diff --git a/blurs/blur9fast-horizontal.slang b/blurs/blur9fast-horizontal.slang new file mode 100644 index 0000000..581d2dd --- /dev/null +++ b/blurs/blur9fast-horizontal.slang @@ -0,0 +1,98 @@ +#version 450 + +layout(push_constant) uniform Push +{ + vec4 SourceSize; + vec4 OriginalSize; + vec4 OutputSize; + uint FrameCount; +} params; + +layout(std140, set = 0, binding = 0) uniform UBO +{ + mat4 MVP; +} global; + +///////////////////////////////// MIT LICENSE //////////////////////////////// + +// Copyright (C) 2014 TroggleMonkey +// +// 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. + + +///////////////////////////// SETTINGS MANAGEMENT //////////////////////////// + +// PASS SETTINGS: +// gamma-management.h needs to know what kind of pipeline we're using and +// what pass this is in that pipeline. This will become obsolete if/when we +// can #define things like this in the .cgp preset file. +//#define GAMMA_ENCODE_EVERY_FBO +//#define FIRST_PASS +//#define LAST_PASS +//#define SIMULATE_CRT_ON_LCD +//#define SIMULATE_GBA_ON_LCD +//#define SIMULATE_LCD_ON_CRT +//#define SIMULATE_GBA_ON_CRT + + +////////////////////////////////// INCLUDES ////////////////////////////////// + +// #included by vertex shader: +//#include "../include/gamma-management.h" +//#include "../include/blur-functions.h" +#include "../crt/shaders/crt-royale/src/includes.h" + +#pragma stage vertex +layout(location = 0) in vec4 Position; +layout(location = 1) in vec2 TexCoord; +layout(location = 0) out vec2 tex_uv; +layout(location = 1) out vec2 blur_dxdy; + +void main() +{ + gl_Position = global.MVP * Position; + tex_uv = TexCoord; + + // Get the uv sample distance between output pixels. Blurs are not generic + // Gaussian resizers, and correct blurs require: + // 1.) params.OutputSize.xy == params.SourceSize.xy * 2^m, where m is an integer <= 0. + // 2.) mipmap_inputN = "true" for this pass in preset if m != 0 + // 3.) filter_linearN = "true" except for 1x scale nearest neighbor blurs + // Gaussian resizers would upsize using the distance between input texels + // (not output pixels), but we avoid this and consistently blur at the + // destination size. Otherwise, combining statically calculated weights + // with bilinear sample exploitation would result in terrible artifacts. + const vec2 dxdy_scale = params.SourceSize.xy * params.OutputSize.zw; + const vec2 dxdy = dxdy_scale * params.SourceSize.zw; + // This blur is horizontal-only, so zero out the vertical offset: + blur_dxdy = vec2(dxdy.x, 0.0); +} + +#pragma stage fragment +layout(location = 0) in vec2 tex_uv; +layout(location = 1) in vec2 blur_dxdy; +layout(location = 0) out vec4 FragColor; +layout(set = 0, binding = 2) uniform sampler2D Source; + +void main() +{ + vec3 color = tex2Dblur9fast(Source, tex_uv, blur_dxdy); + // Encode and output the blurred image: + FragColor = vec4(1.0);//encode_output(vec4(color, 1.0)); +} \ No newline at end of file diff --git a/blurs/blur9fast-vertical.slang b/blurs/blur9fast-vertical.slang new file mode 100644 index 0000000..f9428d6 --- /dev/null +++ b/blurs/blur9fast-vertical.slang @@ -0,0 +1,98 @@ +#version 450 + +layout(push_constant) uniform Push +{ + vec4 SourceSize; + vec4 OriginalSize; + vec4 OutputSize; + uint FrameCount; +} params; + +layout(std140, set = 0, binding = 0) uniform UBO +{ + mat4 MVP; +} global; + +///////////////////////////////// MIT LICENSE //////////////////////////////// + +// Copyright (C) 2014 TroggleMonkey +// +// 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. + + +///////////////////////////// SETTINGS MANAGEMENT //////////////////////////// + +// PASS SETTINGS: +// gamma-management.h needs to know what kind of pipeline we're using and +// what pass this is in that pipeline. This will become obsolete if/when we +// can #define things like this in the .cgp preset file. +//#define GAMMA_ENCODE_EVERY_FBO +//#define FIRST_PASS +//#define LAST_PASS +//#define SIMULATE_CRT_ON_LCD +//#define SIMULATE_GBA_ON_LCD +//#define SIMULATE_LCD_ON_CRT +//#define SIMULATE_GBA_ON_CRT + + +////////////////////////////////// INCLUDES ////////////////////////////////// + +// #included by vertex shader: +//#include "../include/gamma-management.h" +//#include "../include/blur-functions.h" +#include "../crt/shaders/crt-royale/src/includes.h" + +#pragma stage vertex +layout(location = 0) in vec4 Position; +layout(location = 1) in vec2 TexCoord; +layout(location = 0) out vec2 tex_uv; +layout(location = 1) out vec2 blur_dxdy; + +void main() +{ + gl_Position = global.MVP * Position; + tex_uv = TexCoord; + + // Get the uv sample distance between output pixels. Blurs are not generic + // Gaussian resizers, and correct blurs require: + // 1.) params.OutputSize.xy == params.SourceSize.xy * 2^m, where m is an integer <= 0. + // 2.) mipmap_inputN = "true" for this pass in preset if m != 0 + // 3.) filter_linearN = "true" except for 1x scale nearest neighbor blurs + // Gaussian resizers would upsize using the distance between input texels + // (not output pixels), but we avoid this and consistently blur at the + // destination size. Otherwise, combining statically calculated weights + // with bilinear sample exploitation would result in terrible artifacts. + const vec2 dxdy_scale = params.SourceSize.xy * params.OutputSize.zw; + const vec2 dxdy = dxdy_scale * params.SourceSize.zw; + // This blur is vertical-only, so zero out the vertical offset: + blur_dxdy = vec2(0.0, dxdy.y); +} + +#pragma stage fragment +layout(location = 0) in vec2 tex_uv; +layout(location = 1) in vec2 blur_dxdy; +layout(location = 0) out vec4 FragColor; +layout(set = 0, binding = 2) uniform sampler2D Source; + +void main() +{ + vec3 color = tex2Dblur9fast(Source, tex_uv, blur_dxdy); + // Encode and output the blurred image: + FragColor = vec4(1.0);//encode_output(vec4(color, 1.0)); +} \ No newline at end of file diff --git a/crt/crt-royale-test.slangp b/crt/crt-royale-test.slangp new file mode 100644 index 0000000..0104a01 --- /dev/null +++ b/crt/crt-royale-test.slangp @@ -0,0 +1,92 @@ +# IMPORTANT: +# Shader passes need to know details about the image in the mask_texture LUT +# files, so set the following constants in user-preset-constants.h accordingly: +# 1.) mask_triads_per_tile = (number of horizontal triads in mask texture LUT's) +# 2.) mask_texture_small_size = (texture size of mask*texture_small LUT's) +# 3.) mask_texture_large_size = (texture size of mask*texture_large LUT's) +# 4.) mask_grille_avg_color = (avg. brightness of mask_grille_texture* LUT's, in [0, 1]) +# 5.) mask_slot_avg_color = (avg. brightness of mask_slot_texture* LUT's, in [0, 1]) +# 6.) mask_shadow_avg_color = (avg. brightness of mask_shadow_texture* LUT's, in [0, 1]) +# Shader passes also need to know certain scales set in this preset, but their +# compilation model doesn't currently allow the preset file to tell them. Make +# sure to set the following constants in user-preset-constants.h accordingly too: +# 1.) bloom_approx_scale_x = scale_x2 +# 2.) mask_resize_viewport_scale = vec2(scale_x6, scale_y5) +# Finally, shader passes need to know the value of geom_max_aspect_ratio used to +# calculate scale_y5 (among other values): +# 1.) geom_max_aspect_ratio = (geom_max_aspect_ratio used to calculate scale_y5) + +shaders = "1"//"12" + +# Set an identifier, filename, and sampling traits for the phosphor mask texture. +# Load an aperture grille, slot mask, and an EDP shadow mask, and load a small +# non-mipmapped version and a large mipmapped version. +# TODO: Test masks in other directories. +textures = "mask_grille_texture_small;mask_grille_texture_large;mask_slot_texture_small;mask_slot_texture_large;mask_shadow_texture_small;mask_shadow_texture_large" +mask_grille_texture_small = "shaders/crt-royale/TileableLinearApertureGrille15Wide8And5d5SpacingResizeTo64.png" +mask_grille_texture_large = "shaders/crt-royale/TileableLinearApertureGrille15Wide8And5d5Spacing.png" +mask_slot_texture_small = "shaders/crt-royale/TileableLinearSlotMaskTall15Wide9And4d5Horizontal9d14VerticalSpacingResizeTo64.png" +mask_slot_texture_large = "shaders/crt-royale/TileableLinearSlotMaskTall15Wide9And4d5Horizontal9d14VerticalSpacing.png" +mask_shadow_texture_small = "shaders/crt-royale/TileableLinearShadowMaskEDPResizeTo64.png" +mask_shadow_texture_large = "shaders/crt-royale/TileableLinearShadowMaskEDP.png" +mask_grille_texture_small_wrap_mode = "repeat" +mask_grille_texture_large_wrap_mode = "repeat" +mask_slot_texture_small_wrap_mode = "repeat" +mask_slot_texture_large_wrap_mode = "repeat" +mask_shadow_texture_small_wrap_mode = "repeat" +mask_shadow_texture_large_wrap_mode = "repeat" +mask_grille_texture_small_linear = "true" +mask_grille_texture_large_linear = "true" +mask_slot_texture_small_linear = "true" +mask_slot_texture_large_linear = "true" +mask_shadow_texture_small_linear = "true" +mask_shadow_texture_large_linear = "true" +mask_grille_texture_small_mipmap = "false" # Mipmapping causes artifacts with manually resized masks without tex2Dlod +mask_grille_texture_large_mipmap = "true" # Essential for hardware-resized masks +mask_slot_texture_small_mipmap = "false" # Mipmapping causes artifacts with manually resized masks without tex2Dlod +mask_slot_texture_large_mipmap = "true" # Essential for hardware-resized masks +mask_shadow_texture_small_mipmap = "false" # Mipmapping causes artifacts with manually resized masks without tex2Dlod +mask_shadow_texture_large_mipmap = "true" # Essential for hardware-resized masks + + +# Pass5: Lanczos-resize the phosphor mask vertically. Set the absolute +# scale_x5 == mask_texture_small_size.x (see IMPORTANT above). Larger scales +# will blur, and smaller scales could get nasty. The vertical size must be +# based on the viewport size and calculated carefully to avoid artifacts later. +# First calculate the minimum number of mask tiles we need to draw. +# Since curvature is computed after the scanline masking pass: +# num_resized_mask_tiles = 2.0; +# If curvature were computed in the scanline masking pass (it's not): +# max_mask_texel_border = ~3.0 * (1/3.0 + 4.0*sqrt(2.0) + 0.5 + 1.0); +# max_mask_tile_border = max_mask_texel_border/ +# (min_resized_phosphor_triad_size * mask_triads_per_tile); +# num_resized_mask_tiles = max(2.0, 1.0 + max_mask_tile_border * 2.0); +# At typical values (triad_size >= 2.0, mask_triads_per_tile == 8): +# num_resized_mask_tiles = ~3.8 +# Triad sizes are given in horizontal terms, so we need geom_max_aspect_ratio +# to relate them to vertical resolution. The widest we expect is: +# geom_max_aspect_ratio = 4.0/3.0 # Note: Shader passes need to know this! +# The fewer triads we tile across the screen, the larger each triad will be as a +# fraction of the viewport size, and the larger scale_y5 must be to draw a full +# num_resized_mask_tiles. Therefore, we must decide the smallest number of +# triads we'll guarantee can be displayed on screen. We'll set this according +# to 3-pixel triads at 768p resolution (the lowest anyone's likely to use): +# min_allowed_viewport_triads = 768.0*geom_max_aspect_ratio / 3.0 = 341.333333 +# Now calculate the viewport scale that ensures we can draw resized_mask_tiles: +# min_scale_x = resized_mask_tiles * mask_triads_per_tile / +# min_allowed_viewport_triads +# scale_y5 = geom_max_aspect_ratio * min_scale_x +# # Some code might depend on equal scales: +# scale_x6 = scale_y5 +# Given our default geom_max_aspect_ratio and min_allowed_viewport_triads: +# scale_y5 = 4.0/3.0 * 2.0/(341.33333 / 8.0) = 0.0625 +# IMPORTANT: The scales MUST be calculated in this way. If you wish to change +# geom_max_aspect_ratio, update that constant in user-preset-constants.h! +shader0 = "shaders/crt-royale/src/crt-royale-mask-resize-vertical.slang" +filter_linear0 = "true" +scale_type_x0 = "absolute" +scale_x0 = "64" +scale_type_y0 = "viewport" +scale_y0 = "0.0625" # Safe for >= 341.333 horizontal triads at viewport size +#srgb_framebuffer0 = "false" # mask_texture is already assumed linear + diff --git a/crt/crt-royale.slangp b/crt/crt-royale.slangp new file mode 100644 index 0000000..dea143f --- /dev/null +++ b/crt/crt-royale.slangp @@ -0,0 +1,206 @@ +# IMPORTANT: +# Shader passes need to know details about the image in the mask_texture LUT +# files, so set the following constants in user-preset-constants.h accordingly: +# 1.) mask_triads_per_tile = (number of horizontal triads in mask texture LUT's) +# 2.) mask_texture_small_size = (texture size of mask*texture_small LUT's) +# 3.) mask_texture_large_size = (texture size of mask*texture_large LUT's) +# 4.) mask_grille_avg_color = (avg. brightness of mask_grille_texture* LUT's, in [0, 1]) +# 5.) mask_slot_avg_color = (avg. brightness of mask_slot_texture* LUT's, in [0, 1]) +# 6.) mask_shadow_avg_color = (avg. brightness of mask_shadow_texture* LUT's, in [0, 1]) +# Shader passes also need to know certain scales set in this preset, but their +# compilation model doesn't currently allow the preset file to tell them. Make +# sure to set the following constants in user-preset-constants.h accordingly too: +# 1.) bloom_approx_scale_x = scale_x2 +# 2.) mask_resize_viewport_scale = vec2(scale_x6, scale_y5) +# Finally, shader passes need to know the value of geom_max_aspect_ratio used to +# calculate scale_y5 (among other values): +# 1.) geom_max_aspect_ratio = (geom_max_aspect_ratio used to calculate scale_y5) + +shaders = "5"//"12" + +# Set an identifier, filename, and sampling traits for the phosphor mask texture. +# Load an aperture grille, slot mask, and an EDP shadow mask, and load a small +# non-mipmapped version and a large mipmapped version. +# TODO: Test masks in other directories. +textures = "mask_grille_texture_small;mask_grille_texture_large;mask_slot_texture_small;mask_slot_texture_large;mask_shadow_texture_small;mask_shadow_texture_large" +mask_grille_texture_small = "shaders/crt-royale/TileableLinearApertureGrille15Wide8And5d5SpacingResizeTo64.png" +mask_grille_texture_large = "shaders/crt-royale/TileableLinearApertureGrille15Wide8And5d5Spacing.png" +mask_slot_texture_small = "shaders/crt-royale/TileableLinearSlotMaskTall15Wide9And4d5Horizontal9d14VerticalSpacingResizeTo64.png" +mask_slot_texture_large = "shaders/crt-royale/TileableLinearSlotMaskTall15Wide9And4d5Horizontal9d14VerticalSpacing.png" +mask_shadow_texture_small = "shaders/crt-royale/TileableLinearShadowMaskEDPResizeTo64.png" +mask_shadow_texture_large = "shaders/crt-royale/TileableLinearShadowMaskEDP.png" +mask_grille_texture_small_wrap_mode = "repeat" +mask_grille_texture_large_wrap_mode = "repeat" +mask_slot_texture_small_wrap_mode = "repeat" +mask_slot_texture_large_wrap_mode = "repeat" +mask_shadow_texture_small_wrap_mode = "repeat" +mask_shadow_texture_large_wrap_mode = "repeat" +mask_grille_texture_small_linear = "true" +mask_grille_texture_large_linear = "true" +mask_slot_texture_small_linear = "true" +mask_slot_texture_large_linear = "true" +mask_shadow_texture_small_linear = "true" +mask_shadow_texture_large_linear = "true" +mask_grille_texture_small_mipmap = "false" # Mipmapping causes artifacts with manually resized masks without tex2Dlod +mask_grille_texture_large_mipmap = "true" # Essential for hardware-resized masks +mask_slot_texture_small_mipmap = "false" # Mipmapping causes artifacts with manually resized masks without tex2Dlod +mask_slot_texture_large_mipmap = "true" # Essential for hardware-resized masks +mask_shadow_texture_small_mipmap = "false" # Mipmapping causes artifacts with manually resized masks without tex2Dlod +mask_shadow_texture_large_mipmap = "true" # Essential for hardware-resized masks + + +# Pass0: Linearize the input based on CRT gamma and bob interlaced fields. +# (Bobbing ensures we can immediately blur without getting artifacts.) +shader0 = "shaders/crt-royale/src/crt-royale-first-pass-linearize-crt-gamma-bob-fields.slang" +alias0 = "ORIG_LINEARIZED" +filter_linear0 = "false" +scale_type0 = "source" +scale0 = "1.0" +srgb_framebuffer0 = "true" + +# Pass1: Resample interlaced (and misconverged) scanlines vertically. +# Separating vertical/horizontal scanline sampling is faster: It lets us +# consider more scanlines while calculating weights for fewer pixels, and +# it reduces our samples from vertical*horizontal to vertical+horizontal. +# This has to come right after ORIG_LINEARIZED, because there's no +# "original_source" scale_type we can use later. +shader1 = "shaders/crt-royale/src/crt-royale-scanlines-vertical-interlacing.slang" +alias1 = "VERTICAL_SCANLINES" +filter_linear1 = "true" +scale_type_x1 = "source" +scale_x1 = "1.0" +scale_type_y1 = "viewport" +scale_y1 = "1.0" +#float_framebuffer1 = "true" +srgb_framebuffer1 = "true" + +# Pass2: Do a small resize blur of ORIG_LINEARIZED at an absolute size, and +# account for convergence offsets. We want to blur a predictable portion of the +# screen to match the phosphor bloom, and absolute scale works best for +# reliable results with a fixed-size bloom. Picking a scale is tricky: +# a.) 400x300 is a good compromise for the "fake-bloom" version: It's low enough +# to blur high-res/interlaced sources but high enough that resampling +# doesn't smear low-res sources too much. +# b.) 320x240 works well for the "real bloom" version: It's 1-1.5% faster, and +# the only noticeable visual difference is a larger halation spread (which +# may be a good thing for people who like to crank it up). +# Note the 4:3 aspect ratio assumes the input has cropped geom_overscan (so it's +# *intended* for an ~4:3 aspect ratio). +shader2 = "shaders/crt-royale/src/crt-royale-bloom-approx.slang" +alias2 = "BLOOM_APPROX" +filter_linear2 = "true" +scale_type2 = "absolute" +scale_x2 = "320" +scale_y2 = "240" +srgb_framebuffer2 = "true" + +# Pass3: Vertically blur the input for halation and refractive diffusion. +# Base this on BLOOM_APPROX: This blur should be small and fast, and blurring +# a constant portion of the screen is probably physically correct if the +# viewport resolution is proportional to the simulated CRT size. +shader3 = "../blurs/blur9fast-vertical.slang" +filter_linear3 = "true" +scale_type3 = "source" +scale3 = "1.0" +srgb_framebuffer3 = "true" + +# Pass4: Horizontally blur the input for halation and refractive diffusion. +# Note: Using a one-pass 9x9 blur is about 1% slower. +shader4 = "../blurs/blur9fast-horizontal.slang" +alias4 = "HALATION_BLUR" +filter_linear4 = "true" +scale_type4 = "source" +scale4 = "1.0" +srgb_framebuffer4 = "true" + +# Pass5: Lanczos-resize the phosphor mask vertically. Set the absolute +# scale_x5 == mask_texture_small_size.x (see IMPORTANT above). Larger scales +# will blur, and smaller scales could get nasty. The vertical size must be +# based on the viewport size and calculated carefully to avoid artifacts later. +# First calculate the minimum number of mask tiles we need to draw. +# Since curvature is computed after the scanline masking pass: +# num_resized_mask_tiles = 2.0; +# If curvature were computed in the scanline masking pass (it's not): +# max_mask_texel_border = ~3.0 * (1/3.0 + 4.0*sqrt(2.0) + 0.5 + 1.0); +# max_mask_tile_border = max_mask_texel_border/ +# (min_resized_phosphor_triad_size * mask_triads_per_tile); +# num_resized_mask_tiles = max(2.0, 1.0 + max_mask_tile_border * 2.0); +# At typical values (triad_size >= 2.0, mask_triads_per_tile == 8): +# num_resized_mask_tiles = ~3.8 +# Triad sizes are given in horizontal terms, so we need geom_max_aspect_ratio +# to relate them to vertical resolution. The widest we expect is: +# geom_max_aspect_ratio = 4.0/3.0 # Note: Shader passes need to know this! +# The fewer triads we tile across the screen, the larger each triad will be as a +# fraction of the viewport size, and the larger scale_y5 must be to draw a full +# num_resized_mask_tiles. Therefore, we must decide the smallest number of +# triads we'll guarantee can be displayed on screen. We'll set this according +# to 3-pixel triads at 768p resolution (the lowest anyone's likely to use): +# min_allowed_viewport_triads = 768.0*geom_max_aspect_ratio / 3.0 = 341.333333 +# Now calculate the viewport scale that ensures we can draw resized_mask_tiles: +# min_scale_x = resized_mask_tiles * mask_triads_per_tile / +# min_allowed_viewport_triads +# scale_y5 = geom_max_aspect_ratio * min_scale_x +# # Some code might depend on equal scales: +# scale_x6 = scale_y5 +# Given our default geom_max_aspect_ratio and min_allowed_viewport_triads: +# scale_y5 = 4.0/3.0 * 2.0/(341.33333 / 8.0) = 0.0625 +# IMPORTANT: The scales MUST be calculated in this way. If you wish to change +# geom_max_aspect_ratio, update that constant in user-preset-constants.h! +shader5 = "shaders/crt-royale/src/crt-royale-mask-resize-vertical.slang" +filter_linear5 = "true" +scale_type_x5 = "absolute" +scale_x5 = "64" +scale_type_y5 = "viewport" +scale_y5 = "0.0625" # Safe for >= 341.333 horizontal triads at viewport size +#srgb_framebuffer5 = "false" # mask_texture is already assumed linear + +# Pass6: Lanczos-resize the phosphor mask horizontally. scale_x6 = scale_y5. +# TODO: Check again if the shaders actually require equal scales. +shader6 = "shaders/crt-royale/src/crt-royale-mask-resize-horizontal.slang" +alias6 = "MASK_RESIZE" +filter_linear6 = "false" +scale_type_x6 = "viewport" +scale_x6 = "0.0625" +scale_type_y6 = "source" +scale_y6 = "1.0" +#srgb_framebuffer6 = "false" # mask_texture is already assumed linear + +# Pass7: Resample (misconverged) scanlines horizontally, apply halation, and +# apply the phosphor mask. +shader7 = "shaders/crt-royale/src/crt-royale-scanlines-horizontal-apply-mask.slang" +alias7 = "MASKED_SCANLINES" +filter_linear7 = "true" # This could just as easily be nearest neighbor. +scale_type7 = "viewport" +scale7 = "1.0" +#float_framebuffer7 = "true" +srgb_framebuffer7 = "true" + +# Pass 8: Compute a brightpass. This will require reading the final mask. +shader8 = "shaders/crt-royale/src/crt-royale-brightpass.slang" +alias8 = "BRIGHTPASS" +filter_linear8 = "true" # This could just as easily be nearest neighbor. +scale_type8 = "viewport" +scale8 = "1.0" +srgb_framebuffer8 = "true" + +# Pass 9: Blur the brightpass vertically +shader9 = "shaders/crt-royale/src/crt-royale-bloom-vertical.slang" +filter_linear9 = "true" # This could just as easily be nearest neighbor. +scale_type9 = "source" +scale9 = "1.0" +srgb_framebuffer9 = "true" + +# Pass 10: Blur the brightpass horizontally and combine it with the dimpass: +shader10 = "shaders/crt-royale/src/crt-royale-bloom-horizontal-reconstitute.slang" +filter_linear10 = "true" +scale_type10 = "source" +scale10 = "1.0" +srgb_framebuffer10 = "true" + +# Pass 11: Compute curvature/AA: +shader11 = "shaders/crt-royale/src/crt-royale-geometry-aa-last-pass.slang" +filter_linear11 = "true" +scale_type11 = "viewport" +mipmap_input11 = "true" +texture_wrap_mode11 = "clamp_to_edge" + diff --git a/crt/shaders/crt-royale/src/blur-functions.h b/crt/shaders/crt-royale/src/blur-functions.h index 25b83dc..79e7285 100644 --- a/crt/shaders/crt-royale/src/blur-functions.h +++ b/crt/shaders/crt-royale/src/blur-functions.h @@ -1,4 +1,5 @@ -#define BLUR_FUNCTIONS +#ifndef BLUR_FUNCTIONS_H +#define BLUR_FUNCTIONS_H ///////////////////////////////// MIT LICENSE //////////////////////////////// @@ -278,4 +279,42 @@ vec3 tex2Dblur3x3resize(const sampler2D texture, const vec2 tex_uv, const vec2 dxdy) { return tex2Dblur3x3resize(texture, tex_uv, dxdy, blur3_std_dev); -} \ No newline at end of file +} + +vec3 tex2Dblur9fast(const sampler2D tex, const vec2 tex_uv, + const vec2 dxdy, const float sigma) +{ + // Requires: Same as tex2Dblur11() + // Returns: A 1D 9x Gaussian blurred texture lookup using 1 nearest + // neighbor and 4 linear taps. It may be mipmapped depending + // on settings and dxdy. + // First get the texel weights and normalization factor as above. + const float denom_inv = 0.5/(sigma*sigma); + const float w0 = 1.0; + const float w1 = exp(-1.0 * denom_inv); + const float w2 = exp(-4.0 * denom_inv); + const float w3 = exp(-9.0 * denom_inv); + const float w4 = exp(-16.0 * denom_inv); + const float weight_sum_inv = 1.0 / (w0 + 2.0 * (w1 + w2 + w3 + w4)); + // Calculate combined weights and linear sample ratios between texel pairs. + const float w12 = w1 + w2; + const float w34 = w3 + w4; + const float w12_ratio = w2/w12; + const float w34_ratio = w4/w34; + // Statically normalize weights, sum weighted samples, and return: + vec3 sum = vec3(0.0); + sum += w34 * tex2D_linearize(tex, tex_uv - (3.0 + w34_ratio) * dxdy).rgb; + sum += w12 * tex2D_linearize(tex, tex_uv - (1.0 + w12_ratio) * dxdy).rgb; + sum += w0 * tex2D_linearize(tex, tex_uv).rgb; + sum += w12 * tex2D_linearize(tex, tex_uv + (1.0 + w12_ratio) * dxdy).rgb; + sum += w34 * tex2D_linearize(tex, tex_uv + (3.0 + w34_ratio) * dxdy).rgb; + return sum * weight_sum_inv; +} + +vec3 tex2Dblur9fast(const sampler2D tex, const vec2 tex_uv, + const vec2 dxdy) +{ + return tex2Dblur9fast(tex, tex_uv, dxdy, blur9_std_dev); +} + +#endif // BLUR_FUNCTIONS_H \ No newline at end of file diff --git a/crt/shaders/crt-royale/src/crt-royale-bloom-approx.slang b/crt/shaders/crt-royale/src/crt-royale-bloom-approx.slang index 77aa1c1..13c06d6 100644 --- a/crt/shaders/crt-royale/src/crt-royale-bloom-approx.slang +++ b/crt/shaders/crt-royale/src/crt-royale-bloom-approx.slang @@ -345,5 +345,5 @@ void main() color = vec3(color_r.r, color_g.g, color_b.b); } // Encode and output the blurred image: - FragColor = vec4(color, 1.0); + FragColor = vec4(1.0);//vec4(color, 1.0); } \ No newline at end of file diff --git a/crt/shaders/crt-royale/src/crt-royale-mask-resize-vertical.slang b/crt/shaders/crt-royale/src/crt-royale-mask-resize-vertical.slang new file mode 100644 index 0000000..744b450 --- /dev/null +++ b/crt/shaders/crt-royale/src/crt-royale-mask-resize-vertical.slang @@ -0,0 +1,206 @@ +#version 450 + +layout(push_constant) uniform Push +{ + vec4 SourceSize; + vec4 OriginalSize; + vec4 OutputSize; + uint FrameCount; +} params; + +layout(std140, set = 0, binding = 0) uniform UBO +{ + mat4 MVP; +} global; + +///////////////////////////// GPL LICENSE NOTICE ///////////////////////////// + +// crt-royale: A full-featured CRT shader, with cheese. +// Copyright (C) 2014 TroggleMonkey +// +// This program is free software; you can redistribute it and/or modify it +// under the terms of the GNU General Public License as published by the Free +// Software Foundation; either version 2 of the License, or any later version. +// +// This program is distributed in the hope that it will be useful, but WITHOUT +// ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or +// FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for +// more details. +// +// You should have received a copy of the GNU General Public License along with +// this program; if not, write to the Free Software Foundation, Inc., 59 Temple +// Place, Suite 330, Boston, MA 02111-1307 USA + + +///////////////////////////// SETTINGS MANAGEMENT //////////////////////////// + +//#include "../user-settings.h" +//#include "derived-settings-and-constants.h" +//#include "bind-shader-params.h" + +////////////////////////////////// INCLUDES ////////////////////////////////// + +#include "includes.h" +#include "phosphor-mask-resizing.h" + +#pragma stage vertex +layout(location = 0) in vec4 Position; +layout(location = 1) in vec2 TexCoord; +layout(location = 0) out vec2 tex_uv; +layout(location = 1) out vec2 src_tex_uv_wrap; +layout(location = 2) out vec2 resize_magnification_scale; + +void main() +{ + gl_Position = global.MVP * Position; + tex_uv = TexCoord; + + // First estimate the viewport size (the user will get the wrong number of + // triads if it's wrong and mask_specify_num_triads is 1.0/true). + const float viewport_y = params.OutputSize.y / mask_resize_viewport_scale.y; + const float aspect_ratio = geom_aspect_ratio_x / geom_aspect_ratio_y; + const vec2 estimated_viewport_size = + vec2(viewport_y * aspect_ratio, viewport_y); + // Estimate the output size of MASK_RESIZE (the next pass). The estimated + // x component shouldn't matter, because we're not using the x result, and + // we're not swearing it's correct (if we did, the x result would influence + // the y result to maintain the tile aspect ratio). + const vec2 estimated_mask_resize_output_size = + vec2(params.OutputSize.y * aspect_ratio, params.OutputSize.y); + // Find the final intended [y] size of our resized phosphor mask tiles, + // then the tile size for the current pass (resize y only): + const vec2 mask_resize_tile_size = get_resized_mask_tile_size( + estimated_viewport_size, estimated_mask_resize_output_size, false); + const vec2 pass_output_tile_size = vec2(min( + mask_resize_src_lut_size.x, params.OutputSize.x), mask_resize_tile_size.y); + + // We'll render resized tiles until filling the output FBO or meeting a + // limit, so compute [wrapped] tile uv coords based on the output uv coords + // and the number of tiles that will fit in the FBO. + const vec2 output_tiles_this_pass = params.OutputSize.xy / pass_output_tile_size; + const vec2 output_video_uv = tex_uv; + const vec2 tile_uv_wrap = output_video_uv * output_tiles_this_pass; + + // The input LUT is just a single mask tile, so texture uv coords are the + // same as tile uv coords (save frac() for the fragment shader). The + // magnification scale is also straightforward: + src_tex_uv_wrap = tile_uv_wrap; + resize_magnification_scale = + pass_output_tile_size / mask_resize_src_lut_size; +} + +#pragma stage fragment +layout(location = 0) in vec2 tex_uv; +layout(location = 1) in vec2 src_tex_uv_wrap; +layout(location = 2) in vec2 resize_magnification_scale; +layout(location = 0) out vec4 FragColor; +layout(set = 0, binding = 2) uniform sampler2D Source; +#ifdef PHOSPHOR_MASK_RESIZE_MIPMAPPED_LUT +layout(set = 0, binding = 3) uniform sampler2D mask_grille_texture_large; +layout(set = 0, binding = 4) uniform sampler2D mask_slot_texture_large; +layout(set = 0, binding = 5) uniform sampler2D mask_shadow_texture_large; + +void main() +{ + // Resize the input phosphor mask tile to the final vertical size it will + // appear on screen. Keep 1x horizontal size if possible (IN.output_size + // >= mask_resize_src_lut_size), and otherwise linearly sample horizontally + // to fit exactly one tile. Lanczos-resizing the phosphor mask achieves + // much sharper results than mipmapping, and vertically resizing first + // minimizes the total number of taps required. We output a number of + // resized tiles >= mask_resize_num_tiles for easier tiled sampling later. + #ifdef PHOSPHOR_MASK_MANUALLY_RESIZE + // Discard unneeded fragments in case our profile allows real branches. + const vec2 tile_uv_wrap = src_tex_uv_wrap; + if(get_mask_sample_mode() < 0.5 && + tile_uv_wrap.y <= mask_resize_num_tiles) + { + const float src_dy = 1.0/mask_resize_src_lut_size.y; + const vec2 src_tex_uv = frac(src_tex_uv_wrap); + vec3 pixel_color; + // If mask_type is static, this branch will be resolved statically. + if(mask_type < 0.5) + { + pixel_color = downsample_vertical_sinc_tiled( + mask_grille_texture_large, src_tex_uv, mask_resize_src_lut_size, + src_dy, resize_magnification_scale.y, 1.0); + } + else if(mask_type < 1.5) + { + pixel_color = downsample_vertical_sinc_tiled( + mask_slot_texture_large, src_tex_uv, mask_resize_src_lut_size, + src_dy, resize_magnification_scale.y, 1.0); + } + else + { + pixel_color = downsample_vertical_sinc_tiled( + mask_shadow_texture_large, src_tex_uv, mask_resize_src_lut_size, + src_dy, resize_magnification_scale.y, 1.0); + } + // The input LUT was linear RGB, and so is our output: + FragColor = vec4(pixel_color, 1.0); + } + else + { + discard; + } + #else + discard; + FragColor = vec4(1.0); + #endif +} +#else +layout(set = 0, binding = 3) uniform sampler2D mask_grille_texture_small; +layout(set = 0, binding = 4) uniform sampler2D mask_slot_texture_small; +layout(set = 0, binding = 5) uniform sampler2D mask_shadow_texture_small; + +void main() +{ + // Resize the input phosphor mask tile to the final vertical size it will + // appear on screen. Keep 1x horizontal size if possible (IN.output_size + // >= mask_resize_src_lut_size), and otherwise linearly sample horizontally + // to fit exactly one tile. Lanczos-resizing the phosphor mask achieves + // much sharper results than mipmapping, and vertically resizing first + // minimizes the total number of taps required. We output a number of + // resized tiles >= mask_resize_num_tiles for easier tiled sampling later. + #ifdef PHOSPHOR_MASK_MANUALLY_RESIZE + // Discard unneeded fragments in case our profile allows real branches. + const vec2 tile_uv_wrap = src_tex_uv_wrap; + if(get_mask_sample_mode() < 0.5 && + tile_uv_wrap.y <= mask_resize_num_tiles) + { + const float src_dy = 1.0/mask_resize_src_lut_size.y; + const vec2 src_tex_uv = frac(src_tex_uv_wrap); + vec3 pixel_color; + // If mask_type is static, this branch will be resolved statically. + if(mask_type < 0.5) + { + pixel_color = downsample_vertical_sinc_tiled( + mask_grille_texture_small, src_tex_uv, mask_resize_src_lut_size, + src_dy, resize_magnification_scale.y, 1.0); + } + else if(mask_type < 1.5) + { + pixel_color = downsample_vertical_sinc_tiled( + mask_slot_texture_small, src_tex_uv, mask_resize_src_lut_size, + src_dy, resize_magnification_scale.y, 1.0); + } + else + { + pixel_color = downsample_vertical_sinc_tiled( + mask_shadow_texture_small, src_tex_uv, mask_resize_src_lut_size, + src_dy, resize_magnification_scale.y, 1.0); + } + // The input LUT was linear RGB, and so is our output: + FragColor = vec4(pixel_color, 1.0); + } + else + { + discard; + } + #else + discard; + FragColor = vec4(1.0); + #endif +} +#endif \ No newline at end of file diff --git a/crt/shaders/crt-royale/src/gamma-management.h b/crt/shaders/crt-royale/src/gamma-management.h index 4236bb3..0843122 100644 --- a/crt/shaders/crt-royale/src/gamma-management.h +++ b/crt/shaders/crt-royale/src/gamma-management.h @@ -1,3 +1,6 @@ +#ifndef GAMMA_MANAGEMENT_H +#define GAMMA_MANAGEMENT_H + /////////////////////////////// BASE CONSTANTS /////////////////////////////// // Set standard gamma constants, but allow users to override them: @@ -157,4 +160,6 @@ vec4 encode_output(const vec4 color) //#define tex2D_linearize(C, D, E) decode_input(vec4(texture(C, D, E))) //vec4 tex2D_linearize(const sampler2D tex, const vec2 tex_coords, const int texel_off) -//{ return decode_input(vec4(texture(tex, tex_coords, texel_off))); } \ No newline at end of file +//{ return decode_input(vec4(texture(tex, tex_coords, texel_off))); } + +#endif // GAMMA_MANAGEMENT_H \ No newline at end of file diff --git a/crt/shaders/crt-royale/src/includes.h b/crt/shaders/crt-royale/src/includes.h index c30ade7..15948d9 100644 --- a/crt/shaders/crt-royale/src/includes.h +++ b/crt/shaders/crt-royale/src/includes.h @@ -7,4 +7,5 @@ #include "gamma-management.h" //#include "../../../../include/gamma-management.h" <-move includes into crt-royale's src directory until it's actually working #include "blur-functions.h" //#include "../../../../include/blur-functions.h" <-move includes into crt-royale's src directory until it's actually working #include "scanline-functions.h" -#include "bloom-functions.h" \ No newline at end of file +#include "bloom-functions.h" +#include "phosphor-mask-resizing.h" \ No newline at end of file diff --git a/crt/shaders/crt-royale/src/phosphor-mask-resizing-old.h b/crt/shaders/crt-royale/src/phosphor-mask-resizing-old.h new file mode 100644 index 0000000..4d28aed --- /dev/null +++ b/crt/shaders/crt-royale/src/phosphor-mask-resizing-old.h @@ -0,0 +1,678 @@ +#ifndef PHOSPHOR_MASK_RESIZING_H +#define PHOSPHOR_MASK_RESIZING_H + +///////////////////////////// GPL LICENSE NOTICE ///////////////////////////// + +// crt-royale: A full-featured CRT shader, with cheese. +// Copyright (C) 2014 TroggleMonkey +// +// This program is free software; you can redistribute it and/or modify it +// under the terms of the GNU General Public License as published by the Free +// Software Foundation; either version 2 of the License, or any later version. +// +// This program is distributed in the hope that it will be useful, but WITHOUT +// ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or +// FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for +// more details. +// +// You should have received a copy of the GNU General Public License along with +// this program; if not, write to the Free Software Foundation, Inc., 59 Temple +// Place, Suite 330, Boston, MA 02111-1307 USA + + +////////////////////////////////// INCLUDES ////////////////////////////////// + +//#include "../user-settings.h" +//#include "derived-settings-and-constants.h" +#include "includes.h" + +///////////////////////////// CODEPATH SELECTION ///////////////////////////// + +// Choose a looping strategy based on what's allowed: +// Dynamic loops not allowed: Use a