA_RT_METRICS.xyxy, float4( 1.0, 0.0, 0.0, 1.0), texcoord.xyxy); +} +#endif // SMAA_INCLUDE_VS + +#if SMAA_INCLUDE_PS +//----------------------------------------------------------------------------- +// Edge Detection Pixel Shaders (First Pass) + +/** + * Luma Edge Detection + * + * IMPORTANT NOTICE: luma edge detection requires gamma-corrected colors, and + * thus 'colorTex' should be a non-sRGB texture. + */ +float2 SMAALumaEdgeDetectionPS(float2 texcoord, + float4 offset[3], + SMAATexture2D(colorTex) + #if SMAA_PREDICATION + , SMAATexture2D(predicationTex) + #endif + ) { + // Calculate the threshold: + #if SMAA_PREDICATION + float2 threshold = SMAACalculatePredicatedThreshold(texcoord, offset, SMAATexturePass2D(predicationTex)); + #else + float2 threshold = float2(SMAA_THRESHOLD, SMAA_THRESHOLD); + #endif + + // Calculate lumas: + float3 weights = float3(0.2126, 0.7152, 0.0722); + float L = dot(SMAASamplePoint(colorTex, texcoord).rgb, weights); + + float Lleft = dot(SMAASamplePoint(colorTex, offset[0].xy).rgb, weights); + float Ltop = dot(SMAASamplePoint(colorTex, offset[0].zw).rgb, weights); + + // We do the usual threshold: + float4 delta; + delta.xy = abs(L - float2(Lleft, Ltop)); + float2 edges = step(threshold, delta.xy); + + // Then discard if there is no edge: + if (dot(edges, float2(1.0, 1.0)) == 0.0) + discard; + + // Calculate right and bottom deltas: + float Lright = dot(SMAASamplePoint(colorTex, offset[1].xy).rgb, weights); + float Lbottom = dot(SMAASamplePoint(colorTex, offset[1].zw).rgb, weights); + delta.zw = abs(L - float2(Lright, Lbottom)); + + // Calculate the maximum delta in the direct neighborhood: + float2 maxDelta = max(delta.xy, delta.zw); + + // Calculate left-left and top-top deltas: + float Lleftleft = dot(SMAASamplePoint(colorTex, offset[2].xy).rgb, weights); + float Ltoptop = dot(SMAASamplePoint(colorTex, offset[2].zw).rgb, weights); + delta.zw = abs(float2(Lleft, Ltop) - float2(Lleftleft, Ltoptop)); + + // Calculate the final maximum delta: + maxDelta = max(maxDelta.xy, delta.zw); + float finalDelta = max(maxDelta.x, maxDelta.y); + + // Local contrast adaptation: + edges.xy *= step(finalDelta, SMAA_LOCAL_CONTRAST_ADAPTATION_FACTOR * delta.xy); + + return edges; +} + +/** + * Color Edge Detection + * + * IMPORTANT NOTICE: color edge detection requires gamma-corrected colors, and + * thus 'colorTex' should be a non-sRGB texture. + */ +float2 SMAAColorEdgeDetectionPS(float2 texcoord, + float4 offset[3], + SMAATexture2D(colorTex) + #if SMAA_PREDICATION + , SMAATexture2D(predicationTex) + #endif + ) { + // Calculate the threshold: + #if SMAA_PREDICATION + float2 threshold = SMAACalculatePredicatedThreshold(texcoord, offset, predicationTex); + #else + float2 threshold = float2(SMAA_THRESHOLD, SMAA_THRESHOLD); + #endif + + // Calculate color deltas: + float4 delta; + float3 C = SMAASamplePoint(colorTex, texcoord).rgb; + + float3 Cleft = SMAASamplePoint(colorTex, offset[0].xy).rgb; + float3 t = abs(C - Cleft); + delta.x = max(max(t.r, t.g), t.b); + + float3 Ctop = SMAASamplePoint(colorTex, offset[0].zw).rgb; + t = abs(C - Ctop); + delta.y = max(max(t.r, t.g), t.b); + + // We do the usual threshold: + float2 edges = step(threshold, delta.xy); + + // Then discard if there is no edge: + if (dot(edges, float2(1.0, 1.0)) == 0.0) + discard; + + // Calculate right and bottom deltas: + float3 Cright = SMAASamplePoint(colorTex, offset[1].xy).rgb; + t = abs(C - Cright); + delta.z = max(max(t.r, t.g), t.b); + + float3 Cbottom = SMAASamplePoint(colorTex, offset[1].zw).rgb; + t = abs(C - Cbottom); + delta.w = max(max(t.r, t.g), t.b); + + // Calculate the maximum delta in the direct neighborhood: + float2 maxDelta = max(delta.xy, delta.zw); + + // Calculate left-left and top-top deltas: + float3 Cleftleft = SMAASamplePoint(colorTex, offset[2].xy).rgb; + t = abs(Cleft - Cleftleft); + delta.z = max(max(t.r, t.g), t.b); + + float3 Ctoptop = SMAASamplePoint(colorTex, offset[2].zw).rgb; + t = abs(Ctop - Ctoptop); + delta.w = max(max(t.r, t.g), t.b); + + // Calculate the final maximum delta: + maxDelta = max(maxDelta.xy, delta.zw); + float finalDelta = max(maxDelta.x, maxDelta.y); + + // Local contrast adaptation: + edges.xy *= step(finalDelta, SMAA_LOCAL_CONTRAST_ADAPTATION_FACTOR * delta.xy); + + return edges; +} + +/** + * Depth Edge Detection + */ +float2 SMAADepthEdgeDetectionPS(float2 texcoord, + float4 offset[3], + SMAATexture2D(depthTex)) { + float3 neighbours = SMAAGatherNeighbours(texcoord, offset, SMAATexturePass2D(depthTex)); + float2 delta = abs(neighbours.xx - float2(neighbours.y, neighbours.z)); + float2 edges = step(SMAA_DEPTH_THRESHOLD, delta); + + if (dot(edges, float2(1.0, 1.0)) == 0.0) + discard; + + return edges; +} + +//----------------------------------------------------------------------------- +// Diagonal Search Functions + +#if !defined(SMAA_DISABLE_DIAG_DETECTION) + +/** + * Allows to decode two binary values from a bilinear-filtered access. + */ +float2 SMAADecodeDiagBilinearAccess(float2 e) { + // Bilinear access for