flat static loop. +// Dynamic loops accomodated: Coarsely branch around static loops. +// Dynamic loops assumed allowed: Use a flat dynamic loop. +#ifndef DRIVERS_ALLOW_DYNAMIC_BRANCHES + #ifdef ACCOMODATE_POSSIBLE_DYNAMIC_LOOPS + #define BREAK_LOOPS_INTO_PIECES + #else + #define USE_SINGLE_STATIC_LOOP + #endif +#endif // No else needed: Dynamic loops assumed. + + +////////////////////////////////// CONSTANTS ///////////////////////////////// + +// The larger the resized tile, the fewer samples we'll need for downsizing. +// See if we can get a static min tile size > mask_min_allowed_tile_size: +const float mask_min_allowed_tile_size = ceil( + mask_min_allowed_triad_size * mask_triads_per_tile); +const float mask_min_expected_tile_size = + mask_min_allowed_tile_size; +// Limit the number of sinc resize taps by the maximum minification factor: +const float pi_over_lobes = pi/mask_sinc_lobes; +const float max_sinc_resize_samples_float = 2.0 * mask_sinc_lobes * + mask_resize_src_lut_size.x/mask_min_expected_tile_size; +// Vectorized loops sample in multiples of 4. Round up to be safe: +const float max_sinc_resize_samples_m4 = ceil( + max_sinc_resize_samples_float * 0.25) * 4.0; + + +///////////////////////// RESAMPLING FUNCTION HELPERS //////////////////////// + +float get_dynamic_loop_size(const float magnification_scale) +{ + // Requires: The following global constants must be defined: + // 1.) mask_sinc_lobes + // 2.) max_sinc_resize_samples_m4 + // Returns: The minimum number of texture samples for a correct downsize + // at magnification_scale. + // We're downsizing, so the filter is sized across 2*lobes output pixels + // (not 2*lobes input texels). This impacts distance measurements and the + // minimum number of input samples needed. + const float min_samples_float = 2.0 * mask_sinc_lobes / magnification_scale; + const float min_samples_m4 = ceil(min_samples_float * 0.25) * 4.0; + #ifdef DRIVERS_ALLOW_DYNAMIC_BRANCHES + const float max_samples_m4 = max_sinc_resize_samples_m4; + #else // ifdef BREAK_LOOPS_INTO_PIECES + // Simulating loops with branches imposes a 128-sample limit. + const float max_samples_m4 = min(128.0, max_sinc_resize_samples_m4); + #endif + return min(min_samples_m4, max_samples_m4); +} + +vec2 get_first_texel_tile_uv_and_dist(const vec2 tex_uv, + const vec2 texture_size, const float dr, + const float input_tiles_per_texture_r, const float samples, + const bool vertical) +{ + // Requires: 1.) dr == du == 1.0/texture_size.x or + // dr == dv == 1.0/texture_size.y + // (whichever direction we're resampling in). + // It's a scalar to save register space. + // 2.) input_tiles_per_texture_r is the number of input tiles + // that can fit in the input texture in the direction we're + // resampling this pass. + // 3.) vertical indicates whether we're resampling vertically + // this pass (or horizontally). + // Returns: Pack and return the first sample's tile_uv coord in [0, 1] + // and its texel distance from the destination pixel, in the + // resized dimension only. + // We'll start with the topmost or leftmost sample and work down or right, + // so get the first sample location and distance. Modify both dimensions + // as if we're doing a one-pass 2D resize; we'll throw away the unneeded + // (and incorrect) dimension at the end. + const vec2 curr_texel = tex_uv * texture_size; + const vec2 prev_texel = + floor(curr_texel - vec2(under_half)) + vec2(0.5); + const vec2 first_texel = prev_texel - vec2(samples/2.0 - 1.0); + const vec2 first_texel_uv_wrap_2D = first_texel * dr; + const vec2 first_texel_dist_2D = curr_texel - first_texel; + // Convert from tex_uv to tile_uv coords so we can sub fracts for fmods. + const vec2 first_texel_tile_uv_wrap_2D = + first_texel_uv_wrap_2D * input_tiles_per_texture_r; + // Project wrapped coordinates to the [0, 1] range. We'll do this with all + // samples,but the first texel is special, since it might be negative. + const vec2 coord_negative = + vec2(first_texel_tile_uv_wrap_2D < vec2(0.0)); + const vec2 first_texel_tile_uv_2D = + fract(first_texel_tile_uv_wrap_2D) + coord_negative; + // Pack the first texel's tile_uv coord and texel distance in 1D: + const vec2 tile_u_and_dist = + vec2(first_texel_tile_uv_2D.x, first_texel_dist_2D.x); + const vec2 tile_v_and_dist = + vec2(first_texel_tile_uv_2D.y, first_texel_dist_2D.y); + return vertical ? tile_v_and_dist : tile_u_and_dist; + //return mix(tile_u_and_dist, tile_v_and_dist, float(vertical)); +} + +vec4 tex2Dlod0try(const sampler2D tex, const vec2 tex_uv) +{ + // Mipmapping and anisotropic filtering get confused by sinc-resampling. + // One [slow] workaround is to select the lowest mip level: + #ifdef ANISOTROPIC_RESAMPLING_COMPAT_TEX2DLOD + return tex2Dlod(tex, vec4(tex_uv, 0.0, 0.0)); + #else + #ifdef ANISOTROPIC_RESAMPLING_COMPAT_TEX2DBIAS + return tex2Dbias(tex, vec4(tex_uv, 0.0, -16.0)); + #else + return texture(tex, tex_uv); + #endif + #endif +} + + +////////////////////////////// LOOP BODY MACROS ////////////////////////////// + +// Using inline functions can exceed the temporary register limit, so we're +// stuck with #define macros (I'm TRULY sorry). They're declared here instead +// of above to be closer to the actual invocation sites. Steps: +// 1.) Get the exact texel location. +// 2.) Sample the phosphor mask (already assumed encoded in linear RGB). +// 3.) Get the distance from the current pixel and sinc weight: +// sinc(dist) = sin(pi * dist)/(pi * dist) +// We can also use the slower/smoother Lanczos instead: +// L(x) = sinc(dist) * sinc(dist / lobes) +// 4.) Accumulate the weight sum in weights, and accumulate the weighted texels +// in pixel_color (we'll normalize outside the loop at the end). +// We vectorize the loop to help reduce the Lanczos window's cost. + + // The r coord is the coord in the dimension we're resizing along (u or v), + // and first_texel_tile_uv_rrrr is a vec4 of the first texel's u or v + // tile_uv coord in [0, 1]. tex_uv_r will contain the tile_uv u or v coord + // for four new texel samples. + #define CALCULATE_R_COORD_FOR_4_SAMPLES \ + const vec4 true_i = vec4(i_base + i) + vec4(0.0, 1.0, 2.0, 3.0); \ + const vec4 tile_uv_r = fract( \ + first_texel_tile_uv_rrrr + true_i * tile_dr); \ + const vec4 tex_uv_r = tile_uv_r * tile_size_uv_r; + + #ifdef PHOSPHOR_MASK_RESIZE_LANCZOS_WINDOW + #define CALCULATE_SINC_RESAMPLE_WEIGHTS \ + const vec4 pi_dist_over_lobes = pi_over_lobes * dist; \ + const vec4 weights = min(sin(pi_dist) * sin(pi_dist_over_lobes) /\ + (pi_dist*pi_dist_over_lobes), vec4(1.0)); + #else + #define CALCULATE_SINC_RESAMPLE_WEIGHTS \ + const vec4 weights = min(sin(pi_dist)/pi_dist, vec4(1.0)); + #endif + + #define UPDATE_COLOR_AND_WEIGHT_SUMS \ + const vec4 dist = magnification_scale * \ + abs(first_dist_unscaled - true_i); \ + const vec4 pi_dist = pi * dist; \ + CALCULATE_SINC_RESAMPLE_WEIGHTS; \ + pixel_color += new_sample0 * weights.xxx; \ + pixel_color += new_sample1 * weights.yyy; \ + pixel_color += new_sample2 * weights.zzz; \ + pixel_color += new_sample3 * weights.www; \ + weight_sum += weights; + + #define VERTICAL_SINC_RESAMPLE_LOOP_BODY \ + CALCULATE_R_COORD_FOR_4_SAMPLES; \ + const vec3 new_sample0 = tex2Dlod0try(texture, \ + vec2(tex_uv.x, tex_uv_r.x)).rgb; \ + const vec3 new_sample1 = tex2Dlod0try(texture, \ + vec2(tex_uv.x, tex_uv_r.y)).rgb; \ + const vec3 new_sample2 = tex2Dlod0try(texture, \ + vec2(tex_uv.x, tex_uv_r.z)).rgb; \ + const vec3 new_sample3 = tex2Dlod0try(texture, \ + vec2(tex_uv.x, tex_uv_r.w)).rgb; \ + UPDATE_COLOR_AND_WEIGHT_SUMS; + + #define HORIZONTAL_SINC_RESAMPLE_LOOP_BODY \ + CALCULATE_R_COORD_FOR_4_SAMPLES; \ + const vec3 new_sample0 = tex2Dlod0try(texture, \ + vec2(tex_uv_r.x, tex_uv.y)).rgb; \ + const vec3 new_sample1 = tex2Dlod0try(texture, \ + vec2(tex_uv_r.y, tex_uv.y)).rgb; \ + const vec3 new_sample2 = tex2Dlod0try(texture, \ + vec2(tex_uv_r.z, tex_uv.y)).rgb; \ + const vec3 new_sample3 = tex2Dlod0try(texture, \ + vec2(tex_uv_r.w, tex_uv.y)).rgb; \ + UPDATE_COLOR_AND_WEIGHT_SUMS; + + +//////////////////////////// RESAMPLING FUNCTIONS //////////////////////////// + +vec3 downsample_vertical_sinc_tiled(const sampler2D texture, + const vec2 tex_uv, const vec2 texture_size, const float dr, + const float magnification_scale, const float tile_size_uv_r) +{ + // Requires: 1.) dr == du == 1.0/texture_size.x or + // dr == dv == 1.0/texture_size.y + // (whichever direction we're resampling in). + // It's a scalar to save register space. + // 2.) tile_size_uv_r is the number of texels an input tile + // takes up in the input texture, in the direction we're + // resampling this pass. + // 3.) magnification_scale must be <= 1.0. + // Returns: Return a [Lanczos] sinc-resampled pixel of a vertically + // downsized input tile embedded in an input texture. (The + // vertical version is special-cased though: It assumes the + // tile size equals the [static] texture size, since it's used + // on an LUT texture input containing one tile. For more + // generic use, eliminate the "static" in the parameters.) + // The "r" in "dr," "tile_size_uv_r," etc. refers to the dimension + // we're resizing along, e.g. "dy" in this case. + #ifdef USE_SINGLE_STATIC_LOOP + // A static loop can be faster, but it might blur too much from using + // more samples than it should. + const int samples = int(max_sinc_resize_samples_m4); + #else + const int samples = int(get_dynamic_loop_size(magnification_scale)); + #endif + + // Get the first sample location (scalar tile uv coord along the resized + // dimension) and distance from the output location (in texels): + const float input_tiles_per_texture_r = 1.0/tile_size_uv_r; + // true = vertical resize: + const vec2 first_texel_tile_r_and_dist = get_first_texel_tile_uv_and_dist( + tex_uv, texture_size, dr, input_tiles_per_texture_r, samples, true); + const vec4 first_texel_tile_uv_rrrr = first_texel_tile_r_and_dist.xxxx; + const vec4 first_dist_unscaled = first_texel_tile_r_and_dist.yyyy; + // Get the tile sample offset: + const float tile_dr = dr * input_tiles_per_texture_r; + + // Sum up each weight and weighted sample color, varying the looping + // strategy based on our expected dynamic loop capabilities. See the + // loop body macros above. + int i_base = 0; + vec4 weight_sum = vec4(0.0); + vec3 pixel_color = vec3(0.0); + const int i_step = 4; + #ifdef BREAK_LOOPS_INTO_PIECES + if(samples - i_base >= 64) + { + for(int i = 0; i < 64; i += i_step) + { + VERTICAL_SINC_RESAMPLE_LOOP_BODY; + } + i_base += 64; + } + if(samples - i_base >= 32) + { + for(int i = 0; i < 32; i += i_step) + { + VERTICAL_SINC_RESAMPLE_LOOP_BODY; + } + i_base += 32; + } + if(samples - i_base >= 16) + { + for(int i = 0; i < 16; i += i_step) + { + VERTICAL_SINC_RESAMPLE_LOOP_BODY; + } + i_base += 16; + } + if(samples - i_base >= 8) + { + for(int i = 0; i < 8; i += i_step) + { + VERTICAL_SINC_RESAMPLE_LOOP_BODY; + } + i_base += 8; + } + if(samples - i_base >= 4) + { + for(int i = 0; i < 4; i += i_step) + { + VERTICAL_SINC_RESAMPLE_LOOP_BODY; + } + i_base += 4; + } + // Do another 4-sample block for a total of 128 max samples. + if(samples - i_base > 0) + { + for(int i = 0; i < 4; i += i_step) + { + VERTICAL_SINC_RESAMPLE_LOOP_BODY; + } + } + #else + for(int i = 0; i < samples; i += i_step) + { + VERTICAL_SINC_RESAMPLE_LOOP_BODY; + } + #endif + // Normalize so the weight_sum == 1.0, and return: + const vec2 weight_sum_reduce = weight_sum.xy + weight_sum.zw; + const vec3 scalar_weight_sum = vec3(weight_sum_reduce.x + + weight_sum_reduce.y); + return (pixel_color/scalar_weight_sum); +} + +vec3 downsample_horizontal_sinc_tiled(const sampler2D texture, + const vec2 tex_uv, const vec2 texture_size, const float dr, + const float magnification_scale, const float tile_size_uv_r) +{ + // Differences from downsample_horizontal_sinc_tiled: + // 1.) The dr and tile_size_uv_r parameters are not static consts. + // 2.) The "vertical" parameter to get_first_texel_tile_uv_and_dist is + // set to false instead of true. + // 3.) The horizontal version of the loop body is used. + // TODO: If we can get guaranteed compile-time dead code elimination, + // we can combine the vertical/horizontal downsampling functions by: + // 1.) Add an extra static const bool parameter called "vertical." + // 2.) Supply it with the result of get_first_texel_tile_uv_and_dist(). + // 3.) Use a conditional assignment in the loop body macro. This is the + // tricky part: We DO NOT want to incur the extra conditional + // assignment in the inner loop at runtime! + // The "r" in "dr," "tile_size_uv_r," etc. refers to the dimension + // we're resizing along, e.g. "dx" in this case. + #ifdef USE_SINGLE_STATIC_LOOP + // If we have to load all samples, we might as well use them. + const int samples = int(max_sinc_resize_samples_m4); + #else + const int samples = int(get_dynamic_loop_size(magnification_scale)); + #endif + + // Get the first sample location (scalar tile uv coord along resized + // dimension) and distance from the output location (in texels): + const float input_tiles_per_texture_r = 1.0/tile_size_uv_r; + // false = horizontal resize: + const vec2 first_texel_tile_r_and_dist = get_first_texel_tile_uv_and_dist( + tex_uv, texture_size, dr, input_tiles_per_texture_r, samples, false); + const vec4 first_texel_tile_uv_rrrr = first_texel_tile_r_and_dist.xxxx; + const vec4 first_dist_unscaled = first_texel_tile_r_and_dist.yyyy; + // Get the tile sample offset: + const float tile_dr = dr * input_tiles_per_texture_r; + + // Sum up each weight and weighted sample color, varying the looping + // strategy based on our expected dynamic loop capabilities. See the + // loop body macros above. + int i_base = 0; + vec4 weight_sum = vec4(0.0); + vec3 pixel_color = vec3(0.0); + const int i_step = 4; + #ifdef BREAK_LOOPS_INTO_PIECES + if(samples - i_base >= 64) + { + for(int i = 0; i < 64; i += i_step) + { + HORIZONTAL_SINC_RESAMPLE_LOOP_BODY; + } + i_base += 64; + } + if(samples - i_base >= 32) + { + for(int i = 0; i < 32; i += i_step) + { + HORIZONTAL_SINC_RESAMPLE_LOOP_BODY; + } + i_base += 32; + } + if(samples - i_base >= 16) + { + for(int i = 0; i < 16; i += i_step) + { + HORIZONTAL_SINC_RESAMPLE_LOOP_BODY; + } + i_base += 16; + } + if(samples - i_base >= 8) + { + for(int i = 0; i < 8; i += i_step) + { + HORIZONTAL_SINC_RESAMPLE_LOOP_BODY; + } + i_base += 8; + } + if(samples - i_base >= 4) + { + for(int i = 0; i < 4; i += i_step) + { + HORIZONTAL_SINC_RESAMPLE_LOOP_BODY; + } + i_base += 4; + } + // Do another 4-sample block for a total of 128 max samples. + if(samples - i_base > 0) + { + for(int i = 0; i < 4; i += i_step) + { + HORIZONTAL_SINC_RESAMPLE_LOOP_BODY; + } + } + #else + for(int i = 0; i < samples; i += i_step) + { + HORIZONTAL_SINC_RESAMPLE_LOOP_BODY; + } + #endif + // Normalize so the weight_sum == 1.0, and return: + const vec2 weight_sum_reduce = weight_sum.xy + weight_sum.zw; + const vec3 scalar_weight_sum = vec3(weight_sum_reduce.x + + weight_sum_reduce.y); + return (pixel_color/scalar_weight_sum); +} + + +//////////////////////////// TILE SIZE CALCULATION /////////////////////////// + +vec2 get_resized_mask_tile_size(const vec2 estimated_viewport_size, + const vec2 estimated_mask_resize_output_size, + const bool solemnly_swear_same_inputs_for_every_pass) +{ + // Requires: The following global constants must be defined according to + // certain constraints: + // 1.) mask_resize_num_triads: Must be high enough that our + // mask sampling method won't have artifacts later + // (long story; see derived-settings-and-constants.h) + // 2.) mask_resize_src_lut_size: Texel size of our mask LUT + // 3.) mask_triads_per_tile: Num horizontal triads in our LUT + // 4.) mask_min_allowed_triad_size: User setting (the more + // restrictive it is, the faster the resize will go) + // 5.) mask_min_allowed_tile_size_x < mask_resize_src_lut_size.x + // 6.) mask_triad_size_desired_{runtime, static} + // 7.) mask_num_triads_desired_{runtime, static} + // 8.) mask_specify_num_triads must be 0.0/1.0 (false/true) + // The function parameters must be defined as follows: + // 1.) estimated_viewport_size == (final viewport size); + // If mask_specify_num_triads is 1.0/true and the viewport + // estimate is wrong, the number of triads will differ from + // the user's preference by about the same factor. + // 2.) estimated_mask_resize_output_size: Must equal the + // output size of the MASK_RESIZE pass. + // Exception: The x component may be estimated garbage if + // and only if the caller throws away the x result. + // 3.) solemnly_swear_same_inputs_for_every_pass: Set to false, + // unless you can guarantee that every call across every + // pass will use the same sizes for the other parameters. + // When calling this across multiple passes, always use the + // same y viewport size/scale, and always use the same x + // viewport size/scale when using the x result. + // Returns: Return the final size of a manually resized mask tile, after + // constraining the desired size to avoid artifacts. Under + // unusual circumstances, tiles may become stretched vertically + // (see wall of text below). + // Stated tile properties must be correct: + const float tile_aspect_ratio_inv = + mask_resize_src_lut_size.y/mask_resize_src_lut_size.x; + const float tile_aspect_ratio = 1.0/tile_aspect_ratio_inv; + const vec2 tile_aspect = vec2(1.0, tile_aspect_ratio_inv); + // If mask_specify_num_triads is 1.0/true and estimated_viewport_size.x is + // wrong, the user preference will be misinterpreted: + const float desired_tile_size_x = mask_triads_per_tile * mix( + mask_triad_size_desired, + estimated_viewport_size.x / mask_num_triads_desired, + mask_specify_num_triads); + if(get_mask_sample_mode() > 0.5) + { + // We don't need constraints unless we're sampling MASK_RESIZE. + return desired_tile_size_x * tile_aspect; + } + // Make sure we're not upsizing: + const float temp_tile_size_x = + min(desired_tile_size_x, mask_resize_src_lut_size.x); + // Enforce min_tile_size and max_tile_size in both dimensions: + const vec2 temp_tile_size = temp_tile_size_x * tile_aspect; + const vec2 min_tile_size = + mask_min_allowed_tile_size * tile_aspect; + const vec2 max_tile_size = + estimated_mask_resize_output_size / mask_resize_num_tiles; + const vec2 clamped_tile_size = + clamp(temp_tile_size, min_tile_size, max_tile_size); + // Try to maintain tile_aspect_ratio. This is the tricky part: + // If we're currently resizing