fetching 'e' have a 0.25 offset, and we are + // interested in the R and G edges: + // + // +---G---+-------+ + // | x o R x | + // +-------+-------+ + // + // Then, if one of these edge is enabled: + // Red: (0.75 * X + 0.25 * 1) => 0.25 or 1.0 + // Green: (0.75 * 1 + 0.25 * X) => 0.75 or 1.0 + // + // This function will unpack the values (mad + mul + round): + // wolframalpha.com: round(x * abs(5 * x - 5 * 0.75)) plot 0 to 1 + e.r = e.r * abs(5.0 * e.r - 5.0 * 0.75); + return round(e); +} + +float4 SMAADecodeDiagBilinearAccess(float4 e) { + e.rb = e.rb * abs(5.0 * e.rb - 5.0 * 0.75); + return round(e); +} + +/** + * These functions allows to perform diagonal pattern searches. + */ +float2 SMAASearchDiag1(SMAATexture2D(edgesTex), float2 texcoord, float2 dir, out float2 e) { + float4 coord = float4(texcoord, -1.0, 1.0); + float3 t = float3(SMAA_RT_METRICS.xy, 1.0); + while (coord.z < float(SMAA_MAX_SEARCH_STEPS_DIAG - 1) && + coord.w > 0.9) { + coord.xyz = mad(t, float3(dir, 1.0), coord.xyz); + e = SMAASampleLevelZero(edgesTex, coord.xy).rg; + coord.w = dot(e, float2(0.5, 0.5)); + } + return coord.zw; +} + +float2 SMAASearchDiag2(SMAATexture2D(edgesTex), float2 texcoord, float2 dir, out float2 e) { + float4 coord = float4(texcoord, -1.0, 1.0); + coord.x += 0.25 * SMAA_RT_METRICS.x; // See @SearchDiag2Optimization + float3 t = float3(SMAA_RT_METRICS.xy, 1.0); + while (coord.z < float(SMAA_MAX_SEARCH_STEPS_DIAG - 1) && + coord.w > 0.9) { + coord.xyz = mad(t, float3(dir, 1.0), coord.xyz); + + // @SearchDiag2Optimization + // Fetch both edges at once using bilinear filtering: + e = SMAASampleLevelZero(edgesTex, coord.xy).rg; + e = SMAADecodeDiagBilinearAccess(e); + + // Non-optimized version: + // e.g = SMAASampleLevelZero(edgesTex, coord.xy).g; + // e.r = SMAASampleLevelZeroOffset(edgesTex, coord.xy, int2(1, 0)).r; + + coord.w = dot(e, float2(0.5, 0.5)); + } + return coord.zw; +} + +/** + * Similar to SMAAArea, this calculates the area corresponding to a certain + * diagonal distance and crossing edges 'e'. + */ +float2 SMAAAreaDiag(SMAATexture2D(areaTex), float2 dist, float2 e, float offset) { + float2 texcoord = mad(float2(SMAA_AREATEX_MAX_DISTANCE_DIAG, SMAA_AREATEX_MAX_DISTANCE_DIAG), e, dist); + + // We do a scale and bias for mapping to texel space: + texcoord = mad(SMAA_AREATEX_PIXEL_SIZE, texcoord, 0.5 * SMAA_AREATEX_PIXEL_SIZE); + + // Diagonal areas are on the second half of the texture: + texcoord.x += 0.5; + + // Move to proper place, according to the subpixel offset: + texcoord.y += SMAA_AREATEX_SUBTEX_SIZE * offset; + + // Do it! + return SMAA_AREATEX_SELECT(SMAASampleLevelZero(areaTex, texcoord)); +} + +/** + * This searches for diagonal patterns and returns the corresponding weights. + */ +float2 SMAACalculateDiagWeights(SMAATexture2D(edgesTex), SMAATexture2D(areaTex), float2 texcoord, float2 e, float4 subsampleIndices) { + float2 weights = float2(0.0, 0.0); + + // Search for the line ends: + float4 d; + float2 end; + if (e.r > 0.0) { + d.xz = SMAASearchDiag1(SMAATexturePass2D(edgesTex), texcoord, float2(-1.0, 1.0), end); + d.x += float(end.y > 0.9); + } else + d.xz = float2(0.0, 0.0); + d.yw = SMAASearchDiag1(SMAATexturePass2D(edgesTex), texcoord, float2(1.0, -1.0), end); + + SMAA_BRANCH + if (d.x + d.y > 2.0) { // d.x + d.y + 1 > 3 + // Fetch the crossing edges: + float4 coords = mad(float4(-d.x + 0.25, d.x, d.y, -d.y - 0.25), SMAA_RT_METRICS.xyxy, texcoord.xyxy); + float4 c; + c.xy = SMAASampleLevelZeroOffset(edgesTex, coords.xy, int2(-1, 0)).rg; + c.zw = SMAASampleLevelZeroOffset(edgesTex, coords.zw, int2( 1, 0)).rg; + c.yxwz = SMAADecodeDiagBilinearAccess(c.xyzw); + + // Non-optimized version: + // float4 coords = mad(float4(-d.x, d.x, d.y, -d.y), SMAA_RT_METRICS.xyxy, texcoord.xyxy); + // float4 c; + // c.x = SMAASampleLevelZeroOffset(edgesTex, coords.xy, int2(-1, 0)).g; + // c.y = SMAASampleLevelZeroOffset(edgesTex, coords.xy, int2( 0, 0)).r; + // c.z = SMAASampleLevelZeroOffset(edgesTex, coords.zw, int2( 1, 0)).g; + // c.w = SMAASampleLevelZeroOffset(edgesTex, coords.zw, int2( 1, -1)).r; + + // Merge crossing edges at each side into a single value: + float2 cc = mad(float2(2.0, 2.0), c.xz, c.yw); + + // Remove the crossing edge if we didn't found the end of the line: + SMAAMovc(bool2(step(0.9, d.zw)), cc, float2(0.0, 0.0)); + + // Fetch the areas for this line: + weights += SMAAAreaDiag(SMAATexturePass2D(areaTex), d.xy, cc, subsampleIndices.z); + } + + // Search for the line ends: + d.xz = SMAASearchDiag2(SMAATexturePass2D(edgesTex), texcoord, float2(-1.0, -1.0), end); + if (SMAASampleLevelZeroOffset(edgesTex, texcoord, int2(1, 0)).r > 0.0) { + d.yw = SMAASearchDiag2(SMAATexturePass2D(edgesTex), texcoord, float2(1.0, 1.0), end); + d.y += float(end.y > 0.9); + } else + d.yw = float2(0.0, 0.0); + + SMAA_BRANCH + if (d.x + d.y > 2.0) { // d.x + d.y + 1 > 3 + // Fetch the crossing edges: + float4 coords = mad(float4(-d.x, -d.x, d.y, d.y), SMAA_RT_METRICS.xyxy, texcoord.xyxy); + float4 c; + c.x = SMAASampleLevelZeroOffset(edgesTex, coords.xy, int2(-1, 0)).g; + c.y = SMAASampleLevelZeroOffset(edgesTex, coords.xy, int2( 0, -1)).r; + c.zw = SMAASampleLevelZeroOffset(edgesTex, coords.zw, int2( 1, 0)).gr; + float2 cc = mad(float2(2.0, 2.0), c.xz, c.yw); + + // Remove the crossing edge if we didn't found the end of the line: + SMAAMovc(bool2(step(0.9, d.zw)), cc, float2(0.0, 0.0)); + + // Fetch the areas for this line: + weights += SMAAAreaDiag(SMAATexturePass2D(areaTex), d.xy, cc, subsampleIndices.w).gr; + } + + return weights; +} +#endif + +//----------------------------------------------------------------------------- +// Horizontal/Vertical Search Functions + +/** + * This allows to determine how much length should we add in the last step + * of the searches. It takes the bilinearly interpolated edge (see + * @PSEUDO_GATHER4), and adds 0, 1 or 2, depending on which edges and + * crossing edges are active. + */ +float SMAASearchLength(SMAATexture2D(searchTex), float2 e, float offset) { + // The texture is flipped vertically, with left and right cases taking half + // of the space horizontally: + float2 scale = SMAA_SEARCHTEX_SIZE * float2(0.5, -1.0); + float2 bias = SMAA_SEARCHTEX_SIZE * float2(offset, 1.0); + + // Scale and bias to access texel centers: + scale += float2(-1.0, 1.0); + bias += float2( 0.5, -0.5); + + // Convert from pixel coordinates to texcoords: + // (We use SMAA_SEARCHTEX_PACKED_SIZE because the texture is cropped) + scale *= 1.0 / SMAA_SEARCHTEX_PACKED_SIZE; + bias *= 1.0 / SMAA_SEARCHTEX_PACKED_SIZE; + + // Lookup the search texture: + return SMAA_SEARCHTEX_SELECT(SMAASampleLevelZero(searchTex, mad(scale, e, bias))); +} + +/** + * Horizontal/vertical search functions for the 2nd pass. + */ +float SMAASearchXLeft(SMAATexture2D(edgesTex), SMAATexture2D(searchTex), float2 texcoord, float end) { + /** + * @PSEUDO_GATHER4 + * This texcoord has been offset by (-0.25, -0.125) in the vertex shader to + * sample between edge, thus fetching four edges in a row. + * Sampling with different offsets in each direction allows to disambiguate + * which edges are active from the four fetched ones. + */ + float2 e = float2(0.0, 1.0); + while (texcoord.x > end && + e.g > 0.8281 && // Is there some edge not activated? + e.r == 0.0) { // Or is there a crossing edge that breaks the line? + e = SMAASampleLevelZero(edgesTex, texcoord).rg; + texcoord = mad(-float2(2.0, 0.0), SMAA_RT_METRICS.xy, texcoord); + } + + float offset = mad(-(255.0 / 127.0), SMAASearchLength(SMAATexturePass2D(searchTex), e, 0.0), 3.25); + return mad(SMAA_RT_METRICS.x, offset, texcoord.x); + + // Non-optimized version: + // We correct the previous (-0.25, -0.125) offset we applied: + // texcoord.x += 0.25 * SMAA_RT_METRICS.x; + + // The searches are bias by 1, so adjust the coords accordingly: + // texcoord.x += SMAA_RT_METRICS.x; + + // Disambiguate the length added by the last step: + // texcoord.x += 2.0 * SMAA_RT_METRICS.x; // Undo last step + // texcoord.x -= SMAA_RT_METRICS.x * (255.0 / 127.0) * SMAASearchLength(SMAATexturePass2D(searchTex), e, 0.0); + // return mad(SMAA_RT_METRICS.x, offset, texcoord.x); +} + +float SMAASearchXRight(SMAATexture2D(edgesTex), SMAATexture2D(searchTex), float2 texcoord, float end) { + float2 e = float2(0.0, 1.0); + while (texcoord.x < end && + e.g > 0.8281 && // Is there some edge not activated? + e.r == 0.0) { // Or is there a crossing edge that breaks the line? + e = SMAASampleLevelZero(edgesTex, texcoord).rg; + texcoord = mad(float2(2.0, 0.0), SMAA_RT_METRICS.xy, texcoord); + } + float offset = mad(-(255.0 / 127.0), SMAASearchLength(SMAATexturePass2D(searchTex), e, 0.5), 3.25); + return mad(-SMAA_RT_METRICS.x, offset, texcoord.x); +} + +float SMAASearchYUp(SMAATexture2D(edgesTex), SMAATexture2D(searchTex), float2 texcoord, float end) { + float2 e = float2(1.0, 0.0); + while (texcoord.y > end && + e.r > 0.8281 && // Is there some edge not activated? + e.g == 0.0) { // Or is there a crossing edge that breaks the line? + e = SMAASampleLevelZero(edgesTex, texcoord).rg; + texcoord = mad(-float2(0.0, 2.0), SMAA_RT_METRICS.xy, texcoord); + } + float offset = mad(-(255.0 / 127.0), SMAASearchLength(SMAATexturePass2D(searchTex), e.gr, 0.0), 3.25); + return mad(SMAA_RT_METRICS.y, offset, texcoord.y); +} + +float SMAASearchYDown(SMAATexture2D(edgesTex), SMAATexture2D(searchTex), float2 texcoord, float end) { + float2 e = float2(1.0, 0.0); + while (texcoord.y < end && + e.r > 0.8281 && // Is there some edge not activated? + e.g == 0.0) { // Or is there a crossing edge that breaks the line? + e = SMAASampleLevelZero(edgesTex, texcoord).rg; + texcoord = mad(float2(0.0, 2.0), SMAA_RT_METRICS.xy, texcoord); + } + float offset = mad(-(255.0 / 127.0), SMAASearchLength(SMAATexturePass2D(searchTex), e.gr, 0.5), 3.25); + return mad(-SMAA_RT_METRICS.y, offset, texcoord.y); +} + +/** + * Ok, we have the distance and both crossing edges. So, what are the areas + * at each side of current edge? + */ +float2 SMAAArea(SMAATexture2D(areaTex), float2 dist, float e1, float e2, float offset) { + // Rounding prevents