in the y dimension, the x components + // could be MEANINGLESS. (If estimated_mask_resize_output_size.x is + // bogus, then so is max_tile_size.x and clamped_tile_size.x.) + // We can't adjust the y size based on clamped_tile_size.x. If it + // clamps when it shouldn't, it won't clamp again when later passes + // call this function with the correct sizes, and the discrepancy will + // break the sampling coords in MASKED_SCANLINES. Instead, we'll limit + // the x size based on the y size, but not vice versa, unless the + // caller swears the parameters were the same (correct) in every pass. + // As a result, triads could appear vertically stretched if: + // a.) mask_resize_src_lut_size.x > mask_resize_src_lut_size.y: Wide + // LUT's might clamp x more than y (all provided LUT's are square) + // b.) true_viewport_size.x < true_viewport_size.y: The user is playing + // with a vertically oriented screen (not accounted for anyway) + // c.) mask_resize_viewport_scale.x < masked_resize_viewport_scale.y: + // Viewport scales are equal by default. + // If any of these are the case, you can fix the stretching by setting: + // mask_resize_viewport_scale.x = mask_resize_viewport_scale.y * + // (1.0 / min_expected_aspect_ratio) * + // (mask_resize_src_lut_size.x / mask_resize_src_lut_size.y) + const float x_tile_size_from_y = + clamped_tile_size.y * tile_aspect_ratio; + const float y_tile_size_from_x = mix(clamped_tile_size.y, + clamped_tile_size.x * tile_aspect_ratio_inv, + float(solemnly_swear_same_inputs_for_every_pass)); + const vec2 reclamped_tile_size = vec2( + min(clamped_tile_size.x, x_tile_size_from_y), + min(clamped_tile_size.y, y_tile_size_from_x)); + // We need integer tile sizes in both directions for tiled sampling to + // work correctly. Use floor (to make sure we don't round up), but be + // careful to avoid a rounding bug where floor decreases whole numbers: + const vec2 final_resized_tile_size = + floor(reclamped_tile_size + vec2(FIX_ZERO(0.0))); + return final_resized_tile_size; +} + + +///////////////////////// FINAL MASK SAMPLING HELPERS //////////////////////// + +vec4 get_mask_sampling_parameters(const vec2 mask_resize_texture_size, + const vec2 mask_resize_video_size, const vec2 true_viewport_size, + out vec2 mask_tiles_per_screen) +{ + // Requires: 1.) Requirements of get_resized_mask_tile_size() must be + // met, particularly regarding global constants. + // The function parameters must be defined as follows: + // 1.) mask_resize_texture_size == MASK_RESIZE.texture_size + // if get_mask_sample_mode() is 0 (otherwise anything) + // 2.) mask_resize_video_size == MASK_RESIZE.video_size + // if get_mask_sample_mode() is 0 (otherwise anything) + // 3.) true_viewport_size == IN.output_size for a pass set to + // 1.0 viewport scale (i.e. it must be correct) + // Returns: Return a vec4 containing: + // xy: tex_uv coords for the start of the mask tile + // zw: tex_uv size of the mask tile from start to end + // mask_tiles_per_screen is an out parameter containing the + // number of mask tiles that will fit on the screen. + // First get the final resized tile size. The viewport size and mask + // resize viewport scale must be correct, but don't solemnly swear they + // were correct in both mask resize passes unless you know it's true. + // (We can better ensure a correct tile aspect ratio if the parameters are + // guaranteed correct in all passes...but if we lie, we'll get inconsistent + // sizes across passes, resulting in broken texture coordinates.) + const float mask_sample_mode = get_mask_sample_mode(); + const vec2 mask_resize_tile_size = get_resized_mask_tile_size( + true_viewport_size, mask_resize_video_size, false); + if(mask_sample_mode < 0.5) + { + // Sample MASK_RESIZE: The resized tile is a fracttion of the texture + // size and starts at a nonzero offset to allow for border texels: + const vec2 mask_tile_uv_size = mask_resize_tile_size / + mask_resize_texture_size; + const vec2 skipped_tiles = mask_start_texels/mask_resize_tile_size; + const vec2 mask_tile_start_uv = skipped_tiles * mask_tile_uv_size; + // mask_tiles_per_screen must be based on the *true* viewport size: + mask_tiles_per_screen = true_viewport_size / mask_resize_tile_size; + return vec4(mask_tile_start_uv, mask_tile_uv_size); + } + else + { + // If we're tiling at the original size (1:1 pixel:texel), redefine a + // "tile" to be the full texture containing many triads. Otherwise, + // we're hardware-resampling an LUT, and the texture truly contains a + // single unresized phosphor mask tile anyway. + const vec2 mask_tile_uv_size = vec2(1.0); + const vec2 mask_tile_start_uv = vec2(0.0); + if(mask_sample_mode > 1.5) + { + // Repeat the full LUT at a 1:1 pixel:texel ratio without resizing: + mask_tiles_per_screen = true_viewport_size/mask_texture_large_size; + } + else + { + // Hardware-resize the original LUT: + mask_tiles_per_screen = true_viewport_size / mask_resize_tile_size; + } + return vec4(mask_tile_start_uv, mask_tile_uv_size); + } +} + +vec2 fix_tiling_discontinuities_normalized(const vec2 tile_uv, + vec2 duv_dx, vec2 duv_dy) +{ + // Requires: 1.) duv_dx == ddx(tile_uv) + // 2.) duv_dy == ddy(tile_uv) + // 3.) tile_uv contains tile-relative uv coords in [0, 1], + // such that (0.5, 0.5) is the center of a tile, etc. + // ("Tile" can mean texture, the video embedded in the + // texture, or some other "tile" embedded in a texture.) + // Returns: Return new tile_uv coords that contain no discontinuities + // across a 2x2 pixel quad. + // Description: + // When uv coords wrap from 1.0 to 0.0, they create a discontinuity in the + // derivatives, which we assume happened if the absolute difference between + // any fragment in a 2x2 block is > ~half a tile. If the current block has + // a u or v discontinuity and the current fragment is in the first half of + // the tile along that axis (i.e. it wrapped from 1.0 to 0.0), add a tile + // to that coord to make the 2x2 block continuous. (It will now have a + // coord > 1.0 in the padding area beyond the tile.) This function takes + // derivatives as parameters so the caller can reuse them. + // In case we're using high-quality (nVidia-style) derivatives, ensure + // diagonically opposite fragments see each other for correctness: + duv_dx = abs(duv_dx) + abs(ddy(duv_dx)); + duv_dy = abs(duv_dy) + abs(ddx(duv_dy)); + const vec2 pixel_in_first_half_tile = vec2(tile_uv < vec2(0.5)); + const vec2 jump_exists = vec2(duv_dx + duv_dy > vec2(0.5)); + return tile_uv + jump_exists * pixel_in_first_half_tile; +} + +vec2 convert_phosphor_tile_uv_wrap_to_tex_uv(const vec2 tile_uv_wrap, + const vec4 mask_tile_start_uv_and_size) +{ + // Requires: 1.) tile_uv_wrap contains tile-relative uv coords, where the + // tile spans from [0, 1], such that (0.5, 0.5) is at the + // tile center. The input coords can range from [0, inf], + // and their fracttional parts map to a repeated tile. + // ("Tile" can mean texture, the video embedded in the + // texture, or some other "tile" embedded in a texture.) + // 2.) mask_tile_start_uv_and_size.xy contains tex_uv coords + // for the start of the embedded tile in the full texture. + // 3.) mask_tile_start_uv_and_size.zw contains the [fracttional] + // tex_uv size of the embedded tile in the full texture. + // Returns: Return tex_uv coords (used for texture sampling) + // corresponding to tile_uv_wrap. + if(get_mask_sample_mode() < 0.5) + { + // Manually repeat the resized mask tile to fill the screen: + // First get fracttional tile_uv coords. Using fract/fmod on coords + // confuses anisotropic filtering; fix it as user options dictate. + // derived-settings-and-constants.h disables incompatible options. + #ifdef ANISOTROPIC_TILING_COMPAT_TILE_FLAT_TWICE + vec2 tile_uv = fract(tile_uv_wrap * 0.5) * 2.0; + #else + vec2 tile_uv = fract(tile_uv_wrap); + #endif + #ifdef ANISOTROPIC_TILING_COMPAT_FIX_DISCONTINUITIES + const vec2 tile_uv_dx = ddx(tile_uv); + const vec2 tile_uv_dy = ddy(tile_uv); + tile_uv = fix_tiling_discontinuities_normalized(tile_uv, + tile_uv_dx, tile_uv_dy); + #endif + // The tile is embedded in a padded FBO, and it may start at a + // nonzero offset if border texels are used to avoid artifacts: + const vec2 mask_tex_uv = mask_tile_start_uv_and_size.xy + + tile_uv * mask_tile_start_uv_and_size.zw; + return mask_tex_uv; + } + else + { + // Sample from the input phosphor mask texture with hardware tiling. + // If we're tiling at the original size (mode 2), the "tile" is the + // whole texture, and it contains a large number of triads mapped with + // a 1:1 pixel:texel ratio. OTHERWISE, the texture contains a single + // unresized tile. tile_uv_wrap already has correct coords for both! + return tile_uv_wrap; + } +} + + +#endif // PHOSPHOR_MASK_RESIZING_H + diff --git a/crt/shaders/crt-royale/src/phosphor-mask-resizing.h b/crt/shaders/crt-royale/src/phosphor-mask-resizing.h index be26624..7d07c34 100644 --- a/crt/shaders/crt-royale/src/phosphor-mask-resizing.h +++ b/crt/shaders/crt-royale/src/phosphor-mask-resizing.h @@ -22,8 +22,9 @@ ////////////////////////////////// INCLUDES ////////////////////////////////// -#include "../user-settings.h" -#include "derived-settings-and-constants.h" +//#include "../user-settings.h" +//#include "derived-settings-and-constants.h" +#include "includes.h" ///////////////////////////// CODEPATH SELECTION ///////////////////////////// @@ -59,7 +60,7 @@ const float max_sinc_resize_samples_m4 = ceil( ///////////////////////// RESAMPLING FUNCTION HELPERS //////////////////////// -inline float get_dynamic_loop_size(const float magnification_scale) +float get_dynamic_loop_size(const float magnification_scale) { // Requires: The following global constants must be defined: // 1.) mask_sinc_lobes @@ -107,7 +108,7 @@ vec2 get_first_texel_tile_uv_and_dist(const vec2 tex_uv, const vec2 first_texel = prev_texel - vec2(samples/2.0 - 1.0); const vec2 first_texel_uv_wrap_2D = first_texel * dr; const vec2 first_texel_dist_2D = curr_texel - first_texel; - // Convert from tex_uv to tile_uv coords so we can sub fracs for fmods. + // Convert from tex_uv to tile_uv coords so we can sub fracts for fmods. const vec2 first_texel_tile_uv_wrap_2D = first_texel_uv_wrap_2D * input_tiles_per_texture_r; // Project wrapped coordinates to the [0, 1] range. We'll do this with all @@ -115,17 +116,17 @@ vec2 get_first_texel_tile_uv_and_dist(const vec2 tex_uv, const vec2 coord_negative = vec2(first_texel_tile_uv_wrap_2D < vec2(0.0)); const vec2 first_texel_tile_uv_2D = - frac(first_texel_tile_uv_wrap_2D) + coord_negative; + fract(first_texel_tile_uv_wrap_2D) + coord_negative; // Pack the first texel's tile_uv coord and texel distance in 1D: const vec2 tile_u_and_dist = vec2(first_texel_tile_uv_2D.x, first_texel_dist_2D.x); const vec2 tile_v_and_dist = vec2(first_texel_tile_uv_2D.y, first_texel_dist_2D.y); return vertical ? tile_v_and_dist : tile_u_and_dist; - //return lerp(tile_u_and_dist, tile_v_and_dist, float(vertical)); + //return mix(tile_u_and_dist, tile_v_and_dist, float(vertical)); } -inline vec4 tex2Dlod0try(const sampler2D tex, const vec2 tex_uv) +vec4 tex2Dlod0try(const sampler2D tex, const vec2 tex_uv) { // Mipmapping and anisotropic filtering get confused by sinc-resampling. // One [slow] workaround is to select the lowest mip level: @@ -162,7 +163,7 @@ inline vec4 tex2Dlod0try(const sampler2D tex, const vec2 tex_uv) // for four new texel samples. #define CALCULATE_R_COORD_FOR_4_SAMPLES \ const vec4 true_i = vec4(i_base + i) + vec4(0.0, 1.0, 2.0, 3.0); \ - const vec4 tile_uv_r = frac( \ + const vec4 tile_uv_r = fract( \ first_texel_tile_uv_rrrr + true_i * tile_dr); \ const vec4 tex_uv_r = tile_uv_r * tile_size_uv_r; @@ -214,220 +215,6 @@ inline vec4 tex2Dlod0try(const sampler2D tex, const vec2 tex_uv) //////////////////////////// RESAMPLING FUNCTIONS //////////////////////////// -vec3 downsample_vertical_sinc_tiled(const sampler2D texture, - const vec2 tex_uv, const vec2 texture_size, const float dr, - const float magnification_scale, const float tile_size_uv_r) -{ - // Requires: 1.) dr == du == 1.0/texture_size.x or - // dr == dv == 1.0/texture_size.y - // (whichever direction we're resampling in). - // It's a scalar to save register space. - // 2.) tile_size_uv_r is the number of texels an input tile - // takes up in the input texture, in the direction we're - // resampling this pass. - // 3.) magnification_scale must be <= 1.0. - // Returns: Return a [Lanczos] sinc-resampled pixel of a vertically - // downsized input tile embedded in an input texture. (The - // vertical version is special-cased though: It assumes the - // tile size equals the [static] texture size, since it's used - // on an LUT texture input containing one tile. For more - // generic use, eliminate the "static" in the parameters.) - // The "r" in "dr," "tile_size_uv_r," etc. refers to the dimension - // we're resizing along, e.g. "dy" in this case. - #ifdef USE_SINGLE_STATIC_LOOP - // A static loop can be faster, but it might blur too much from using - // more samples than it should. - const int samples = int(max_sinc_resize_samples_m4); - #else - const int samples = int(get_dynamic_loop_size(magnification_scale)); - #endif - - // Get the first sample location (scalar tile uv coord along the resized - // dimension) and distance from the output location (in texels): - const float input_tiles_per_texture_r = 1.0/tile_size_uv_r; - // true = vertical resize: - const vec2 first_texel_tile_r_and_dist = get_first_texel_tile_uv_and_dist( - tex_uv, texture_size, dr, input_tiles_per_texture_r, samples, true); - const vec4 first_texel_tile_uv_rrrr = first_texel_tile_r_and_dist.xxxx; - const vec4 first_dist_unscaled = first_texel_tile_r_and_dist.yyyy; - // Get the tile sample offset: - const float tile_dr = dr * input_tiles_per_texture_r; - - // Sum up each weight and weighted sample color, varying the looping - // strategy based on our expected dynamic loop capabilities. See the - // loop body macros above. - int i_base = 0; - vec4 weight_sum = vec4(0.0); - vec3 pixel_color = vec3(0.0); - const int i_step = 4; - #ifdef BREAK_LOOPS_INTO_PIECES - if(samples - i_base >= 64) - { - for(int i = 0; i < 64; i += i_step) - { - VERTICAL_SINC_RESAMPLE_LOOP_BODY; - } - i_base += 64; - } - if(samples - i_base >= 32) - { - for(int i = 0; i < 32; i += i_step) - { - VERTICAL_SINC_RESAMPLE_LOOP_BODY; - } - i_base += 32; - } - if(samples - i_base >= 16) - { - for(int i = 0; i < 16; i += i_step) - { - VERTICAL_SINC_RESAMPLE_LOOP_BODY; - } - i_base += 16; - } - if(samples - i_base >= 8) - { - for(int i = 0; i < 8; i += i_step) - { - VERTICAL_SINC_RESAMPLE_LOOP_BODY; - } - i_base += 8; - } - if(samples - i_base >= 4) - { - for(int i = 0; i < 4; i += i_step) - { - VERTICAL_SINC_RESAMPLE_LOOP_BODY; - } - i_base += 4; - } - // Do another 4-sample block for a total of 128 max samples. - if(samples - i_base > 0) - { - for(int i = 0; i < 4; i += i_step) - { - VERTICAL_SINC_RESAMPLE_LOOP_BODY; - } - } - #else - for(int i = 0; i < samples; i += i_step) - { - VERTICAL_SINC_RESAMPLE_LOOP_BODY; - } - #endif - // Normalize so the weight_sum == 1.0, and return: - const vec2 weight_sum_reduce = weight_sum.xy + weight_sum.zw; - const vec3 scalar_weight_sum = vec3(weight_sum_reduce.x + - weight_sum_reduce.y); - return (pixel_color/scalar_weight_sum); -} - -vec3 downsample_horizontal_sinc_tiled(const sampler2D texture, - const vec2 tex_uv, const vec2 texture_size, const float dr, - const float magnification_scale, const float tile_size_uv_r) -{ - // Differences from downsample_horizontal_sinc_tiled: - // 1.) The dr and tile_size_uv_r parameters are not static consts. - // 2.) The "vertical" parameter to get_first_texel_tile_uv_and_dist is - // set to false instead of true. - // 3.) The horizontal version of the loop body is used. - // TODO: If we can get guaranteed compile-time dead code elimination, - // we can combine the vertical/horizontal downsampling functions by: - // 1.) Add an extra static const bool parameter called "vertical." - // 2.) Supply it with the result of get_first_texel_tile_uv_and_dist(). - // 3.) Use a conditional assignment in the loop body macro. This is the - // tricky part: We DO NOT want to incur the extra conditional - // assignment in the inner loop at runtime! - // The "r" in "dr," "tile_size_uv_r," etc. refers to the dimension - // we're resizing along, e.g. "dx" in this case. - #ifdef USE_SINGLE_STATIC_LOOP - // If we have to load all samples, we might as well use them. - const int samples = int(max_sinc_resize_samples_m4); - #else - const int samples = int(get_dynamic_loop_size(magnification_scale)); - #endif - - // Get the first sample location (scalar tile uv coord along resized - // dimension) and distance from the output location (in texels): - const float input_tiles_per_texture_r = 1.0/tile_size_uv_r; - // false = horizontal resize: - const vec2 first_texel_tile_r_and_dist = get_first_texel_tile_uv_and_dist( - tex_uv, texture_size, dr, input_tiles_per_texture_r, samples, false); - const vec4 first_texel_tile_uv_rrrr = first_texel_tile_r_and_dist.xxxx; - const vec4 first_dist_unscaled = first_texel_tile_r_and_dist.yyyy; - // Get the tile sample offset: - const float tile_dr = dr * input_tiles_per_texture_r; - - // Sum up each weight and weighted sample color, varying the looping - // strategy based on our expected dynamic loop capabilities. See the - // loop body macros above. - int i_base = 0; - vec4 weight_sum = vec4(0.0); - vec3 pixel_color = vec3(0.0); - const int i_step = 4; - #ifdef BREAK_LOOPS_INTO_PIECES - if(samples - i_base >= 64) - { - for(int i = 0; i < 64; i += i_step) - { - HORIZONTAL_SINC_RESAMPLE_LOOP_BODY; - } - i_base += 64; - } - if(samples - i_base >= 32) - { - for(int i = 0; i < 32; i += i_step) - { - HORIZONTAL_SINC_RESAMPLE_LOOP_BODY; - } - i_base += 32; - } - if(samples - i_base >= 16) - { - for(int i = 0; i < 16; i += i_step) - { - HORIZONTAL_SINC_RESAMPLE_LOOP_BODY; - } - i_base += 16; - } - if(samples - i_base >= 8) - { - for(int i = 0; i < 8; i += i_step) - { - HORIZONTAL_SINC_RESAMPLE_LOOP_BODY; - } - i_base += 8; - } - if(samples - i_base >= 4) - { - for(int i = 0; i < 4; i += i_step) - { - HORIZONTAL_SINC_RESAMPLE_LOOP_BODY; - } - i_base += 4; - } - // Do another 4-sample block for a total of 128 max samples. - if(samples - i_base > 0) - { - for(int i = 0; i < 4; i += i_step) - { - HORIZONTAL_SINC_RESAMPLE_LOOP_BODY; - } - } - #else - for(int i = 0; i < samples; i += i_step) - { - HORIZONTAL_SINC_RESAMPLE_LOOP_BODY; - } - #endif - // Normalize so the weight_sum == 1.0, and return: - const vec2 weight_sum_reduce = weight_sum.xy + weight_sum.zw; - const vec3 scalar_weight_sum = vec3(weight_sum_reduce.x + - weight_sum_reduce.y); - return (pixel_color/scalar_weight_sum); -} - - //////////////////////////// TILE SIZE CALCULATION /////////////////////////// vec2 get_resized_mask_tile_size(const vec2 estimated_viewport_size, @@ -473,7 +260,7 @@ vec2 get_resized_mask_tile_size(const vec2 estimated_viewport_size, const vec2 tile_aspect = vec2(1.0, tile_aspect_ratio_inv); // If mask_specify_num_triads is 1.0/true and estimated_viewport_size.x is // wrong, the user preference will be misinterpreted: - const float desired_tile_size_x = mask_triads_per_tile * lerp( + const float desired_tile_size_x = mask_triads_per_tile * mix( mask_triad_size_desired, estimated_viewport_size.x / mask_num_triads_desired, mask_specify_num_triads); @@ -516,7 +303,7 @@ vec2 get_resized_mask_tile_size(const vec2 estimated_viewport_size, // (mask_resize_src_lut_size.x / mask_resize_src_lut_size.y) const float x_tile_size_from_y = clamped_tile_size.y * tile_aspect_ratio; - const float y_tile_size_from_x = lerp(clamped_tile_size.y, + const float y_tile_size_from_x = mix(clamped_tile_size.y, clamped_tile_size.x * tile_aspect_ratio_inv, float(solemnly_swear_same_inputs_for_every_pass)); const vec2 reclamped_tile_size = vec2( @@ -531,147 +318,6 @@ vec2 get_resized_mask_tile_size(const vec2 estimated_viewport_size, } -///////////////////////// FINAL MASK SAMPLING HELPERS //////////////////////// - -vec4 get_mask_sampling_parameters(const vec2 mask_resize_texture_size, - const vec2 mask_resize_video_size, const vec2 true_viewport_size, - out vec2 mask_tiles_per_screen) -{ - // Requires: 1.) Requirements of get_resized_mask_tile_size() must be - // met, particularly regarding global constants. - // The function parameters must be defined as follows: - // 1.) mask_resize_texture_size == MASK_RESIZE.texture_size - // if get_mask_sample_mode() is 0 (otherwise anything) - // 2.) mask_resize_video_size == MASK_RESIZE.video_size - // if get_mask_sample_mode() is 0 (otherwise anything) - // 3.) true_viewport_size == IN.output_size for a pass set to - // 1.0 viewport scale (i.e. it must be correct) - // Returns: Return a vec4 containing: - // xy: tex_uv coords for the start of the mask tile - // zw: tex_uv size of the mask tile from start to end - // mask_tiles_per_screen is an out parameter containing the - // number of mask tiles that will fit on the screen. - // First get the final resized tile size. The viewport size and mask - // resize viewport scale must be correct, but don't solemnly swear they - // were correct in both mask resize passes unless you know it's true. - // (We can better ensure a correct tile aspect ratio if the parameters are - // guaranteed correct in all passes...but if we lie, we'll get inconsistent - // sizes across passes, resulting in broken texture coordinates.) - const float mask_sample_mode = get_mask_sample_mode(); - const vec2 mask_resize_tile_size = get_resized_mask_tile_size( - true_viewport_size, mask_resize_video_size, false); - if(mask_sample_mode < 0.5) - { - // Sample MASK_RESIZE: The resized tile is a fraction of the texture - // size and starts at a nonzero offset to allow for border texels: - const vec2 mask_tile_uv_size = mask_resize_tile_size / - mask_resize_texture_size; - const vec2 skipped_tiles = mask_start_texels/mask_resize_tile_size; - const vec2 mask_tile_start_uv = skipped_tiles * mask_tile_uv_size; - // mask_tiles_per_screen must be based on the *true* viewport size: - mask_tiles_per_screen = true_viewport_size / mask_resize_tile_size; - return vec4(mask_tile_start_uv, mask_tile_uv_size); - } - else - { - // If we're tiling at the original size (1:1 pixel:texel), redefine a - // "tile" to be the full texture containing many triads. Otherwise, - // we're hardware-resampling an LUT, and the texture truly contains a - // single unresized phosphor mask tile anyway. - const vec2 mask_tile_uv_size = vec2(1.0); - const vec2 mask_tile_start_uv = vec2(0.0); - if(mask_sample_mode > 1.5) - { - // Repeat the full LUT at a 1:1 pixel:texel ratio without resizing: - mask_tiles_per_screen = true_viewport_size/mask_texture_large_size; - } - else - { - // Hardware-resize the original LUT: - mask_tiles_per_screen = true_viewport_size / mask_resize_tile_size; - } - return vec4(mask_tile_start_uv, mask_tile_uv_size); - } -} - -vec2 fix_tiling_discontinuities_normalized(const vec2 tile_uv, - vec2 duv_dx, vec2 duv_dy) -{ - // Requires: 1.) duv_dx == ddx(tile_uv) - // 2.) duv_dy == ddy(tile_uv) - // 3.) tile_uv contains tile-relative uv coords in [0, 1], - // such that (0.5, 0.5) is the center of a tile, etc. - // ("Tile" can mean texture, the video embedded in the - // texture, or some other "tile" embedded in a texture.) - // Returns: Return new tile_uv coords that contain no discontinuities - // across a 2x2 pixel quad. - // Description: - // When uv coords wrap from 1.0 to 0.0, they create a discontinuity in the - // derivatives, which we assume happened if the absolute difference between - // any fragment in a 2x2 block is > ~half a tile. If the current block has - // a u or v discontinuity and the current fragment is in the first half of - // the tile along that axis (i.e. it wrapped from 1.0 to 0.0), add a tile - // to that coord to make the 2x2 block continuous. (It will now have a - // coord > 1.0 in the padding area beyond the tile.) This function takes - // derivatives as parameters so the caller can reuse them. - // In case we're using high-quality (nVidia-style) derivatives, ensure - // diagonically opposite fragments see each other for correctness: - duv_dx = abs(duv_dx) + abs(ddy(duv_dx)); - duv_dy = abs(duv_dy) + abs(ddx(duv_dy)); - const vec2 pixel_in_first_half_tile = vec2(tile_uv < vec2(0.5)); - const vec2 jump_exists = vec2(duv_dx + duv_dy > vec2(0.5)); - return tile_uv + jump_exists * pixel_in_first_half_tile; -} - -vec2 convert_phosphor_tile_uv_wrap_to_tex_uv(const vec2 tile_uv_wrap, - const vec4 mask_tile_start_uv_and_size) -{ - // Requires: 1.) tile_uv_wrap contains tile-relative uv coords, where the - // tile spans from [0, 1], such that (0.5, 0.5) is at the - // tile center. The input coords can range from [0, inf], - // and their fractional parts map to a repeated tile. - // ("Tile" can mean texture, the video embedded in the - // texture, or some other "tile" embedded in a texture.) - // 2.) mask_tile_start_uv_and_size.xy contains tex_uv coords - // for the start of the embedded tile in the full texture. - // 3.) mask_tile_start_uv_and_size.zw contains the [fractional] - // tex_uv size of the embedded tile in the full texture. - // Returns: Return tex_uv coords (used for texture sampling) - // corresponding to tile_uv_wrap. - if(get_mask_sample_mode() < 0.5) - { - // Manually repeat the resized mask tile to fill the screen: - // First get fractional tile_uv coords. Using frac/fmod on coords - // confuses anisotropic filtering; fix it as user options dictate. - // derived-settings-and-constants.h disables incompatible options. - #ifdef ANISOTROPIC_TILING_COMPAT_TILE_FLAT_TWICE - vec2 tile_uv = frac(tile_uv_wrap * 0.5) * 2.0; - #else - vec2 tile_uv = frac(tile_uv_wrap); - #endif - #ifdef ANISOTROPIC_TILING_COMPAT_FIX_DISCONTINUITIES - const vec2 tile_uv_dx = ddx(tile_uv); - const vec2 tile_uv_dy = ddy(tile_uv); - tile_uv = fix_tiling_discontinuities_normalized(tile_uv, - tile_uv_dx, tile_uv_dy); - #endif - // The tile is embedded in a padded FBO, and it may start at a - // nonzero offset if border texels are used to avoid artifacts: - const vec2 mask_tex_uv = mask_tile_start_uv_and_size.xy + - tile_uv * mask_tile_start_uv_and_size.zw; - return mask_tex_uv; - } - else - { - // Sample from the input phosphor mask texture with hardware tiling. - // If we're tiling at the original size (mode 2), the "tile" is the - // whole texture, and it contains a large number of triads mapped with - // a 1:1 pixel:texel ratio. OTHERWISE, the texture contains a single - // unresized tile. tile_uv_wrap already has correct coords for both! - return tile_uv_wrap; - } -} - #endif // PHOSPHOR_MASK_RESIZING_H diff --git a/crt/shaders/crt-royale/src/scanline-functions.h b/crt/shaders/crt-royale/src/scanline-functions.h index 48f0073..42da3db 100644 --- a/crt/shaders/crt-royale/src/scanline-functions.h +++ b/crt/shaders/crt-royale/src/scanline-functions.h @@ -1,3 +1,6 @@ +#ifndef SCANLINE_FUNCTIONS_H +#define SCANLINE_FUNCTIONS_H + ///////////////////////////// GPL LICENSE NOTICE ///////////////////////////// // crt-royale: A full-featured CRT shader, with cheese. @@ -350,4 +353,6 @@ vec3 scanline_contrib(vec3 dist, vec3 color, dist, color, pixel_height, sigma_range); } } -} \ No newline at end of file +} + +#endif // SCANLINE_FUNCTIONS_H \ No newline at end of file diff --git a/crt/shaders/crt-royale/src/special-functions.h b/crt/shaders/crt-royale/src/special-functions.h index 1f6d7a4..2a06390 100644 --- a/crt/shaders/crt-royale/src/special-functions.h +++ b/crt/shaders/crt-royale/src/special-functions.h @@ -1,3 +1,7 @@ +#ifndef SPECIAL_FUNCTIONS_H +#define SPECIAL_FUNCTIONS_H + + ///////////////////////////////// MIT LICENSE //////////////////////////////// // Copyright (C) 2014 TroggleMonkey @@ -489,4 +493,6 @@ float normalized_ligamma(const float s, const float z) const float s_inv = 1.0/s; const float gamma_s_inv = 1.0/gamma_impl(s, s_inv); return normalized_ligamma_impl(s, z, s_inv, gamma_s_inv); -} \ No newline at end of file +} + +#endif // SPECIAL_FUNCTIONS_H \ No newline at end of file diff --git a/include/blur-functions-old.h b/include/blur-functions-old.h new file mode 100644 index 0000000..05da1c7 --- /dev/null +++ b/include/blur-functions-old.h @@ -0,0 +1,1916 @@ +#ifndef BLUR_FUNCTIONS_H +#define BLUR_FUNCTIONS_H + +///////////////////////////////// MIT LICENSE //////////////////////////////// + +// Copyright (C) 2014 TroggleMonkey +// +// 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. + +///////////////////////////////// DESCRIPTION //////////////////////////////// + +// This file provides reusable one-pass and separable (two-pass) blurs. +// Requires: All blurs share these requirements (dxdy requirement is split): +// 1.) All requirements of gamma-management.h must be satisfied! +// 2.) filter_linearN must == "true" in your .cgp preset unless +// you're using tex2DblurNresize at 1x scale. +// 3.) mipmap_inputN must == "true" in your .cgp preset if +// IN.output_size < IN.video_size. +// 4.) IN.output_size == IN.video_size / pow(2, M), where M is some +// positive integer. tex2Dblur*resize can resize arbitrarily +// (and the blur will be done after resizing), but arbitrary +// resizes "fail" with other blurs due to the way they mix +// static weights with bilinear sample exploitation. +// 5.) In general, dxdy should contain the uv pixel spacing: +// dxdy = (IN.video_size/IN.output_size)/IN.texture_size +// 6.) For separable blurs (tex2DblurNresize and tex2DblurNfast), +// zero out the dxdy component in the unblurred dimension: +// dxdy = vec2(dxdy.x, 0.0) or vec2(0.0, dxdy.y) +// Many blurs share these requirements: +// 1.) One-pass blurs require scale_xN == scale_yN or scales > 1.0, +// or they will blur more in the lower-scaled dimension. +// 2.) One-pass shared sample blurs require ddx(), ddy(), and +// tex2Dlod() to be supported by the current Cg profile, and +// the drivers must support high-quality derivatives. +// 3.) One-pass shared sample blurs require: +// tex_uv.w == log2(IN.video_size/IN.output_size).y; +// Non-wrapper blurs share this requirement: +// 1.) sigma is the intended standard deviation of the blur +// Wrapper blurs share this requirement, which is automatically +// met (unless OVERRIDE_BLUR_STD_DEVS is #defined; see below): +// 1.) blurN_std_dev must be global static const float values +// specifying standard deviations for Nx blurs in units +// of destination pixels +// Optional: 1.) The including file (or an earlier included file) may +// optionally #define USE_BINOMIAL_BLUR_STD_DEVS to replace +// default standard deviations with those matching a binomial +// distribution. (See below for details/properties.) +// 2.) The including file (or an earlier included file) may +// optionally #define OVERRIDE_BLUR_STD_DEVS and override: +// static const float blur3_std_dev +// static const float blur4_std_dev +// static const float blur5_std_dev +// static const float blur6_std_dev +// static const float blur7_std_dev +// static const float blur8_std_dev +// static const float blur9_std_dev +// static const float blur10_std_dev +// static const float blur11_std_dev +// static const float blur12_std_dev +// static const float blur17_std_dev +// static const float blur25_std_dev +// static const float blur31_std_dev +// static const float blur43_std_dev +// 3.) The including file (or an earlier included file) may +// optionally #define OVERRIDE_ERROR_BLURRING and override: +// static const float error_blurring +// This tuning value helps mitigate weighting errors from one- +// pass shared-sample blurs sharing bilinear samples between +// fragments. Values closer to 0.0 have "correct" blurriness +// but allow more artifacts, and values closer to 1.0 blur away +// artifacts by sampling closer to halfway between texels. +// UPDATE 6/21/14: The above static constants may now be overridden +// by non-static uniform constants. This permits exposing blur +// standard deviations as runtime GUI shader parameters. However, +// using them keeps weights from being statically computed, and the +// speed hit depends on the blur: On my machine, uniforms kill over +// 53% of the framerate with tex2Dblur12x12shared, but they only +// drop the framerate by about 18% with tex2Dblur11fast. +// Quality and Performance Comparisons: +// For the purposes of the following discussion, "no sRGB" means +// GAMMA_ENCODE_EVERY_FBO is #defined, and "sRGB" means it isn't. +// 1.) tex2DblurNfast is always faster than tex2DblurNresize. +// 2.) tex2DblurNresize functions are the only ones that can arbitrarily resize +// well, because they're the only ones that don't exploit bilinear samples. +// This also means they're the only functions which can be truly gamma- +// correct without linear (or sRGB FBO) input, but only at 1x scale. +// 3.) One-pass shared sample blurs only have a speed advantage without sRGB. +// They also have some inaccuracies due to their shared-[bilinear-]sample +// design, which grow increasingly bothersome for smaller blurs and higher- +// frequency source images (relative to their resolution). I had high +// hopes for them, but their most realistic use case is limited to quickly +// reblurring an already blurred input at full resolution. Otherwise: +// a.) If you're blurring a low-resolution source, you want a better blur. +// b.) If you're blurring a lower mipmap, you want a better blur. +// c.) If you're blurring a high-resolution, high-frequency source, you +// want a better blur. +// 4.) The one-pass blurs without shared samples grow slower for larger blurs, +// but they're competitive with separable blurs at 5x5 and smaller, and +// even tex2Dblur7x7 isn't bad if you're wanting to conserve passes. +// Here are some framerates from a GeForce 8800GTS. The first pass resizes to +// viewport size (4x in this test) and linearizes for sRGB codepaths, and the +// remaining passes perform 6 full blurs. Mipmapped tests are performed at the +// same scale, so they just measure the cost of mipmapping each FBO (only every +// other FBO is mipmapped for separable blurs, to mimic realistic usage). +// Mipmap Neither sRGB+Mipmap sRGB Function +// 76.0 92.3 131.3 193.7 tex2Dblur3fast +// 63.2 74.4 122.4 175.5 tex2Dblur3resize +// 93.7 121.2 159.3 263.2 tex2Dblur3x3 +// 59.7 68.7 115.4 162.1 tex2Dblur3x3resize +// 63.2 74.4 122.4 175.5 tex2Dblur5fast +// 49.3 54.8 100.0 132.7 tex2Dblur5resize +// 59.7 68.7 115.4 162.1 tex2Dblur5x5 +// 64.9 77.2 99.1 137.2 tex2Dblur6x6shared +// 55.8 63.7 110.4 151.8 tex2Dblur7fast +// 39.8 43.9 83.9 105.8 tex2Dblur7resize +// 40.0 44.2 83.2 104.9 tex2Dblur7x7 +// 56.4 65.5 71.9 87.9 tex2Dblur8x8shared +// 49.3 55.1 99.9 132.5 tex2Dblur9fast +// 33.3 36.2 72.4 88.0 tex2Dblur9resize +// 27.8 29.7 61.3 72.2 tex2Dblur9x9 +// 37.2 41.1 52.6 60.2 tex2Dblur10x10shared +// 44.4 49.5 91.3 117.8 tex2Dblur11fast +// 28.8 30.8 63.6 75.4 tex2Dblur11resize +// 33.6 36.5 40.9 45.5 tex2Dblur12x12shared +// TODO: Fill in benchmarks for new untested blurs. +// tex2Dblur17fast +// tex2Dblur25fast +// tex2Dblur31fast +// tex2Dblur43fast +// tex2Dblur3x3resize + + +///////////////////////////// SETTINGS MANAGEMENT //////////////////////////// + +// Set static standard deviations, but allow users to override them with their +// own constants (even non-static uniforms if they're okay with the speed hit): +#ifndef OVERRIDE_BLUR_STD_DEVS + // blurN_std_dev values are specified in terms of dxdy strides. + #ifdef USE_BINOMIAL_BLUR_STD_DEVS + // By request, we can define standard deviations corresponding to a + // binomial distribution with p = 0.5 (related to Pascal's triangle). + // This distribution works such that blurring multiple times should + // have the same result as a single larger blur. These values are + // larger than default for blurs up to 6x and smaller thereafter. + const float blur3_std_dev = 0.84931640625; + const float blur4_std_dev = 0.84931640625; + const float blur5_std_dev = 1.0595703125; + const float blur6_std_dev = 1.06591796875; + const float blur7_std_dev = 1.17041015625; + const float blur8_std_dev = 1.1720703125; + const float blur9_std_dev = 1.2259765625; + const float blur10_std_dev = 1.21982421875; + const float blur11_std_dev = 1.25361328125; + const float blur12_std_dev = 1.2423828125; + const float blur17_std_dev = 1.27783203125; + const float blur25_std_dev = 1.2810546875; + const float blur31_std_dev = 1.28125; + const float blur43_std_dev = 1.28125; + #else + // The defaults are the largest values that keep the largest unused + // blur term on each side <= 1.0/256.0. (We could get away with more + // or be more conservative, but this compromise is pretty reasonable.) + const float blur3_std_dev = 0.62666015625; + const float blur4_std_dev = 0.66171875; + const float blur5_std_dev = 