precision errors of bilinear filtering: + float2 texcoord = mad(float2(SMAA_AREATEX_MAX_DISTANCE, SMAA_AREATEX_MAX_DISTANCE), round(4.0 * float2(e1, e2)), dist); + + // We do a scale and bias for mapping to texel space: + texcoord = mad(SMAA_AREATEX_PIXEL_SIZE, texcoord, 0.5 * SMAA_AREATEX_PIXEL_SIZE); + + // Move to proper place, according to the subpixel offset: + texcoord.y = mad(SMAA_AREATEX_SUBTEX_SIZE, offset, texcoord.y); + + // Do it! + return SMAA_AREATEX_SELECT(SMAASampleLevelZero(areaTex, texcoord)); +} + +//----------------------------------------------------------------------------- +// Corner Detection Functions + +void SMAADetectHorizontalCornerPattern(SMAATexture2D(edgesTex), inout float2 weights, float4 texcoord, float2 d) { + #if !defined(SMAA_DISABLE_CORNER_DETECTION) + float2 leftRight = step(d.xy, d.yx); + float2 rounding = (1.0 - SMAA_CORNER_ROUNDING_NORM) * leftRight; + + rounding /= leftRight.x + leftRight.y; // Reduce blending for pixels in the center of a line. + + float2 factor = float2(1.0, 1.0); + factor.x -= rounding.x * SMAASampleLevelZeroOffset(edgesTex, texcoord.xy, int2(0, 1)).r; + factor.x -= rounding.y * SMAASampleLevelZeroOffset(edgesTex, texcoord.zw, int2(1, 1)).r; + factor.y -= rounding.x * SMAASampleLevelZeroOffset(edgesTex, texcoord.xy, int2(0, -2)).r; + factor.y -= rounding.y * SMAASampleLevelZeroOffset(edgesTex, texcoord.zw, int2(1, -2)).r; + + weights *= saturate(factor); + #endif +} + +void SMAADetectVerticalCornerPattern(SMAATexture2D(edgesTex), inout float2 weights, float4 texcoord, float2 d) { + #if !defined(SMAA_DISABLE_CORNER_DETECTION) + float2 leftRight = step(d.xy, d.yx); + float2 rounding = (1.0 - SMAA_CORNER_ROUNDING_NORM) * leftRight; + + rounding /= leftRight.x + leftRight.y; + + float2 factor = float2(1.0, 1.0); + factor.x -= rounding.x * SMAASampleLevelZeroOffset(edgesTex, texcoord.xy, int2( 1, 0)).g; + factor.x -= rounding.y * SMAASampleLevelZeroOffset(edgesTex, texcoord.zw, int2( 1, 1)).g; + factor.y -= rounding.x * SMAASampleLevelZeroOffset(edgesTex, texcoord.xy, int2(-2, 0)).g; + factor.y -= rounding.y * SMAASampleLevelZeroOffset(edgesTex, texcoord.zw, int2(-2, 1)).g; + + weights *= saturate(factor); + #endif +} + +//----------------------------------------------------------------------------- +// Blending Weight Calculation Pixel Shader (Second Pass) + +float4 SMAABlendingWeightCalculationPS(float2 texcoord, + float2 pixcoord, + float4 offset[3], + SMAATexture2D(edgesTex), + SMAATexture2D(areaTex), + SMAATexture2D(searchTex), + float4 subsampleIndices) { // Just pass zero for SMAA 1x, see @SUBSAMPLE_INDICES. + float4 weights = float4(0.0, 0.0, 0.0, 0.0); + + float2 e = SMAASample(edgesTex, texcoord).rg; + + SMAA_BRANCH + if (e.g > 0.0) { // Edge at north + #if !defined(SMAA_DISABLE_DIAG_DETECTION) + // Diagonals have both north and west edges, so searching for them in + // one of the boundaries is enough. + weights.rg = SMAACalculateDiagWeights(SMAATexturePass2D(edgesTex), SMAATexturePass2D(areaTex), texcoord, e, subsampleIndices); + + // We give priority to diagonals, so if we find a diagonal we skip + // horizontal/vertical processing. + SMAA_BRANCH + if (weights.r == -weights.g) { // weights.r + weights.g == 0.0 + #endif + + float2 d; + + // Find the distance to the left: + float3 coords; + coords.x = SMAASearchXLeft(SMAATexturePass2D(edgesTex), SMAATexturePass2D(searchTex), offset[0].xy, offset[2].x); + coords.y = offset[1].y; // offset[1].y = texcoord.y - 0.25 * SMAA_RT_METRICS.y (@CROSSING_OFFSET) + d.x = coords.x; + + // Now fetch the left crossing edges, two at a time using bilinear + // filtering. Sampling at -0.25 (see @CROSSING_OFFSET) enables to + // discern what value each edge has: + float e1 = SMAASampleLevelZero(edgesTex, coords.xy).r; + + // Find the distance to the right: + coords.z = SMAASearchXRight(SMAATexturePass2D(edgesTex), SMAATexturePass2D(searchTex), offset[0].zw, offset[2].y); + d.y = coords.z; + + // We want the distances to be in pixel units (doing this here allow to + // better interleave arithmetic and memory accesses): + d = abs(round(mad(SMAA_RT_METRICS.zz, d, -pixcoord.xx))); + + // SMAAArea below needs a sqrt, as the areas texture is compressed + // quadratically: + float2 sqrt_d = sqrt(d); + + // Fetch the right crossing edges: + float e2 = SMAASampleLevelZeroOffset(edgesTex, coords.zy, int2(1, 0)).r; + + // Ok, we know how this pattern looks like, now it is time for getting + // the actual area: + weights.rg = SMAAArea(SMAATexturePass2D(areaTex), sqrt_d, e1, e2, subsampleIndices.y); + + // Fix corners: + coords.y = texcoord.y; + SMAADetectHorizontalCornerPattern(SMAATexturePass2D(edgesTex), weights.rg, coords.xyzy, d); + + #if !defined(SMAA_DISABLE_DIAG_DETECTION) + } else + e.r = 0.0; // Skip vertical processing. + #endif + } + + SMAA_BRANCH + if (e.r > 0.0) { // Edge at west + float2 d; + + // Find the distance to the top: + float3 coords; + coords.y = SMAASearchYUp(SMAATexturePass2D(edgesTex), SMAATexturePass2D(searchTex), offset[1].xy, offset[2].z); + coords.x = offset[0].x; // offset[1].x = texcoord.x - 0.25 * SMAA_RT_METRICS.x; + d.x = coords.y; + + // Fetch the top crossing edges: + float e1 = SMAASampleLevelZero(edgesTex, coords.xy).g; + + // Find the distance to the bottom: + coords.z = SMAASearchYDown(SMAATexturePass2D(edgesTex), SMAATexturePass2D(searchTex), offset[1].zw, offset[2].w); + d.y = coords.z; + + // We want the distances to be in pixel units: + d = abs(round(mad(SMAA_RT_METRICS.ww, d, -pixcoord.yy))); + + // SMAAArea below needs a sqrt, as the areas texture is compressed + // quadratically: + float2 sqrt_d = sqrt(d); + + // Fetch the bottom crossing edges: + float e2 = SMAASampleLevelZeroOffset(edgesTex, coords.xz, int2(0, 1)).g; + + // Get the area for this direction: + weights.ba = SMAAArea(SMAATexturePass2D(areaTex), sqrt_d, e1, e2, subsampleIndices.x); + + // Fix corners: + coords.x = texcoord.x; + SMAADetectVerticalCornerPattern(SMAATexturePass2D(edgesTex), weights.ba, coords.xyxz, d); + } + + return weights; +} + +//----------------------------------------------------------------------------- +// Neighborhood Blending Pixel Shader (Third Pass) + +float4 SMAANeighborhoodBlendingPS(float2 texcoord, + float4 offset, + SMAATexture2D(colorTex), + SMAATexture2D(blendTex) + #if SMAA_REPROJECTION + , SMAATexture2D(velocityTex) + #endif + ) { + // Fetch the blending weights for current pixel: + float4 a; + a.x = SMAASample(blendTex, offset.xy).a; // Right + a.y = SMAASample(blendTex, offset.zw).g; // Top + a.wz = SMAASample(blendTex, texcoord).xz; // Bottom / Left + + // Is there any blending weight with a value greater than 0.0? + SMAA_BRANCH + if (dot(a, float4(1.0, 1.0, 1.0, 1.0)) < 1e-5) { + float4 color = SMAASampleLevelZero(colorTex, texcoord); + + #if SMAA_REPROJECTION + float2 velocity = SMAA_DECODE_VELOCITY(SMAASampleLevelZero(velocityTex, texcoord)); + + // Pack velocity into the alpha channel: + color.a = sqrt(5.0 * length(velocity)); + #endif + + return color; + } else { + bool h = max(a.x, a.z) > max(a.y, a.w); // max(horizontal) > max(vertical) + + // Calculate the blending offsets: + float4 blendingOffset = float4(0.0, a.y, 0.0, a.w); + float2 blendingWeight = a.yw; + SMAAMovc(bool4(h, h, h, h), blendingOffset, float4(a.x, 0.0, a.z, 0.0)); + SMAAMovc(bool2(h, h), blendingWeight, a.xz); + blendingWeight /= dot(blendingWeight, float2(1.0, 1.0)); + + // Calculate the texture coordinates: + float4 blendingCoord = mad(blendingOffset, float4(SMAA_RT_METRICS.xy, -SMAA_RT_METRICS.xy), texcoord.xyxy); + + // We exploit bilinear filtering to mix current pixel with the chosen + // neighbor: + float4 color = blendingWeight.x * SMAASampleLevelZero(colorTex, blendingCoord.xy); + color += blendingWeight.y * SMAASampleLevelZero(colorTex, blendingCoord.zw); + + #if SMAA_REPROJECTION + // Antialias velocity for proper reprojection in a later stage: + float2 velocity = blendingWeight.x * SMAA_DECODE_VELOCITY(SMAASampleLevelZero(velocityTex, blendingCoord.xy)); + velocity += blendingWeight.y * SMAA_DECODE_VELOCITY(SMAASampleLevelZero(velocityTex, blendingCoord.zw)); + + // Pack velocity into the alpha channel: + color.a = sqrt(5.0 * length(velocity)); + #endif + + return color; + } +} + +//----------------------------------------------------------------------------- +// Temporal Resolve Pixel Shader (Optional Pass) + +float4 SMAAResolvePS(float2 texcoord, + SMAATexture2D(currentColorTex), + SMAATexture2D(previousColorTex) + #if SMAA_REPROJECTION + , SMAATexture2D(velocityTex) + #endif + ) { + #if SMAA_REPROJECTION + // Velocity is assumed to be calculated for motion blur, so we need to + // inverse it for reprojection: + float2 velocity = -SMAA_DECODE_VELOCITY(SMAASamplePoint(velocityTex, texcoord).rg); + + // Fetch current pixel: + float4 current = SMAASamplePoint(currentColorTex, texcoord); + + // Reproject current coordinates and fetch previous pixel: + float4 previous = SMAASamplePoint(previousColorTex, texcoord + velocity); + + // Attenuate the previous pixel if the velocity is different: + float delta = abs(current.a * current.a - previous.a * previous.a) / 5.0; + float weight = 0.5 * saturate(1.0 - sqrt(delta) * SMAA_REPROJECTION_WEIGHT_SCALE); + + // Blend the pixels according to the calculated weight: + return lerp(current, previous, weight); + #else + // Just blend the pixels: + float4 current = SMAASamplePoint(currentColorTex, texcoord); + float4 previous = SMAASamplePoint(previousColorTex, texcoord); + return lerp(current, previous, 0.5); + #endif +} + +//----------------------------------------------------------------------------- +// Separate Multisamples Pixel Shader (Optional Pass) + +#ifdef SMAALoad +void SMAASeparatePS(float4 position, + float2 texcoord, + out float4 target0, + out float4 target1, + SMAATexture2DMS2(colorTexMS)) { + int2 pos = int2(position.xy); + target0 = SMAALoad(colorTexMS, pos, 0); + target1 = SMAALoad(colorTexMS, pos, 1); +} +#endif + +//----------------------------------------------------------------------------- +#endif // SMAA_INCLUDE_PS diff --git a/src/video_core/host_shaders/antialiasing/Vulkan/vulkan_smaa_pass0_post.frag b/src/video_core/host_shaders/antialiasing/Vulkan/vulkan_smaa_pass0_post.frag new file mode 100644 index 000000000..722a27b36 --- /dev/null +++ b/src/video_core/host_shaders/antialiasing/Vulkan/vulkan_smaa_pass0_post.frag @@ -0,0 +1,7 @@ +void main() { + if (SMAA_EDT == 0.0) { + color = vec4(SMAALumaEdgeDetectionPS(frag_tex_coord, offset, color_texture), 0.0, 0.0); + } else if (SMAA_EDT <= 1.0) { + color = vec4(SMAAColorEdgeDetectionPS(frag_tex_coord, offset, color_texture), 0.0, 0.0); + } +} diff --git a/src/video_core/host_shaders/antialiasing/Vulkan/vulkan_smaa_pass0_post.vert b/src/video_core/host_shaders/antialiasing/Vulkan/vulkan_smaa_pass0_post.vert new file mode 100644 index 000000000..eed5f71cb --- /dev/null +++ b/src/video_core/host_shaders/antialiasing/Vulkan/vulkan_smaa_pass0_post.vert @@ -0,0 +1,5 @@ +void main() { + gl_Position = vec4(vert_position, 0.0, 1.0); + frag_tex_coord = vert_tex_coord; + SMAAEdgeDetectionVS(vert_tex_coord, offset); +} \ No newline at end of file diff --git a/src/video_core/host_shaders/antialiasing/Vulkan/vulkan_smaa_pass0_pre.frag b/src/video_core/host_shaders/antialiasing/Vulkan/vulkan_smaa_pass0_pre.frag new file mode 100644 index 000000000..51a1e2779 --- /dev/null +++ b/src/video_core/host_shaders/antialiasing/Vulkan/vulkan_smaa_pass0_pre.frag @@ -0,0 +1,18 @@ +//? #version 450 +// SPDX-License-Identifier: Unlicense +//----------------------------------------------------------------------------- +// Edge Detection Shaders (First Pass) + +uniform vec4 i_resolution; +#define SMAA_RT_METRICS vec4(i_resolution.z, i_resolution.w, i_resolution.x, i_resolution.y) +#define SMAA_GLSL_4 +#define SMAA_PRESET_ULTRA +#define SMAA_EDT 1.0 + +layout(location = 0) in vec2 frag_tex_coord; +layout(location = 1) in vec4 offset[3]; +layout(location = 0) out vec4 color; +layout(binding = 0) uniform sampler2D color_texture; + +#define SMAA_INCLUDE_VS 0 +//#include "SMAA.hlsl" \ No newline at end of file diff --git a/src/video_core/host_shaders/antialiasing/Vulkan/vulkan_smaa_pass0_pre.vert b/src/video_core/host_shaders/antialiasing/Vulkan/vulkan_smaa_pass0_pre.vert new file mode 100644 index 000000000..4bdce548b --- /dev/null +++ b/src/video_core/host_shaders/antialiasing/Vulkan/vulkan_smaa_pass0_pre.vert @@ -0,0 +1,18 @@ +//? #version 450 +// SPDX-License-Identifier: Unlicense +//----------------------------------------------------------------------------- +// Edge Detection Shaders (First Pass) + +uniform vec4 i_resolution; +#define SMAA_RT_METRICS vec4(i_resolution.z, i_resolution.w, i_resolution.x, i_resolution.y) +#define SMAA_GLSL_4 +#define SMAA_PRESET_ULTRA +#define SMAA_EDT 1.0 + +layout(location = 0) in vec2 vert_position; +layout(location = 1) in vec2 vert_tex_coord; +layout(location = 0) out vec2 frag_tex_coord; +layout(location = 1) out vec4 offset[3]; + +#define SMAA_INCLUDE_PS 0 +//#include "SMAA.hlsl" \ No newline at end of file diff --git a/src/video_core/host_shaders/antialiasing/Vulkan/vulkan_smaa_pass1_post.frag b/src/video_core/host_shaders/antialiasing/Vulkan/vulkan_smaa_pass1_post.frag new file mode 100644 index 000000000..d407fa1a1 --- /dev/null +++ b/src/video_core/host_shaders/antialiasing/Vulkan/vulkan_smaa_pass1_post.frag @@ -0,0 +1,4 @@ +void main() { + vec4 subsampleIndices = vec4(0.0); + color = SMAABlendingWeightCalculationPS(frag_tex_coord, pixcoord, offset, color_texture, areaTex, searchTex, subsampleIndices); +} diff --git a/src/video_core/host_shaders/antialiasing/Vulkan/vulkan_smaa_pass1_post.vert b/src/video_core/host_shaders/antialiasing/Vulkan/vulkan_smaa_pass1_post.vert new file mode 100644 index 000000000..110300dbd --- /dev/null +++ b/src/video_core/host_shaders/antialiasing/Vulkan/vulkan_smaa_pass1_post.vert @@ -0,0 +1,5 @@ +void main() { + gl_Position = vec4(vert_position, 0.0, 1.0); + frag_tex_coord = vert_tex_coord; + SMAABlendingWeightCalculationVS(vert_tex_coord, pixcoord, offset); +} \ No newline at end of file diff --git a/src/video_core/host_shaders/antialiasing/Vulkan/vulkan_smaa_pass1_pre.frag b/src/video_core/host_shaders/antialiasing/Vulkan/vulkan_smaa_pass1_pre.frag new file mode 100644 index 000000000..279a85300 --- /dev/null +++ b/src/video_core/host_shaders/antialiasing/Vulkan/vulkan_smaa_pass1_pre.frag @@ -0,0 +1,21 @@ +//? #version 450 +// SPDX-License-Identifier: Unlicense +//----------------------------------------------------------------------------- +// Blending Weight Calculation Shader (Second Pass) + +uniform vec4 i_resolution; +#define SMAA_RT_METRICS vec4(i_resolution.z, i_resolution.w, i_resolution.x, i_resolution.y) +#define SMAA_GLSL_4 +#define SMAA_PRESET_ULTRA +#define SMAA_EDT 1.0 + +layout(location = 0) in vec2 frag_tex_coord; +layout(location = 1) in vec2 pixcoord; +layout(location = 2) in vec4 offset[3]; +layout(location = 0) out vec4 color; +layout(binding = 0) uniform sampler2D color_texture; +uniform sampler2D areaTex; +uniform sampler2D searchTex; + +#define SMAA_INCLUDE_VS 0 +//#include "SMAA.hlsl" \ No newline at end of file diff --git a/src/video_core/host_shaders/antialiasing/Vulkan/vulkan_smaa_pass1_pre.vert b/src/video_core/host_shaders/antialiasing/Vulkan/vulkan_smaa_pass1_pre.vert new file mode 100644 index 000000000..f75990c39 --- /dev/null +++ b/src/video_core/host_shaders/antialiasing/Vulkan/vulkan_smaa_pass1_pre.vert @@ -0,0 +1,19 @@ +//? #version 450 +// SPDX-License-Identifier: Unlicense +//----------------------------------------------------------------------------- +// Blending Weight Calculation Shader (Second Pass) + +uniform vec4 i_resolution; +#define SMAA_RT_METRICS vec4(i_resolution.z, i_resolution.w, i_resolution.x, i_resolution.y) +#define SMAA_GLSL_4 +#define SMAA_PRESET_ULTRA +#define SMAA_EDT 1.0 + +layout(location = 0) in vec2 vert_position; +layout(location = 1) in vec2 vert_tex_coord; +layout(location = 0) out vec2 frag_tex_coord; +layout(location = 1) out vec2 pixcoord; +layout(location = 2) out vec4 offset[3]; + +#define SMAA_INCLUDE_PS 0 +//#include "SMAA.hlsl" \ No newline at end of file diff --git a/src/video_core/host_shaders/antialiasing/Vulkan/vulkan_smaa_pass2_post.frag b/src/video_core/host_shaders/antialiasing/Vulkan/vulkan_smaa_pass2_post.frag new file mode 100644 index 000000000..8341214c5 --- /dev/null +++ b/src/video_core/host_shaders/antialiasing/Vulkan/vulkan_smaa_pass2_post.frag @@ -0,0 +1,11 @@ +vec3 LinearTosRGB(vec3 c) { + return mix(c * 12.92, 1.055 * pow(c, vec3(1.0/2.4)) - 0.055, step(0.0031308, c)); +} + +void main() { + vec4 pixel = SMAANeighborhoodBlendingPS(frag_tex_coord, offset, SMAA_Input, color_texture); + if (convert_colors == 2){ + pixel = vec4(LinearTosRGB(pixel.rgb), pixel.a); + } + color = pixel; +} diff --git a/src/video_core/host_shaders/antialiasing/Vulkan/vulkan_smaa_pass2_post.vert b/src/video_core/host_shaders/antialiasing/Vulkan/vulkan_smaa_pass2_post.vert new file mode 100644 index 000000000..6a3d9ea1e --- /dev/null +++ b/src/video_core/host_shaders/antialiasing/Vulkan/vulkan_smaa_pass2_post.vert @@ -0,0 +1,5 @@ +void main() { + gl_Position = vec4(vert_position, 0.0, 1.0); + frag_tex_coord = vert_tex_coord; + SMAANeighborhoodBlendingVS(vert_tex_coord, offset); +} \ No newline at end of file diff --git a/src/video_core/host_shaders/antialiasing/Vulkan/vulkan_smaa_pass2_pre.frag b/src/video_core/host_shaders/antialiasing/Vulkan/vulkan_smaa_pass2_pre.frag new file mode 100644 index 000000000..edd0d1137 --- /dev/null +++ b/src/video_core/host_shaders/antialiasing/Vulkan/vulkan_smaa_pass2_pre.frag @@ -0,0 +1,18 @@ +//? #version 450 +// SPDX-License-Identifier: Unlicense +//----------------------------------------------------------------------------- +// Neighborhood Blending Shader (Third Pass) + +uniform vec4 i_resolution; +uniform int convert_colors; +#define SMAA_RT_METRICS vec4(i_resolution.z, i_resolution.w, i_resolution.x, i_resolution.y) +#define SMAA_GLSL_4 + +layout(location = 0) in vec2 frag_tex_coord; +layout(location = 1) in vec4 offset; +layout(location = 0) out vec4 color; +layout(binding = 0) uniform sampler2D color_texture; +uniform sampler2D SMAA_Input; + +#define SMAA_INCLUDE_VS 0 +//#include "SMAA.hlsl" \ No newline at end of file diff --git a/src/video_core/host_shaders/antialiasing/Vulkan/vulkan_smaa_pass2_pre.vert b/src/video_core/host_shaders/antialiasing/Vulkan/vulkan_smaa_pass2_pre.vert new file mode 100644 index 000000000..24baa2cc6 --- /dev/null +++ b/src/video_core/host_shaders/antialiasing/Vulkan/vulkan_smaa_pass2_pre.vert @@ -0,0 +1,16 @@ +//? #version 450 +// SPDX-License-Identifier: Unlicense +//----------------------------------------------------------------------------- +// Neighborhood Blending Shader (Third Pass) + +uniform vec4 i_resolution; +#define SMAA_RT_METRICS vec4(i_resolution.z, i_resolution.w, i_resolution.x, i_resolution.y) +#define SMAA_GLSL_4 + +layout(location = 0) in vec2 vert_position; +layout(location = 1) in vec2 vert_tex_coord; +layout(location = 0) out vec2 frag_tex_coord; +layout(location = 1) out vec4 offset; + +#define SMAA_INCLUDE_PS 0 +//#include "SMAA.hlsl" \ No newline at end of file diff --git a/src/video_core/renderer_opengl/renderer_opengl.cpp b/src/video_core/renderer_opengl/renderer_opengl.cpp index f7639b8cc..44f2fd85e 100644 --- a/src/video_core/renderer_opengl/renderer_opengl.cpp +++ b/src/video_core/renderer_opengl/renderer_opengl.cpp @@ -25,21 +25,21 @@ #include "video_core/host_shaders/opengl_simple_present_frag.h" #include "video_core/host_shaders/opengl_simple_present_vert.h" -#include "video_core/host_shaders/antialiasing/opengl_fxaa_frag.h" -#include "video_core/host_shaders/antialiasing/opengl_fxaa_vert.h" -#include "video_core/host_shaders/antialiasing/opengl_smaa_pass0_pre_frag.h" -#include "video_core/host_shaders/antialiasing/opengl_smaa_pass0_pre_vert.h" -#include "video_core/host_shaders/antialiasing/opengl_smaa_pass0_post_frag.h" -#include "video_core/host_shaders/antialiasing/opengl_smaa_pass0_post_vert.h" -#include "video_core/host_shaders/antialiasing/opengl_smaa_pass1_pre_frag.h" -#include "video_core/host_shaders/antialiasing/opengl_smaa_pass1_pre_vert.h" -#include "video_core/host_shaders/antialiasing/opengl_smaa_pass1_post_frag.h" -#include "video_core/host_shaders/antialiasing/opengl_smaa_pass1_post_vert.h" -#include "video_core/host_shaders/antialiasing/opengl_smaa_pass2_pre_frag.h" -#include "video_core/host_shaders/antialiasing/opengl_smaa_pass2_pre_vert.h" -#include "video_core/host_shaders/antialiasing/opengl_smaa_pass2_post_frag.h" -#include "video_core/host_shaders/antialiasing/opengl_smaa_pass2_post_vert.h" -#include "video_core/host_shaders/antialiasing/smaa_hlsl.h" +#include "video_core/host_shaders/antialiasing/OpenGL/opengl_fxaa_frag.h" +#include "video_core/host_shaders/antialiasing/OpenGL/opengl_fxaa_vert.h" +#include "video_core/host_shaders/antialiasing/OpenGL/opengl_smaa_pass0_pre_frag.h" +#include "video_core/host_shaders/antialiasing/OpenGL/opengl_smaa_pass0_pre_vert.h" +#include "video_core/host_shaders/antialiasing/OpenGL/opengl_smaa_pass0_post_frag.h" +#include "video_core/host_shaders/antialiasing/OpenGL/opengl_smaa_pass0_post_vert.h" +#include "video_core/host_shaders/antialiasing/OpenGL/opengl_smaa_pass1_pre_frag.h" +#include "video_core/host_shaders/antialiasing/OpenGL/opengl_smaa_pass1_pre_vert.h" +#include "video_core/host_shaders/antialiasing/OpenGL/opengl_smaa_pass1_post_frag.h" +#include "video_core/host_shaders/antialiasing/OpenGL/opengl_smaa_pass1_post_vert.h" +#include "video_core/host_shaders/antialiasing/OpenGL/opengl_smaa_pass2_pre_frag.h" +#include "video_core/host_shaders/antialiasing/OpenGL/opengl_smaa_pass2_pre_vert.h" +#include "video_core/host_shaders/antialiasing/OpenGL/opengl_smaa_pass2_post_frag.h" +#include "video_core/host_shaders/antialiasing/OpenGL/opengl_smaa_pass2_post_vert.h" +#include "video_core/host_shaders/antialiasing/OpenGL/opengl_smaa_hlsl.h" #include "video_core/host_shaders/antialiasing/AreaTex.h" #include "video_core/host_shaders/antialiasing/SearchTex.h" #include "video_core/host_shaders/scaling/opengl_area_sampling_frag.h" @@ -501,31 +501,31 @@ void RendererOpenGL::ReloadShader(Settings::StereoRenderOption render_3d) { std::string SMAA_PASS_0_shader_frag_data = fragment_shader_precision_OES; SMAA_PASS_0_shader_frag_data += HostShaders::OPENGL_SMAA_PASS0_PRE_FRAG; - SMAA_PASS_0_shader_frag_data += HostShaders::SMAA_HLSL; + SMAA_PASS_0_shader_frag_data += HostShaders::OPENGL_SMAA_HLSL; SMAA_PASS_0_shader_frag_data += HostShaders::OPENGL_SMAA_PASS0_POST_FRAG; std::string SMAA_PASS_0_shader_vert_data; SMAA_PASS_0_shader_vert_data += HostShaders::OPENGL_SMAA_PASS0_PRE_VERT; - SMAA_PASS_0_shader_vert_data += HostShaders::SMAA_HLSL; + SMAA_PASS_0_shader_vert_data += HostShaders::OPENGL_SMAA_HLSL; SMAA_PASS_0_shader_vert_data += HostShaders::OPENGL_SMAA_PASS0_POST_VERT; SMAA_PASS_0_shader.Create(SMAA_PASS_0_shader_vert_data, SMAA_PASS_0_shader_frag_data); std::string SMAA_PASS_1_shader_frag_data = fragment_shader_precision_OES; SMAA_PASS_1_shader_frag_data += HostShaders::OPENGL_SMAA_PASS1_PRE_FRAG; - SMAA_PASS_1_shader_frag_data += HostShaders::SMAA_HLSL; + SMAA_PASS_1_shader_frag_data += HostShaders::OPENGL_SMAA_HLSL; SMAA_PASS_1_shader_frag_data += HostShaders::OPENGL_SMAA_PASS1_POST_FRAG; std::string SMAA_PASS_1_shader_vert_data; SMAA_PASS_1_shader_vert_data += HostShaders::OPENGL_SMAA_PASS1_PRE_VERT; - SMAA_PASS_1_shader_vert_data += HostShaders::SMAA_HLSL; + SMAA_PASS_1_shader_vert_data += HostShaders::OPENGL_SMAA_HLSL; SMAA_PASS_1_shader_vert_data += HostShaders::OPENGL_SMAA_PASS1_POST_VERT; SMAA_PASS_1_shader.Create(SMAA_PASS_1_shader_vert_data, SMAA_PASS_1_shader_frag_data); std::string SMAA_PASS_2_shader_frag_data = fragment_shader_precision_OES; SMAA_PASS_2_shader_frag_data += HostShaders::OPENGL_SMAA_PASS2_PRE_FRAG; - SMAA_PASS_2_shader_frag_data += HostShaders::SMAA_HLSL; + SMAA_PASS_2_shader_frag_data += HostShaders::OPENGL_SMAA_HLSL; SMAA_PASS_2_shader_frag_data += HostShaders::OPENGL_SMAA_PASS2_POST_FRAG; std::string SMAA_PASS_2_shader_vert_data; SMAA_PASS_2_shader_vert_data += HostShaders::OPENGL_SMAA_PASS2_PRE_VERT; - SMAA_PASS_2_shader_vert_data += HostShaders::SMAA_HLSL; + SMAA_PASS_2_shader_vert_data += HostShaders::OPENGL_SMAA_HLSL; SMAA_PASS_2_shader_vert_data += HostShaders::OPENGL_SMAA_PASS2_POST_VERT; SMAA_PASS_2_shader.Create(SMAA_PASS_2_shader_vert_data, SMAA_PASS_2_shader_frag_data); diff --git a/src/video_core/renderer_vulkan/renderer_vulkan.h b/src/video_core/renderer_vulkan/renderer_vulkan.h index 14c9bd34f..96bf3d9c3 100644 --- a/src/video_core/renderer_vulkan/renderer_vulkan.h +++ b/src/video_core/renderer_vulkan/renderer_vulkan.h @@ -62,8 +62,12 @@ struct PresentUniformData { int screen_id_r = 0; int layer = 0; int reverse_interlaced = 0; + int convert_colors; + int areatex; + int searchtex; + int smaa_input; }; -static_assert(sizeof(PresentUniformData) == 112, +static_assert(sizeof(PresentUniformData) == 128, "PresentUniformData does not structure in shader!"); class RendererVulkan : public VideoCore::RendererBase {