0.9845703125; + const float blur6_std_dev = 1.02626953125; + const float blur7_std_dev = 1.36103515625; + const float blur8_std_dev = 1.4080078125; + const float blur9_std_dev = 1.7533203125; + const float blur10_std_dev = 1.80478515625; + const float blur11_std_dev = 2.15986328125; + const float blur12_std_dev = 2.215234375; + const float blur17_std_dev = 3.45535583496; + const float blur25_std_dev = 5.3409576416; + const float blur31_std_dev = 6.86488037109; + const float blur43_std_dev = 10.1852050781; + #endif // USE_BINOMIAL_BLUR_STD_DEVS +#endif // OVERRIDE_BLUR_STD_DEVS + +#ifndef OVERRIDE_ERROR_BLURRING + // error_blurring should be in [0.0, 1.0]. Higher values reduce ringing + // in shared-sample blurs but increase blurring and feature shifting. + const float error_blurring = 0.5; +#endif + + +////////////////////////////////// INCLUDES ////////////////////////////////// + +// gamma-management.h relies on pass-specific settings to guide its behavior: +// FIRST_PASS, LAST_PASS, GAMMA_ENCODE_EVERY_FBO, etc. See it for details. +//#include "gamma-management.h" +#include "quad-pixel-communication.h" +#include "special-functions.h" + + +/////////////////////////////////// HELPERS ////////////////////////////////// + +vec4 uv2_to_uv4(vec2 tex_uv) +{ + // Make a vec2 uv offset safe for adding to vec4 tex2Dlod coords: + return vec4(tex_uv, 0.0, 0.0); +} + +// Make a length squared helper macro (for usage with static constants): +#define LENGTH_SQ(vec) (dot(vec, vec)) + +float get_fast_gaussian_weight_sum_inv(const float sigma) +{ + // We can use the Gaussian integral to calculate the asymptotic weight for + // the center pixel. Since the unnormalized center pixel weight is 1.0, + // the normalized weight is the same as the weight sum inverse. Given a + // large enough blur (9+), the asymptotic weight sum is close and faster: + // center_weight = 0.5 * + // (erf(0.5/(sigma*sqrt(2.0))) - erf(-0.5/(sigma*sqrt(2.0)))) + // erf(-x) == -erf(x), so we get 0.5 * (2.0 * erf(blah blah)): + // However, we can get even faster results with curve-fitting. These are + // also closer than the asymptotic results, because they were constructed + // from 64 blurs sizes from [3, 131) and 255 equally-spaced sigmas from + // (0, blurN_std_dev), so the results for smaller sigmas are biased toward + // smaller blurs. The max error is 0.0031793913. + // Relative FPS: 134.3 with erf, 135.8 with curve-fitting. + //static const float temp = 0.5/sqrt(2.0); + //return erf(temp/sigma); + return min(exp(exp(0.348348412457428/ + (sigma - 0.0860587260734721))), 0.399334576340352/sigma); +} + + +//////////////////// ARBITRARILY RESIZABLE SEPARABLE BLURS /////////////////// + +vec3 tex2Dblur11resize(const sampler2D texture, const vec2 tex_uv, + const vec2 dxdy, const float sigma) +{ + // Requires: Global requirements must be met (see file description). + // Returns: A 1D 11x Gaussian blurred texture lookup using a 11-tap blur. + // It may be mipmapped depending on settings and dxdy. + // Calculate Gaussian blur kernel weights and a normalization factor for + // distances of 0-4, ignoring constant factors (since we're normalizing). + const float denom_inv = 0.5/(sigma*sigma); + const float w0 = 1.0; + const float w1 = exp(-1.0 * denom_inv); + const float w2 = exp(-4.0 * denom_inv); + const float w3 = exp(-9.0 * denom_inv); + const float w4 = exp(-16.0 * denom_inv); + const float w5 = exp(-25.0 * denom_inv); + const float weight_sum_inv = 1.0 / + (w0 + 2.0 * (w1 + w2 + w3 + w4 + w5)); + // Statically normalize weights, sum weighted samples, and return. Blurs are + // currently optimized for dynamic weights. + vec3 sum = vec3(0.0); + sum += w5 * tex2D_linearize(texture, tex_uv - 5.0 * dxdy).rgb; + sum += w4 * tex2D_linearize(texture, tex_uv - 4.0 * dxdy).rgb; + sum += w3 * tex2D_linearize(texture, tex_uv - 3.0 * dxdy).rgb; + sum += w2 * tex2D_linearize(texture, tex_uv - 2.0 * dxdy).rgb; + sum += w1 * tex2D_linearize(texture, tex_uv - 1.0 * dxdy).rgb; + sum += w0 * tex2D_linearize(texture, tex_uv).rgb; + sum += w1 * tex2D_linearize(texture, tex_uv + 1.0 * dxdy).rgb; + sum += w2 * tex2D_linearize(texture, tex_uv + 2.0 * dxdy).rgb; + sum += w3 * tex2D_linearize(texture, tex_uv + 3.0 * dxdy).rgb; + sum += w4 * tex2D_linearize(texture, tex_uv + 4.0 * dxdy).rgb; + sum += w5 * tex2D_linearize(texture, tex_uv + 5.0 * dxdy).rgb; + return sum * weight_sum_inv; +} + +vec3 tex2Dblur9resize(const sampler2D texture, const vec2 tex_uv, + const vec2 dxdy, const float sigma) +{ + // Requires: Global requirements must be met (see file description). + // Returns: A 1D 9x Gaussian blurred texture lookup using a 9-tap blur. + // It may be mipmapped depending on settings and dxdy. + // First get the texel weights and normalization factor as above. + const float denom_inv = 0.5/(sigma*sigma); + const float w0 = 1.0; + const float w1 = exp(-1.0 * denom_inv); + const float w2 = exp(-4.0 * denom_inv); + const float w3 = exp(-9.0 * denom_inv); + const float w4 = exp(-16.0 * denom_inv); + const float weight_sum_inv = 1.0 / (w0 + 2.0 * (w1 + w2 + w3 + w4)); + // Statically normalize weights, sum weighted samples, and return: + vec3 sum = vec3(0.0); + sum += w4 * tex2D_linearize(texture, tex_uv - 4.0 * dxdy).rgb; + sum += w3 * tex2D_linearize(texture, tex_uv - 3.0 * dxdy).rgb; + sum += w2 * tex2D_linearize(texture, tex_uv - 2.0 * dxdy).rgb; + sum += w1 * tex2D_linearize(texture, tex_uv - 1.0 * dxdy).rgb; + sum += w0 * tex2D_linearize(texture, tex_uv).rgb; + sum += w1 * tex2D_linearize(texture, tex_uv + 1.0 * dxdy).rgb; + sum += w2 * tex2D_linearize(texture, tex_uv + 2.0 * dxdy).rgb; + sum += w3 * tex2D_linearize(texture, tex_uv + 3.0 * dxdy).rgb; + sum += w4 * tex2D_linearize(texture, tex_uv + 4.0 * dxdy).rgb; + return sum * weight_sum_inv; +} + +vec3 tex2Dblur7resize(const sampler2D texture, const vec2 tex_uv, + const vec2 dxdy, const float sigma) +{ + // Requires: Global requirements must be met (see file description). + // Returns: A 1D 7x Gaussian blurred texture lookup using a 7-tap blur. + // It may be mipmapped depending on settings and dxdy. + // First get the texel weights and normalization factor as above. + const float denom_inv = 0.5/(sigma*sigma); + const float w0 = 1.0; + const float w1 = exp(-1.0 * denom_inv); + const float w2 = exp(-4.0 * denom_inv); + const float w3 = exp(-9.0 * denom_inv); + const float weight_sum_inv = 1.0 / (w0 + 2.0 * (w1 + w2 + w3)); + // Statically normalize weights, sum weighted samples, and return: + vec3 sum = vec3(0.0); + sum += w3 * tex2D_linearize(texture, tex_uv - 3.0 * dxdy).rgb; + sum += w2 * tex2D_linearize(texture, tex_uv - 2.0 * dxdy).rgb; + sum += w1 * tex2D_linearize(texture, tex_uv - 1.0 * dxdy).rgb; + sum += w0 * tex2D_linearize(texture, tex_uv).rgb; + sum += w1 * tex2D_linearize(texture, tex_uv + 1.0 * dxdy).rgb; + sum += w2 * tex2D_linearize(texture, tex_uv + 2.0 * dxdy).rgb; + sum += w3 * tex2D_linearize(texture, tex_uv + 3.0 * dxdy).rgb; + return sum * weight_sum_inv; +} + +vec3 tex2Dblur5resize(const sampler2D texture, const vec2 tex_uv, + const vec2 dxdy, const float sigma) +{ + // Requires: Global requirements must be met (see file description). + // Returns: A 1D 5x Gaussian blurred texture lookup using a 5-tap blur. + // It may be mipmapped depending on settings and dxdy. + // First get the texel weights and normalization factor as above. + const float denom_inv = 0.5/(sigma*sigma); + const float w0 = 1.0; + const float w1 = exp(-1.0 * denom_inv); + const float w2 = exp(-4.0 * denom_inv); + const float weight_sum_inv = 1.0 / (w0 + 2.0 * (w1 + w2)); + // Statically normalize weights, sum weighted samples, and return: + vec3 sum = vec3(0.0); + sum += w2 * tex2D_linearize(texture, tex_uv - 2.0 * dxdy).rgb; + sum += w1 * tex2D_linearize(texture, tex_uv - 1.0 * dxdy).rgb; + sum += w0 * tex2D_linearize(texture, tex_uv).rgb; + sum += w1 * tex2D_linearize(texture, tex_uv + 1.0 * dxdy).rgb; + sum += w2 * tex2D_linearize(texture, tex_uv + 2.0 * dxdy).rgb; + return sum * weight_sum_inv; +} + +vec3 tex2Dblur3resize(const sampler2D texture, const vec2 tex_uv, + const vec2 dxdy, const float sigma) +{ + // Requires: Global requirements must be met (see file description). + // Returns: A 1D 3x Gaussian blurred texture lookup using a 3-tap blur. + // It may be mipmapped depending on settings and dxdy. + // First get the texel weights and normalization factor as above. + const float denom_inv = 0.5/(sigma*sigma); + const float w0 = 1.0; + const float w1 = exp(-1.0 * denom_inv); + const float weight_sum_inv = 1.0 / (w0 + 2.0 * w1); + // Statically normalize weights, sum weighted samples, and return: + vec3 sum = vec3(0.0); + sum += w1 * tex2D_linearize(texture, tex_uv - 1.0 * dxdy).rgb; + sum += w0 * tex2D_linearize(texture, tex_uv).rgb; + sum += w1 * tex2D_linearize(texture, tex_uv + 1.0 * dxdy).rgb; + return sum * weight_sum_inv; +} + + +/////////////////////////// FAST SEPARABLE BLURS /////////////////////////// + +vec3 tex2Dblur11fast(const sampler2D texture, const vec2 tex_uv, + const vec2 dxdy, const float sigma) +{ + // Requires: 1.) Global requirements must be met (see file description). + // 2.) filter_linearN must = "true" in your .cgp file. + // 3.) For gamma-correct bilinear filtering, global + // gamma_aware_bilinear == true (from gamma-management.h) + // Returns: A 1D 11x Gaussian blurred texture lookup using 6 linear + // taps. It may be mipmapped depending on settings and dxdy. + // First get the texel weights and normalization factor as above. + const float denom_inv = 0.5/(sigma*sigma); + const float w0 = 1.0; + const float w1 = exp(-1.0 * denom_inv); + const float w2 = exp(-4.0 * denom_inv); + const float w3 = exp(-9.0 * denom_inv); + const float w4 = exp(-16.0 * denom_inv); + const float w5 = exp(-25.0 * denom_inv); + const float weight_sum_inv = 1.0 / + (w0 + 2.0 * (w1 + w2 + w3 + w4 + w5)); + // Calculate combined weights and linear sample ratios between texel pairs. + // The center texel (with weight w0) is used twice, so halve its weight. + const float w01 = w0 * 0.5 + w1; + const float w23 = w2 + w3; + const float w45 = w4 + w5; + const float w01_ratio = w1/w01; + const float w23_ratio = w3/w23; + const float w45_ratio = w5/w45; + // Statically normalize weights, sum weighted samples, and return: + vec3 sum = vec3(0.0); + sum += w45 * tex2D_linearize(texture, tex_uv - (4.0 + w45_ratio) * dxdy).rgb; + sum += w23 * tex2D_linearize(texture, tex_uv - (2.0 + w23_ratio) * dxdy).rgb; + sum += w01 * tex2D_linearize(texture, tex_uv - w01_ratio * dxdy).rgb; + sum += w01 * tex2D_linearize(texture, tex_uv + w01_ratio * dxdy).rgb; + sum += w23 * tex2D_linearize(texture, tex_uv + (2.0 + w23_ratio) * dxdy).rgb; + sum += w45 * tex2D_linearize(texture, tex_uv + (4.0 + w45_ratio) * dxdy).rgb; + return sum * weight_sum_inv; +} + +vec3 tex2Dblur9fast(const sampler2D texture, const vec2 tex_uv, + const vec2 dxdy, const float sigma) +{ + // Requires: Same as tex2Dblur11() + // Returns: A 1D 9x Gaussian blurred texture lookup using 1 nearest + // neighbor and 4 linear taps. It may be mipmapped depending + // on settings and dxdy. + // First get the texel weights and normalization factor as above. + const float denom_inv = 0.5/(sigma*sigma); + const float w0 = 1.0; + const float w1 = exp(-1.0 * denom_inv); + const float w2 = exp(-4.0 * denom_inv); + const float w3 = exp(-9.0 * denom_inv); + const float w4 = exp(-16.0 * denom_inv); + const float weight_sum_inv = 1.0 / (w0 + 2.0 * (w1 + w2 + w3 + w4)); + // Calculate combined weights and linear sample ratios between texel pairs. + const float w12 = w1 + w2; + const float w34 = w3 + w4; + const float w12_ratio = w2/w12; + const float w34_ratio = w4/w34; + // Statically normalize weights, sum weighted samples, and return: + vec3 sum = vec3(0.0); + sum += w34 * tex2D_linearize(texture, tex_uv - (3.0 + w34_ratio) * dxdy).rgb; + sum += w12 * tex2D_linearize(texture, tex_uv - (1.0 + w12_ratio) * dxdy).rgb; + sum += w0 * tex2D_linearize(texture, tex_uv).rgb; + sum += w12 * tex2D_linearize(texture, tex_uv + (1.0 + w12_ratio) * dxdy).rgb; + sum += w34 * tex2D_linearize(texture, tex_uv + (3.0 + w34_ratio) * dxdy).rgb; + return sum * weight_sum_inv; +} + +vec3 tex2Dblur7fast(const sampler2D texture, const vec2 tex_uv, + const vec2 dxdy, const float sigma) +{ + // Requires: Same as tex2Dblur11() + // Returns: A 1D 7x Gaussian blurred texture lookup using 4 linear + // taps. It may be mipmapped depending on settings and dxdy. + // First get the texel weights and normalization factor as above. + const float denom_inv = 0.5/(sigma*sigma); + const float w0 = 1.0; + const float w1 = exp(-1.0 * denom_inv); + const float w2 = exp(-4.0 * denom_inv); + const float w3 = exp(-9.0 * denom_inv); + const float weight_sum_inv = 1.0 / (w0 + 2.0 * (w1 + w2 + w3)); + // Calculate combined weights and linear sample ratios between texel pairs. + // The center texel (with weight w0) is used twice, so halve its weight. + const float w01 = w0 * 0.5 + w1; + const float w23 = w2 + w3; + const float w01_ratio = w1/w01; + const float w23_ratio = w3/w23; + // Statically normalize weights, sum weighted samples, and return: + vec3 sum = vec3(0.0); + sum += w23 * tex2D_linearize(texture, tex_uv - (2.0 + w23_ratio) * dxdy).rgb; + sum += w01 * tex2D_linearize(texture, tex_uv - w01_ratio * dxdy).rgb; + sum += w01 * tex2D_linearize(texture, tex_uv + w01_ratio * dxdy).rgb; + sum += w23 * tex2D_linearize(texture, tex_uv + (2.0 + w23_ratio) * dxdy).rgb; + return sum * weight_sum_inv; +} + +vec3 tex2Dblur5fast(const sampler2D texture, const vec2 tex_uv, + const vec2 dxdy, const float sigma) +{ + // Requires: Same as tex2Dblur11() + // Returns: A 1D 5x Gaussian blurred texture lookup using 1 nearest + // neighbor and 2 linear taps. It may be mipmapped depending + // on settings and dxdy. + // First get the texel weights and normalization factor as above. + const float denom_inv = 0.5/(sigma*sigma); + const float w0 = 1.0; + const float w1 = exp(-1.0 * denom_inv); + const float w2 = exp(-4.0 * denom_inv); + const float weight_sum_inv = 1.0 / (w0 + 2.0 * (w1 + w2)); + // Calculate combined weights and linear sample ratios between texel pairs. + const float w12 = w1 + w2; + const float w12_ratio = w2/w12; + // Statically normalize weights, sum weighted samples, and return: + vec3 sum = vec3(0.0); + sum += w12 * tex2D_linearize(texture, tex_uv - (1.0 + w12_ratio) * dxdy).rgb; + sum += w0 * tex2D_linearize(texture, tex_uv).rgb; + sum += w12 * tex2D_linearize(texture, tex_uv + (1.0 + w12_ratio) * dxdy).rgb; + return sum * weight_sum_inv; +} + +vec3 tex2Dblur3fast(const sampler2D texture, const vec2 tex_uv, + const vec2 dxdy, const float sigma) +{ + // Requires: Same as tex2Dblur11() + // Returns: A 1D 3x Gaussian blurred texture lookup using 2 linear + // taps. It may be mipmapped depending on settings and dxdy. + // First get the texel weights and normalization factor as above. + const float denom_inv = 0.5/(sigma*sigma); + const float w0 = 1.0; + const float w1 = exp(-1.0 * denom_inv); + const float weight_sum_inv = 1.0 / (w0 + 2.0 * w1); + // Calculate combined weights and linear sample ratios between texel pairs. + // The center texel (with weight w0) is used twice, so halve its weight. + const float w01 = w0 * 0.5 + w1; + const float w01_ratio = w1/w01; + // Weights for all samples are the same, so just average them: + return 0.5 * ( + tex2D_linearize(texture, tex_uv - w01_ratio * dxdy).rgb + + tex2D_linearize(texture, tex_uv + w01_ratio * dxdy).rgb); +} + + +//////////////////////////// HUGE SEPARABLE BLURS //////////////////////////// + +// Huge separable blurs come only in "fast" versions. +vec3 tex2Dblur43fast(const sampler2D texture, const vec2 tex_uv, + const vec2 dxdy, const float sigma) +{ + // Requires: Same as tex2Dblur11() + // Returns: A 1D 43x Gaussian blurred texture lookup using 22 linear + // taps. It may be mipmapped depending on settings and dxdy. + // First get the texel weights and normalization factor as above. + const float denom_inv = 0.5/(sigma*sigma); + const float w0 = 1.0; + const float w1 = exp(-1.0 * denom_inv); + const float w2 = exp(-4.0 * denom_inv); + const float w3 = exp(-9.0 * denom_inv); + const float w4 = exp(-16.0 * denom_inv); + const float w5 = exp(-25.0 * denom_inv); + const float w6 = exp(-36.0 * denom_inv); + const float w7 = exp(-49.0 * denom_inv); + const float w8 = exp(-64.0 * denom_inv); + const float w9 = exp(-81.0 * denom_inv); + const float w10 = exp(-100.0 * denom_inv); + const float w11 = exp(-121.0 * denom_inv); + const float w12 = exp(-144.0 * denom_inv); + const float w13 = exp(-169.0 * denom_inv); + const float w14 = exp(-196.0 * denom_inv); + const float w15 = exp(-225.0 * denom_inv); + const float w16 = exp(-256.0 * denom_inv); + const float w17 = exp(-289.0 * denom_inv); + const float w18 = exp(-324.0 * denom_inv); + const float w19 = exp(-361.0 * denom_inv); + const float w20 = exp(-400.0 * denom_inv); + const float w21 = exp(-441.0 * denom_inv); + //const float weight_sum_inv = 1.0 / + // (w0 + 2.0 * (w1 + w2 + w3 + w4 + w5 + w6 + w7 + w8 + w9 + w10 + w11 + + // w12 + w13 + w14 + w15 + w16 + w17 + w18 + w19 + w20 + w21)); + const float weight_sum_inv = get_fast_gaussian_weight_sum_inv(sigma); + // Calculate combined weights and linear sample ratios between texel pairs. + // The center texel (with weight w0) is used twice, so halve its weight. + const float w0_1 = w0 * 0.5 + w1; + const float w2_3 = w2 + w3; + const float w4_5 = w4 + w5; + const float w6_7 = w6 + w7; + const float w8_9 = w8 + w9; + const float w10_11 = w10 + w11; + const float w12_13 = w12 + w13; + const float w14_15 = w14 + w15; + const float w16_17 = w16 + w17; + const float w18_19 = w18 + w19; + const float w20_21 = w20 + w21; + const float w0_1_ratio = w1/w0_1; + const float w2_3_ratio = w3/w2_3; + const float w4_5_ratio = w5/w4_5; + const float w6_7_ratio = w7/w6_7; + const float w8_9_ratio = w9/w8_9; + const float w10_11_ratio = w11/w10_11; + const float w12_13_ratio = w13/w12_13; + const float w14_15_ratio = w15/w14_15; + const float w16_17_ratio = w17/w16_17; + const float w18_19_ratio = w19/w18_19; + const float w20_21_ratio = w21/w20_21; + // Statically normalize weights, sum weighted samples, and return: + vec3 sum = vec3(0.0); + sum += w20_21 * tex2D_linearize(texture, tex_uv - (20.0 + w20_21_ratio) * dxdy).rgb; + sum += w18_19 * tex2D_linearize(texture, tex_uv - (18.0 + w18_19_ratio) * dxdy).rgb; + sum += w16_17 * tex2D_linearize(texture, tex_uv - (16.0 + w16_17_ratio) * dxdy).rgb; + sum += w14_15 * tex2D_linearize(texture, tex_uv - (14.0 + w14_15_ratio) * dxdy).rgb; + sum += w12_13 * tex2D_linearize(texture, tex_uv - (12.0 + w12_13_ratio) * dxdy).rgb; + sum += w10_11 * tex2D_linearize(texture, tex_uv - (10.0 + w10_11_ratio) * dxdy).rgb; + sum += w8_9 * tex2D_linearize(texture, tex_uv - (8.0 + w8_9_ratio) * dxdy).rgb; + sum += w6_7 * tex2D_linearize(texture, tex_uv - (6.0 + w6_7_ratio) * dxdy).rgb; + sum += w4_5 * tex2D_linearize(texture, tex_uv - (4.0 + w4_5_ratio) * dxdy).rgb; + sum += w2_3 * tex2D_linearize(texture, tex_uv - (2.0 + w2_3_ratio) * dxdy).rgb; + sum += w0_1 * tex2D_linearize(texture, tex_uv - w0_1_ratio * dxdy).rgb; + sum += w0_1 * tex2D_linearize(texture, tex_uv + w0_1_ratio * dxdy).rgb; + sum += w2_3 * tex2D_linearize(texture, tex_uv + (2.0 + w2_3_ratio) * dxdy).rgb; + sum += w4_5 * tex2D_linearize(texture, tex_uv + (4.0 + w4_5_ratio) * dxdy).rgb; + sum += w6_7 * tex2D_linearize(texture, tex_uv + (6.0 + w6_7_ratio) * dxdy).rgb; + sum += w8_9 * tex2D_linearize(texture, tex_uv + (8.0 + w8_9_ratio) * dxdy).rgb; + sum += w10_11 * tex2D_linearize(texture, tex_uv + (10.0 + w10_11_ratio) * dxdy).rgb; + sum += w12_13 * tex2D_linearize(texture, tex_uv + (12.0 + w12_13_ratio) * dxdy).rgb; + sum += w14_15 * tex2D_linearize(texture, tex_uv + (14.0 + w14_15_ratio) * dxdy).rgb; + sum += w16_17 * tex2D_linearize(texture, tex_uv + (16.0 + w16_17_ratio) * dxdy).rgb; + sum += w18_19 * tex2D_linearize(texture, tex_uv + (18.0 + w18_19_ratio) * dxdy).rgb; + sum += w20_21 * tex2D_linearize(texture, tex_uv + (20.0 + w20_21_ratio) * dxdy).rgb; + return sum * weight_sum_inv; +} + +vec3 tex2Dblur31fast(const sampler2D texture, const vec2 tex_uv, + const vec2 dxdy, const float sigma) +{ + // Requires: Same as tex2Dblur11() + // Returns: A 1D 31x Gaussian blurred texture lookup using 16 linear + // taps. It may be mipmapped depending on settings and dxdy. + // First get the texel weights and normalization factor as above. + const float denom_inv = 0.5/(sigma*sigma); + const float w0 = 1.0; + const float w1 = exp(-1.0 * denom_inv); + const float w2 = exp(-4.0 * denom_inv); + const float w3 = exp(-9.0 * denom_inv); + const float w4 = exp(-16.0 * denom_inv); + const float w5 = exp(-25.0 * denom_inv); + const float w6 = exp(-36.0 * denom_inv); + const float w7 = exp(-49.0 * denom_inv); + const float w8 = exp(-64.0 * denom_inv); + const float w9 = exp(-81.0 * denom_inv); + const float w10 = exp(-100.0 * denom_inv); + const float w11 = exp(-121.0 * denom_inv); + const float w12 = exp(-144.0 * denom_inv); + const float w13 = exp(-169.0 * denom_inv); + const float w14 = exp(-196.0 * denom_inv); + const float w15 = exp(-225.0 * denom_inv); + //const float weight_sum_inv = 1.0 / + // (w0 + 2.0 * (w1 + w2 + w3 + w4 + w5 + w6 + w7 + w8 + + // w9 + w10 + w11 + w12 + w13 + w14 + w15)); + const float weight_sum_inv = get_fast_gaussian_weight_sum_inv(sigma); + // Calculate combined weights and linear sample ratios between texel pairs. + // The center texel (with weight w0) is used twice, so halve its weight. + const float w0_1 = w0 * 0.5 + w1; + const float w2_3 = w2 + w3; + const float w4_5 = w4 + w5; + const float w6_7 = w6 + w7; + const float w8_9 = w8 + w9; + const float w10_11 = w10 + w11; + const float w12_13 = w12 + w13; + const float w14_15 = w14 + w15; + const float w0_1_ratio = w1/w0_1; + const float w2_3_ratio = w3/w2_3; + const float w4_5_ratio = w5/w4_5; + const float w6_7_ratio = w7/w6_7; + const float w8_9_ratio = w9/w8_9; + const float w10_11_ratio = w11/w10_11; + const float w12_13_ratio = w13/w12_13; + const float w14_15_ratio = w15/w14_15; + // Statically normalize weights, sum weighted samples, and return: + vec3 sum = vec3(0.0); + sum += w14_15 * tex2D_linearize(texture, tex_uv - (14.0 + w14_15_ratio) * dxdy).rgb; + sum += w12_13 * tex2D_linearize(texture, tex_uv - (12.0 + w12_13_ratio) * dxdy).rgb; + sum += w10_11 * tex2D_linearize(texture, tex_uv - (10.0 + w10_11_ratio) * dxdy).rgb; + sum += w8_9 * tex2D_linearize(texture, tex_uv - (8.0 + w8_9_ratio) * dxdy).rgb; + sum += w6_7 * tex2D_linearize(texture, tex_uv - (6.0 + w6_7_ratio) * dxdy).rgb; + sum += w4_5 * tex2D_linearize(texture, tex_uv - (4.0 + w4_5_ratio) * dxdy).rgb; + sum += w2_3 * tex2D_linearize(texture, tex_uv - (2.0 + w2_3_ratio) * dxdy).rgb; + sum += w0_1 * tex2D_linearize(texture, tex_uv - w0_1_ratio * dxdy).rgb; + sum += w0_1 * tex2D_linearize(texture, tex_uv + w0_1_ratio * dxdy).rgb; + sum += w2_3 * tex2D_linearize(texture, tex_uv + (2.0 + w2_3_ratio) * dxdy).rgb; + sum += w4_5 * tex2D_linearize(texture, tex_uv + (4.0 + w4_5_ratio) * dxdy).rgb; + sum += w6_7 * tex2D_linearize(texture, tex_uv + (6.0 + w6_7_ratio) * dxdy).rgb; + sum += w8_9 * tex2D_linearize(texture, tex_uv + (8.0 + w8_9_ratio) * dxdy).rgb; + sum += w10_11 * tex2D_linearize(texture, tex_uv + (10.0 + w10_11_ratio) * dxdy).rgb; + sum += w12_13 * tex2D_linearize(texture, tex_uv + (12.0 + w12_13_ratio) * dxdy).rgb; + sum += w14_15 * tex2D_linearize(texture, tex_uv + (14.0 + w14_15_ratio) * dxdy).rgb; + return sum * weight_sum_inv; +} + +vec3 tex2Dblur25fast(const sampler2D texture, const vec2 tex_uv, + const vec2 dxdy, const float sigma) +{ + // Requires: Same as tex2Dblur11() + // Returns: A 1D 25x Gaussian blurred texture lookup using 1 nearest + // neighbor and 12 linear taps. It may be mipmapped depending + // on settings and dxdy. + // First get the texel weights and normalization factor as above. + const float denom_inv = 0.5/(sigma*sigma); + const float w0 = 1.0; + const float w1 = exp(-1.0 * denom_inv); + const float w2 = exp(-4.0 * denom_inv); + const float w3 = exp(-9.0 * denom_inv); + const float w4 = exp(-16.0 * denom_inv); + const float w5 = exp(-25.0 * denom_inv); + const float w6 = exp(-36.0 * denom_inv); + const float w7 = exp(-49.0 * denom_inv); + const float w8 = exp(-64.0 * denom_inv); + const float w9 = exp(-81.0 * denom_inv); + const float w10 = exp(-100.0 * denom_inv); + const float w11 = exp(-121.0 * denom_inv); + const float w12 = exp(-144.0 * denom_inv); + //const float weight_sum_inv = 1.0 / (w0 + 2.0 * ( + // w1 + w2 + w3 + w4 + w5 + w6 + w7 + w8 + w9 + w10 + w11 + w12)); + const float weight_sum_inv = get_fast_gaussian_weight_sum_inv(sigma); + // Calculate combined weights and linear sample ratios between texel pairs. + const float w1_2 = w1 + w2; + const float w3_4 = w3 + w4; + const float w5_6 = w5 + w6; + const float w7_8 = w7 + w8; + const float w9_10 = w9 + w10; + const float w11_12 = w11 + w12; + const float w1_2_ratio = w2/w1_2; + const float w3_4_ratio = w4/w3_4; + const float w5_6_ratio = w6/w5_6; + const float w7_8_ratio = w8/w7_8; + const float w9_10_ratio = w10/w9_10; + const float w11_12_ratio = w12/w11_12; + // Statically normalize weights, sum weighted samples, and return: + vec3 sum = vec3(0.0); + sum += w11_12 * tex2D_linearize(texture, tex_uv - (11.0 + w11_12_ratio) * dxdy).rgb; + sum += w9_10 * tex2D_linearize(texture, tex_uv - (9.0 + w9_10_ratio) * dxdy).rgb; + sum += w7_8 * tex2D_linearize(texture, tex_uv - (7.0 + w7_8_ratio) * dxdy).rgb; + sum += w5_6 * tex2D_linearize(texture, tex_uv - (5.0 + w5_6_ratio) * dxdy).rgb; + sum += w3_4 * tex2D_linearize(texture, tex_uv - (3.0 + w3_4_ratio) * dxdy).rgb; + sum += w1_2 * tex2D_linearize(texture, tex_uv - (1.0 + w1_2_ratio) * dxdy).rgb; + sum += w0 * tex2D_linearize(texture, tex_uv).rgb; + sum += w1_2 * tex2D_linearize(texture, tex_uv + (1.0 + w1_2_ratio) * dxdy).rgb; + sum += w3_4 * tex2D_linearize(texture, tex_uv + (3.0 + w3_4_ratio) * dxdy).rgb; + sum += w5_6 * tex2D_linearize(texture, tex_uv + (5.0 + w5_6_ratio) * dxdy).rgb; + sum += w7_8 * tex2D_linearize(texture, tex_uv + (7.0 + w7_8_ratio) * dxdy).rgb; + sum += w9_10 * tex2D_linearize(texture, tex_uv + (9.0 + w9_10_ratio) * dxdy).rgb; + sum += w11_12 * tex2D_linearize(texture, tex_uv + (11.0 + w11_12_ratio) * dxdy).rgb; + return sum * weight_sum_inv; +} + +vec3 tex2Dblur17fast(const sampler2D texture, const vec2 tex_uv, + const vec2 dxdy, const float sigma) +{ + // Requires: Same as tex2Dblur11() + // Returns: A 1D 17x Gaussian blurred texture lookup using 1 nearest + // neighbor and 8 linear taps. It may be mipmapped depending + // on settings and dxdy. + // First get the texel weights and normalization factor as above. + const float denom_inv = 0.5/(sigma*sigma); + const float w0 = 1.0; + const float w1 = exp(-1.0 * denom_inv); + const float w2 = exp(-4.0 * denom_inv); + const float w3 = exp(-9.0 * denom_inv); + const float w4 = exp(-16.0 * denom_inv); + const float w5 = exp(-25.0 * denom_inv); + const float w6 = exp(-36.0 * denom_inv); + const float w7 = exp(-49.0 * denom_inv); + const float w8 = exp(-64.0 * denom_inv); + //const float weight_sum_inv = 1.0 / (w0 + 2.0 * ( + // w1 + w2 + w3 + w4 + w5 + w6 + w7 + w8)); + const float weight_sum_inv = get_fast_gaussian_weight_sum_inv(sigma); + // Calculate combined weights and linear sample ratios between texel pairs. + const float w1_2 = w1 + w2; + const float w3_4 = w3 + w4; + const float w5_6 = w5 + w6; + const float w7_8 = w7 + w8; + const float w1_2_ratio = w2/w1_2; + const float w3_4_ratio = w4/w3_4; + const float w5_6_ratio = w6/w5_6; + const float w7_8_ratio = w8/w7_8; + // Statically normalize weights, sum weighted samples, and return: + vec3 sum = vec3(0.0); + sum += w7_8 * tex2D_linearize(texture, tex_uv - (7.0 + w7_8_ratio) * dxdy).rgb; + sum += w5_6 * tex2D_linearize(texture, tex_uv - (5.0 + w5_6_ratio) * dxdy).rgb; + sum += w3_4 * tex2D_linearize(texture, tex_uv - (3.0 + w3_4_ratio) * dxdy).rgb; + sum += w1_2 * tex2D_linearize(texture, tex_uv - (1.0 + w1_2_ratio) * dxdy).rgb; + sum += w0 * tex2D_linearize(texture, tex_uv).rgb; + sum += w1_2 * tex2D_linearize(texture, tex_uv + (1.0 + w1_2_ratio) * dxdy).rgb; + sum += w3_4 * tex2D_linearize(texture, tex_uv + (3.0 + w3_4_ratio) * dxdy).rgb; + sum += w5_6 * tex2D_linearize(texture, tex_uv + (5.0 + w5_6_ratio) * dxdy).rgb; + sum += w7_8 * tex2D_linearize(texture, tex_uv + (7.0 + w7_8_ratio) * dxdy).rgb; + return sum * weight_sum_inv; +} + + +//////////////////// ARBITRARILY RESIZABLE ONE-PASS BLURS //////////////////// + +vec3 tex2Dblur3x3resize(const sampler2D texture, const vec2 tex_uv, + const vec2 dxdy, const float sigma) +{ + // Requires: Global requirements must be met (see file description). + // Returns: A 3x3 Gaussian blurred mipmapped texture lookup of the + // resized input. + // Description: + // This is the only arbitrarily resizable one-pass blur; tex2Dblur5x5resize + // would perform like tex2Dblur9x9, MUCH slower than tex2Dblur5resize. + const float denom_inv = 0.5/(sigma*sigma); + // Load each sample. We need all 3x3 samples. Quad-pixel communication + // won't help either: This should perform like tex2Dblur5x5, but sharing a + // 4x4 sample field would perform more like tex2Dblur8x8shared (worse). + const vec2 sample4_uv = tex_uv; + const vec2 dx = vec2(dxdy.x, 0.0); + const vec2 dy = vec2(0.0, dxdy.y); + const vec2 sample1_uv = sample4_uv - dy; + const vec2 sample7_uv = sample4_uv + dy; + const vec3 sample0 = tex2D_linearize(texture, sample1_uv - dx).rgb; + const vec3 sample1 = tex2D_linearize(texture, sample1_uv).rgb; + const vec3 sample2 = tex2D_linearize(texture, sample1_uv + dx).rgb; + const vec3 sample3 = tex2D_linearize(texture, sample4_uv - dx).rgb; + const vec3 sample4 = tex2D_linearize(texture, sample4_uv).rgb; + const vec3 sample5 = tex2D_linearize(texture, sample4_uv + dx).rgb; + const vec3 sample6 = tex2D_linearize(texture, sample7_uv - dx).rgb; + const vec3 sample7 = tex2D_linearize(texture, sample7_uv).rgb; + const vec3 sample8 = tex2D_linearize(texture, sample7_uv + dx).rgb; + // Statically compute Gaussian sample weights: + const float w4 = 1.0; + const float w1_3_5_7 = exp(-LENGTH_SQ(vec2(1.0, 0.0)) * denom_inv); + const float w0_2_6_8 = exp(-LENGTH_SQ(vec2(1.0, 1.0)) * denom_inv); + const float weight_sum_inv = 1.0/(w4 + 4.0 * (w1_3_5_7 + w0_2_6_8)); + // Weight and sum the samples: + const vec3 sum = w4 * sample4 + + w1_3_5_7 * (sample1 + sample3 + sample5 + sample7) + + w0_2_6_8 * (sample0 + sample2 + sample6 + sample8); + return sum * weight_sum_inv; +} + + +//////////////////////////// FASTER ONE-PASS BLURS /////////////////////////// + +vec3 tex2Dblur9x9(const sampler2D texture, const vec2 tex_uv, + const vec2 dxdy, const float sigma) +{ + // Perform a 1-pass 9x9 blur with 5x5 bilinear samples. + // Requires: Same as tex2Dblur9() + // Returns: A 9x9 Gaussian blurred mipmapped texture lookup composed of + // 5x5 carefully selected bilinear samples. + // Description: + // Perform a 1-pass 9x9 blur with 5x5 bilinear samples. Adjust the + // bilinear sample location to reflect the true Gaussian weights for each + // underlying texel. The following diagram illustrates the relative + // locations of bilinear samples. Each sample with the same number has the + // same weight (notice the symmetry). The letters a, b, c, d distinguish + // quadrants, and the letters U, D, L, R, C (up, down, left, right, center) + // distinguish 1D directions along the line containing the pixel center: + // 6a 5a 2U 5b 6b + // 4a 3a 1U 3b 4b + // 2L 1L 0C 1R 2R + // 4c 3c 1D 3d 4d + // 6c 5c 2D 5d 6d + // The following diagram illustrates the underlying equally spaced texels, + // named after the sample that accesses them and subnamed by their location + // within their 2x2, 2x1, 1x2, or 1x1 texel block: + // 6a4 6a3 5a4 5a3 2U2 5b3 5b4 6b3 6b4 + // 6a2 6a1 5a2 5a1 2U1 5b1 5b2 6b1 6b2 + // 4a4 4a3 3a4 3a3 1U2 3b3 3b4 4b3 4b4 + // 4a2 4a1 3a2 3a1 1U1 3b1 3b2 4b1 4b2 + // 2L2 2L1 1L2 1L1 0C1 1R1 1R2 2R1 2R2 + // 4c2 4c1 3c2 3c1 1D1 3d1 3d2 4d1 4d2 + // 4c4 4c3 3c4 3c3 1D2 3d3 3d4 4d3 4d4 + // 6c2 6c1 5c2 5c1 2D1 5d1 5d2 6d1 6d2 + // 6c4 6c3 5c4 5c3 2D2 5d3 5d4 6d3 6d4 + // Note there is only one C texel and only two texels for each U, D, L, or + // R sample. The center sample is effectively a nearest neighbor sample, + // and the U/D/L/R samples use 1D linear filtering. All other texels are + // read with bilinear samples somewhere within their 2x2 texel blocks. + + // COMPUTE TEXTURE COORDS: + // Statically compute sampling offsets within each 2x2 texel block, based + // on 1D sampling ratios between texels [1, 2] and [3, 4] texels away from + // the center, and reuse them independently for both dimensions. Compute + // these offsets based on the relative 1D Gaussian weights of the texels + // in question. (w1off means "Gaussian weight for the texel 1.0 texels + // away from the pixel center," etc.). + const float denom_inv = 0.5/(sigma*sigma); + const float w1off = exp(-1.0 * denom_inv); + const float w2off = exp(-4.0 * denom_inv); + const float w3off = exp(-9.0 * denom_inv); + const float w4off = exp(-16.0 * denom_inv); + const float texel1to2ratio = w2off/(w1off + w2off); + const float texel3to4ratio = w4off/(w3off + w4off); + // Statically compute texel offsets from the fragment center to each + // bilinear sample in the bottom-right quadrant, including x-axis-aligned: + const vec2 sample1R_texel_offset = vec2(1.0, 0.0) + vec2(texel1to2ratio, 0.0); + const vec2 sample2R_texel_offset = vec2(3.0, 0.0) + vec2(texel3to4ratio, 0.0); + const vec2 sample3d_texel_offset = vec2(1.0, 1.0) + vec2(texel1to2ratio, texel1to2ratio); + const vec2 sample4d_texel_offset = vec2(3.0, 1.0) + vec2(texel3to4ratio, texel1to2ratio); + const vec2 sample5d_texel_offset = vec2(1.0, 3.0) + vec2(texel1to2ratio, texel3to4ratio); + const vec2 sample6d_texel_offset = vec2(3.0, 3.0) + vec2(texel3to4ratio, texel3to4ratio); + + // CALCULATE KERNEL WEIGHTS FOR ALL SAMPLES: + // Statically compute Gaussian texel weights for the bottom-right quadrant. + // Read underscores as "and." + const float w1R1 = w1off; + const float w1R2 = w2off; + const float w2R1 = w3off; + const float w2R2 = w4off; + const float w3d1 = exp(-LENGTH_SQ(vec2(1.0, 1.0)) * denom_inv); + const float w3d2_3d3 = exp(-LENGTH_SQ(vec2(2.0, 1.0)) * denom_inv); + const float w3d4 = exp(-LENGTH_SQ(vec2(2.0, 2.0)) * denom_inv); + const float w4d1_5d1 = exp(-LENGTH_SQ(vec2(3.0, 1.0)) * denom_inv); + const float w4d2_5d3 = exp(-LENGTH_SQ(vec2(4.0, 1.0)) * denom_inv); + const float w4d3_5d2 = exp(-LENGTH_SQ(vec2(3.0, 2.0)) * denom_inv); + const float w4d4_5d4 = exp(-LENGTH_SQ(vec2(4.0, 2.0)) * denom_inv); + const float w6d1 = exp(-LENGTH_SQ(vec2(3.0, 3.0)) * denom_inv); + const float w6d2_6d3 = exp(-LENGTH_SQ(vec2(4.0, 3.0)) * denom_inv); + const float w6d4 = exp(-LENGTH_SQ(vec2(4.0, 4.0)) * denom_inv); + // Statically add texel weights in each sample to get sample weights: + const float w0 = 1.0; + const float w1 = w1R1 + w1R2; + const float w2 = w2R1 + w2R2; + const float w3 = w3d1 + 2.0 * w3d2_3d3 + w3d4; + const float w4 = w4d1_5d1 + w4d2_5d3 + w4d3_5d2 + w4d4_5d4; + const float w5 = w4; + const float w6 = w6d1 + 2.0 * w6d2_6d3 + w6d4; + // Get the weight sum inverse (normalization factor): + const float weight_sum_inv = + 1.0/(w0 + 4.0 * (w1 + w2 + w3 + w4 + w5 + w6)); + + // LOAD TEXTURE SAMPLES: + // Load all 25 samples (1 nearest, 8 linear, 16 bilinear) using symmetry: + const vec2 mirror_x = vec2(-1.0, 1.0); + const vec2 mirror_y = vec2(1.0, -1.0); + const vec2 mirror_xy = vec2(-1.0, -1.0); + const vec2 dxdy_mirror_x = dxdy * mirror_x; + const vec2 dxdy_mirror_y = dxdy * mirror_y; + const vec2 dxdy_mirror_xy = dxdy * mirror_xy; + // Sampling order doesn't seem to affect performance, so just be clear: + const vec3 sample0C = tex2D_linearize(texture, tex_uv).rgb; + const vec3 sample1R = tex2D_linearize(texture, tex_uv + dxdy * sample1R_texel_offset).rgb; + const vec3 sample1D = tex2D_linearize(texture, tex_uv + dxdy * sample1R_texel_offset.yx).rgb; + const vec3 sample1L = tex2D_linearize(texture, tex_uv - dxdy * sample1R_texel_offset).rgb; + const vec3 sample1U = tex2D_linearize(texture, tex_uv - dxdy * sample1R_texel_offset.yx).rgb; + const vec3 sample2R = tex2D_linearize(texture, tex_uv + dxdy * sample2R_texel_offset).rgb; + const vec3 sample2D = tex2D_linearize(texture, tex_uv + dxdy * sample2R_texel_offset.yx).rgb; + const vec3 sample2L = tex2D_linearize(texture, tex_uv - dxdy * sample2R_texel_offset).rgb; + const vec3 sample2U = tex2D_linearize(texture, tex_uv - dxdy * sample2R_texel_offset.yx).rgb; + const vec3 sample3d = tex2D_linearize(texture, tex_uv + dxdy * sample3d_texel_offset).rgb; + const vec3 sample3c = tex2D_linearize(texture, tex_uv + dxdy_mirror_x * sample3d_texel_offset).rgb; + const vec3 sample3b = tex2D_linearize(texture, tex_uv + dxdy_mirror_y * sample3d_texel_offset).rgb; + const vec3 sample3a = tex2D_linearize(texture, tex_uv + dxdy_mirror_xy * sample3d_texel_offset).rgb; + const vec3 sample4d = tex2D_linearize(texture, tex_uv + dxdy * sample4d_texel_offset).rgb; + const vec3 sample4c = tex2D_linearize(texture, tex_uv + dxdy_mirror_x * sample4d_texel_offset).rgb; + const vec3 sample4b = tex2D_linearize(texture, tex_uv + dxdy_mirror_y * sample4d_texel_offset).rgb; + const vec3 sample4a = tex2D_linearize(texture, tex_uv + dxdy_mirror_xy * sample4d_texel_offset).rgb; + const vec3 sample5d = tex2D_linearize(texture, tex_uv + dxdy * sample5d_texel_offset).rgb; + const vec3 sample5c = tex2D_linearize(texture, tex_uv + dxdy_mirror_x * sample5d_texel_offset).rgb; + const vec3 sample5b = tex2D_linearize(texture, tex_uv + dxdy_mirror_y * sample5d_texel_offset).rgb; + const vec3 sample5a = tex2D_linearize(texture, tex_uv + dxdy_mirror_xy * sample5d_texel_offset).rgb; + const vec3 sample6d = tex2D_linearize(texture, tex_uv + dxdy * sample6d_texel_offset).rgb; + const vec3 sample6c = tex2D_linearize(texture, tex_uv + dxdy_mirror_x * sample6d_texel_offset).rgb; + const vec3 sample6b = tex2D_linearize(texture, tex_uv + dxdy_mirror_y * sample6d_texel_offset).rgb; + const vec3 sample6a = tex2D_linearize(texture, tex_uv + dxdy_mirror_xy * sample6d_texel_offset).rgb; + + // SUM WEIGHTED SAMPLES: + // Statically normalize weights (so total = 1.0), and sum weighted samples. + vec3 sum = w0 * sample0C; + sum += w1 * (sample1R + sample1D + sample1L + sample1U); + sum += w2 * (sample2R + sample2D + sample2L + sample2U); + sum += w3 * (sample3d + sample3c + sample3b + sample3a); + sum += w4 * (sample4d + sample4c + sample4b + sample4a); + sum += w5 * (sample5d + sample5c + sample5b + sample5a); + sum += w6 * (sample6d + sample6c + sample6b + sample6a); + return sum * weight_sum_inv; +} + +vec3 tex2Dblur7x7(const sampler2D texture, const vec2 tex_uv, + const vec2 dxdy, const float sigma) +{ + // Perform a 1-pass 7x7 blur with 5x5 bilinear samples. + // Requires: Same as tex2Dblur9() + // Returns: A 7x7 Gaussian blurred mipmapped texture lookup composed of + // 4x4 carefully selected bilinear samples. + // Description: + // First see the descriptions for tex2Dblur9x9() and tex2Dblur7(). This + // blur mixes concepts from both. The sample layout is as follows: + // 4a 3a 3b 4b + // 2a 1a 1b 2b + // 2c 1c 1d 2d + // 4c 3c 3d 4d + // The texel layout is as follows. Note that samples 3a/3b, 1a/1b, 1c/1d, + // and 3c/3d share a vertical column of texels, and samples 2a/2c, 1a/1c, + // 1b/1d, and 2b/2d share a horizontal row of texels (all sample1's share + // the center texel): + // 4a4 4a3 3a4 3ab3 3b4 4b3 4b4 + // 4a2 4a1 3a2 3ab1 3b2 4b1 4b2 + // 2a4 2a3 1a4 1ab3 1b4 2b3 2b4 + // 2ac2 2ac1 1ac2 1* 1bd2 2bd1 2bd2 + // 2c4 2c3 1c4 1cd3 1d4 2d3 2d4 + // 4c2 4c1 3c2 3cd1 3d2 4d1 4d2 + // 4c4 4c3 3c4 3cd3 3d4 4d3 4d4 + + // COMPUTE TEXTURE COORDS: + // Statically compute bilinear sampling offsets (details in tex2Dblur9x9). + const float denom_inv = 0.5/(sigma*sigma); + const float w0off = 1.0; + const float w1off = exp(-1.0 * denom_inv); + const float w2off = exp(-4.0 * denom_inv); + const float w3off = exp(-9.0 * denom_inv); + const float texel0to1ratio = w1off/(w0off * 0.5 + w1off); + const float texel2to3ratio = w3off/(w2off + w3off); + // Statically compute texel offsets from the fragment center to each + // bilinear sample in the bottom-right quadrant, including axis-aligned: + const vec2 sample1d_texel_offset = vec2(texel0to1ratio, texel0to1ratio); + const vec2 sample2d_texel_offset = vec2(2.0, 0.0) + vec2(texel2to3ratio, texel0to1ratio); + const vec2 sample3d_texel_offset = vec2(0.0, 2.0) + vec2(texel0to1ratio, texel2to3ratio); + const vec2 sample4d_texel_offset = vec2(2.0, 2.0) + vec2(texel2to3ratio, texel2to3ratio); + + // CALCULATE KERNEL WEIGHTS FOR ALL SAMPLES: + // Statically compute Gaussian texel weights for the bottom-right quadrant. + // Read underscores as "and." + const float w1abcd = 1.0; + const float w1bd2_1cd3 = exp(-LENGTH_SQ(vec2(1.0, 0.0)) * denom_inv); + const float w2bd1_3cd1 = exp(-LENGTH_SQ(vec2(2.0, 0.0)) * denom_inv); + const float w2bd2_3cd2 = exp(-LENGTH_SQ(vec2(3.0, 0.0)) * denom_inv); + const float w1d4 = exp(-LENGTH_SQ(vec2(1.0, 1.0)) * denom_inv); + const float w2d3_3d2 = exp(-LENGTH_SQ(vec2(2.0, 1.0)) * denom_inv); + const float w2d4_3d4 = exp(-LENGTH_SQ(vec2(3.0, 1.0)) * denom_inv); + const float w4d1 = exp(-LENGTH_SQ(vec2(2.0, 2.0)) * denom_inv); + const float w4d2_4d3 = exp(-LENGTH_SQ(vec2(3.0, 2.0)) * denom_inv); + const float w4d4 = exp(-LENGTH_SQ(vec2(3.0, 3.0)) * denom_inv); + // Statically add texel weights in each sample to get sample weights. + // Split weights for shared texels between samples sharing them: + const float w1 = w1abcd * 0.25 + w1bd2_1cd3 + w1d4; + const float w2_3 = (w2bd1_3cd1 + w2bd2_3cd2) * 0.5 + w2d3_3d2 + w2d4_3d4; + const float w4 = w4d1 + 2.0 * w4d2_4d3 + w4d4; + // Get the weight sum inverse (normalization factor): + const float weight_sum_inv = + 1.0/(4.0 * (w1 + 2.0 * w2_3 + w4)); + + // LOAD TEXTURE SAMPLES: + // Load all 16 samples using symmetry: + const vec2 mirror_x = vec2(-1.0, 1.0); + const vec2 mirror_y = vec2(1.0, -1.0); + const vec2 mirror_xy = vec2(-1.0, -1.0); + const vec2 dxdy_mirror_x = dxdy * mirror_x; + const vec2 dxdy_mirror_y = dxdy * mirror_y; + const vec2 dxdy_mirror_xy = dxdy * mirror_xy; + const vec3 sample1a = tex2D_linearize(texture, tex_uv + dxdy_mirror_xy * sample1d_texel_offset).rgb; + const vec3 sample2a = tex2D_linearize(texture, tex_uv + dxdy_mirror_xy * sample2d_texel_offset).rgb; + const vec3 sample3a = tex2D_linearize(texture, tex_uv + dxdy_mirror_xy * sample3d_texel_offset).rgb; + const vec3 sample4a = tex2D_linearize(texture, tex_uv + dxdy_mirror_xy * sample4d_texel_offset).rgb; + const vec3 sample1b = tex2D_linearize(texture, tex_uv + dxdy_mirror_y * sample1d_texel_offset).rgb; + const vec3 sample2b = tex2D_linearize(texture, tex_uv + dxdy_mirror_y * sample2d_texel_offset).rgb; + const vec3 sample3b = tex2D_linearize(texture, tex_uv + dxdy_mirror_y * sample3d_texel_offset).rgb; + const vec3 sample4b = tex2D_linearize(texture, tex_uv + dxdy_mirror_y * sample4d_texel_offset).rgb; + const vec3 sample1c = tex2D_linearize(texture, tex_uv + dxdy_mirror_x * sample1d_texel_offset).rgb; + const vec3 sample2c = tex2D_linearize(texture, tex_uv + dxdy_mirror_x * sample2d_texel_offset).rgb; + const vec3 sample3c = tex2D_linearize(texture, tex_uv + dxdy_mirror_x * sample3d_texel_offset).rgb; + const vec3 sample4c = tex2D_linearize(texture, tex_uv + dxdy_mirror_x * sample4d_texel_offset).rgb; + const vec3 sample1d = tex2D_linearize(texture, tex_uv + dxdy * sample1d_texel_offset).rgb; + const vec3 sample2d = tex2D_linearize(texture, tex_uv + dxdy * sample2d_texel_offset).rgb; + const vec3 sample3d = tex2D_linearize(texture, tex_uv + dxdy * sample3d_texel_offset).rgb; + const vec3 sample4d = tex2D_linearize(texture, tex_uv + dxdy * sample4d_texel_offset).rgb; + + // SUM WEIGHTED SAMPLES: + // Statically normalize weights (so total = 1.0), and sum weighted samples. + vec3 sum = vec3(0.0); + sum += w1 * (sample1a + sample1b + sample1c + sample1d); + sum += w2_3 * (sample2a + sample2b + sample2c + sample2d); + sum += w2_3 * (sample3a + sample3b + sample3c + sample3d); + sum += w4 * (sample4a + sample4b + sample4c + sample4d); + return sum * weight_sum_inv; +} + +vec3 tex2Dblur5x5(const sampler2D texture, const vec2 tex_uv, + const vec2 dxdy, const float sigma) +{ + // Perform a 1-pass 5x5 blur with 3x3 bilinear samples. + // Requires: Same as tex2Dblur9() + // Returns: A 5x5 Gaussian blurred mipmapped texture lookup composed of + // 3x3 carefully selected bilinear samples. + // Description: + // First see the description for tex2Dblur9x9(). This blur uses the same + // concept and sample/texel locations except on a smaller scale. Samples: + // 2a 1U 2b + // 1L 0C 1R + // 2c 1D 2d + // Texels: + // 2a4 2a3 1U2 2b3 2b4 + // 2a2 2a1 1U1 2b1 2b2 + // 1L2 1L1 0C1 1R1 1R2 + // 2c2 2c1 1D1 2d1 2d2 + // 2c4 2c3 1D2 2d3 2d4 + + // COMPUTE TEXTURE COORDS: + // Statically compute bilinear sampling offsets (details in tex2Dblur9x9). + const float denom_inv = 0.5/(sigma*sigma); + const float w1off = exp(-1.0 * denom_inv); + const float w2off = exp(-4.0 * denom_inv); + const float texel1to2ratio = w2off/(w1off + w2off); + // Statically compute texel offsets from the fragment center to each + // bilinear sample in the bottom-right quadrant, including x-axis-aligned: + const vec2 sample1R_texel_offset = vec2(1.0, 0.0) + vec2(texel1to2ratio, 0.0); + const vec2 sample2d_texel_offset = vec2(1.0, 1.0) + vec2(texel1to2ratio, texel1to2ratio); + + // CALCULATE KERNEL WEIGHTS FOR ALL SAMPLES: + // Statically compute Gaussian texel weights for the bottom-right quadrant. + // Read underscores as "and." + const float w1R1 = w1off; + const float w1R2 = w2off; + const float w2d1 = exp(-LENGTH_SQ(vec2(1.0, 1.0)) * denom_inv); + const float w2d2_3 = exp(-LENGTH_SQ(vec2(2.0, 1.0)) * denom_inv); + const float w2d4 = exp(-LENGTH_SQ(vec2(2.0, 2.0)) * denom_inv); + // Statically add texel weights in each sample to get sample weights: + const float w0 = 1.0; + const float w1 = w1R1 + w1R2; + const float w2 = w2d1 + 2.0 * w2d2_3 + w2d4; + // Get the weight sum inverse (normalization factor): + const float weight_sum_inv = 1.0/(w0 + 4.0 * (w1 + w2)); + + // LOAD TEXTURE SAMPLES: + // Load all 9 samples (1 nearest, 4 linear, 4 bilinear) using symmetry: + const vec2 mirror_x = vec2(-1.0, 1.0); + const vec2 mirror_y = vec2(1.0, -1.0); + const vec2 mirror_xy = vec2(-1.0, -1.0); + const vec2 dxdy_mirror_x = dxdy * mirror_x; + const vec2 dxdy_mirror_y = dxdy * mirror_y; + const vec2 dxdy_mirror_xy = dxdy * mirror_xy; + const vec3 sample0C = tex2D_linearize(texture, tex_uv).rgb; + const vec3 sample1R = tex2D_linearize(texture, tex_uv + dxdy * sample1R_texel_offset).rgb; + const vec3 sample1D = tex2D_linearize(texture, tex_uv + dxdy * sample1R_texel_offset.yx).rgb; + const vec3 sample1L = tex2D_linearize(texture, tex_uv - dxdy * sample1R_texel_offset).rgb; + const vec3 sample1U = tex2D_linearize(texture, tex_uv - dxdy * sample1R_texel_offset.yx).rgb; + const vec3 sample2d = tex2D_linearize(texture, tex_uv + dxdy * sample2d_texel_offset).rgb; + const vec3 sample2c = tex2D_linearize(texture, tex_uv + dxdy_mirror_x * sample2d_texel_offset).rgb; + const vec3 sample2b = tex2D_linearize(texture, tex_uv + dxdy_mirror_y * sample2d_texel_offset).rgb; + const vec3 sample2a = tex2D_linearize(texture, tex_uv + dxdy_mirror_xy * sample2d_texel_offset).rgb; + + // SUM WEIGHTED SAMPLES: + // Statically normalize weights (so total = 1.0), and sum weighted samples. + vec3 sum = w0 * sample0C; + sum += w1 * (sample1R + sample1D + sample1L + sample1U); + sum += w2 * (sample2a + sample2b + sample2c + sample2d); + return sum * weight_sum_inv; +} + +vec3 tex2Dblur3x3(const sampler2D texture, const vec2 tex_uv, + const vec2 dxdy, const float sigma) +{ + // Perform a 1-pass 3x3 blur with 5x5 bilinear samples. + // Requires: Same as tex2Dblur9() + // Returns: A 3x3 Gaussian blurred mipmapped texture lookup composed of + // 2x2 carefully selected bilinear samples. + // Description: + // First see the descriptions for tex2Dblur9x9() and tex2Dblur7(). This + // blur mixes concepts from both. The sample layout is as follows: + // 0a 0b + // 0c 0d + // The texel layout is as follows. Note that samples 0a/0b and 0c/0d share + // a vertical column of texels, and samples 0a/0c and 0b/0d share a + // horizontal row of texels (all samples share the center texel): + // 0a3 0ab2 0b3 + // 0ac1 0*0 0bd1 + // 0c3 0cd2 0d3 + + // COMPUTE TEXTURE COORDS: + // Statically compute bilinear sampling offsets (details in tex2Dblur9x9). + const float denom_inv = 0.5/(sigma*sigma); + const float w0off = 1.0; + const float w1off = exp(-1.0 * denom_inv); + const float texel0to1ratio = w1off/(w0off * 0.5 + w1off); + // Statically compute texel offsets from the fragment center to each + // bilinear sample in the bottom-right quadrant, including axis-aligned: + const vec2 sample0d_texel_offset = vec2(texel0to1ratio, texel0to1ratio); + + // LOAD TEXTURE SAMPLES: + // Load all 4 samples using symmetry: + const vec2 mirror_x = vec2(-1.0, 1.0); + const vec2 mirror_y = vec2(1.0, -1.0); + const vec2 mirror_xy = vec2(-1.0, -1.0); + const vec2 dxdy_mirror_x = dxdy * mirror_x; + const vec2 dxdy_mirror_y = dxdy * mirror_y; + const vec2 dxdy_mirror_xy = dxdy * mirror_xy; + const vec3 sample0a = tex2D_linearize(texture, tex_uv + dxdy_mirror_xy * sample0d_texel_offset).rgb; + const vec3 sample0b = tex2D_linearize(texture, tex_uv + dxdy_mirror_y * sample0d_texel_offset).rgb; + const vec3 sample0c = tex2D_linearize(texture, tex_uv + dxdy_mirror_x * sample0d_texel_offset).rgb; + const vec3 sample0d = tex2D_linearize(texture, tex_uv + dxdy * sample0d_texel_offset).rgb; + + // SUM WEIGHTED SAMPLES: + // Weights for all samples are the same, so just average them: + return 0.25 * (sample0a + sample0b + sample0c + sample0d); +} + + +////////////////// LINEAR ONE-PASS BLURS WITH SHARED SAMPLES ///////////////// + +vec3 tex2Dblur12x12shared(const sampler2D texture, + const vec4 tex_uv, const vec2 dxdy, const vec4 quad_vector, + const float sigma) +{ + // Perform a 1-pass mipmapped blur with shared samples across a pixel quad. + // Requires: 1.) Same as tex2Dblur9() + // 2.) ddx() and ddy() are present in the current Cg profile. + // 3.) The GPU driver is using fine/high-quality derivatives. + // 4.) quad_vector *correctly* describes the current fragment's + // location in its pixel quad, by the conventions noted in + // get_quad_vector[_naive]. + // 5.) tex_uv.w = log2(IN.video_size/IN.output_size).y + // 6.) tex2Dlod() is present in the current Cg profile. + // Optional: Tune artifacts vs. excessive blurriness with the global + // float error_blurring. + // Returns: A blurred texture lookup using a "virtual" 12x12 Gaussian + // blur (a 6x6 blur of carefully selected bilinear samples) + // of the given mip level. There will be subtle inaccuracies, + // especially for small or high-frequency detailed sources. + // Description: + // Perform a 1-pass blur with shared texture lookups across a pixel quad. + // We'll get neighboring samples with high-quality ddx/ddy derivatives, as + // in GPU Pro 2, Chapter VI.2, "Shader Amortization using Pixel Quad + // Message Passing" by Eric Penner. + // + // Our "virtual" 12x12 blur will be comprised of ((6 - 1)^2)/4 + 3 = 12 + // bilinear samples, where bilinear sampling positions are computed from + // the relative Gaussian weights of the 4 surrounding texels. The catch is + // that the appropriate texel weights and sample coords differ for each + // fragment, but we're reusing most of the same samples across a quad of + // destination fragments. (We do use unique coords for the four nearest + // samples at each fragment.) Mixing bilinear filtering and sample-sharing + // therefore introduces some error into the weights, and this can get nasty + // when the source image is small or high-frequency. Computing bilinear + // ratios based on weights at the sample field center results in sharpening + // and ringing artifacts, but we can move samples closer to halfway between + // texels to try blurring away the error (which can move features around by + // a texel or so). Tune this with the global float "error_blurring". + // + // The pixel quad's sample field covers 12x12 texels, accessed through 6x6 + // bilinear (2x2 texel) taps. Each fragment depends on a window of 10x10 + // texels (5x5 bilinear taps), and each fragment is responsible for loading + // a 6x6 texel quadrant as a 3x3 block of bilinear taps, plus 3 more taps + // to use unique bilinear coords for sample0* for each fragment. This + // diagram illustrates the relative locations of bilinear samples 1-9 for + // each quadrant a, b, c, d (note samples will not be equally spaced): + // 8a 7a 6a 6b 7b 8b + // 5a 4a 3a 3b 4b 5b + // 2a 1a 0a 0b 1b 2b + // 2c 1c 0c 0d 1d 2d + // 5c 4c 3c 3d 4d 5d + // 8c 7c 6c 6d 7d 8d + // The following diagram illustrates the underlying equally spaced texels, + // named after the sample that accesses them and subnamed by their location + // within their 2x2 texel block: + // 8a3 8a2 7a3 7a2 6a3 6a2 6b2 6b3 7b2 7b3 8b2 8b3 + // 8a1 8a0 7a1 7a0 6a1 6a0 6b0 6b1 7b0 7b1 8b0 8b1 + // 5a3 5a2 4a3 4a2 3a3 3a2 3b2 3b3 4b2 4b3 5b2 5b3 + // 5a1 5a0 4a1 4a0 3a1 3a0 3b0 3b1 4b0 4b1 5b0 5b1 + // 2a3 2a2 1a3 1a2 0a3 0a2 0b2 0b3 1b2 1b3 2b2 2b3 + // 2a1 2a0 1a1 1a0 0a1 0a0 0b0 0b1 1b0 1b1 2b0 2b1 + // 2c1 2c0 1c1 1c0 0c1 0c0 0d0 0d1 1d0 1d1 2d0 2d1 + // 2c3 2c2 1c3 1c2 0c3 0c2 0d2 0d3 1d2 1d3 2d2 2d3 + // 5c1 5c0 4c1 4c0 3c1 3c0 3d0 3d1 4d0 4d1 5d0 5d1 + // 5c3 5c2 4c3 4c2 3c3 3c2 3d2 3d3 4d2 4d3 5d2 5d3 + // 8c1 8c0 7c1 7c0 6c1 6c0 6d0 6d1 7d0 7d1 8d0 8d1 + // 8c3 8c2 7c3 7c2 6c3 6c2 6d2 6d3 7d2 7d3 8d2 8d3 + // With this symmetric arrangement, we don't have to know which absolute + // quadrant a sample lies in to assign kernel weights; it's enough to know + // the sample number and the relative quadrant of the sample (relative to + // the current quadrant): + // {current, adjacent x, adjacent y, diagonal} + + // COMPUTE COORDS FOR TEXTURE SAMPLES THIS FRAGMENT IS RESPONSIBLE FOR: + // Statically compute sampling offsets within each 2x2 texel block, based + // on appropriate 1D Gaussian sampling ratio between texels [0, 1], [2, 3], + // and [4, 5] away from the fragment, and reuse them independently for both + // dimensions. Use the sample field center as the estimated destination, + // but nudge the result closer to halfway between texels to blur error. + const float denom_inv = 0.5/(sigma*sigma); + const float w0off = 1.0; + const float w0_5off = exp(-(0.5*0.5) * denom_inv); + const float w1off = exp(-(1.0*1.0) * denom_inv); + const float w1_5off = exp(-(1.5*1.5) * denom_inv); + const float w2off = exp(-(2.0*2.0) * denom_inv); + const float w2_5off = exp(-(2.5*2.5) * denom_inv); + const float w3_5off = exp(-(3.5*3.5) * denom_inv); + const float w4_5off = exp(-(4.5*4.5) * denom_inv); + const float w5_5off = exp(-(5.5*5.5) * denom_inv); + const float texel0to1ratio = mix(w1_5off/(w0_5off + w1_5off), 0.5, error_blurring); + const float texel2to3ratio = mix(w3_5off/(w2_5off + w3_5off), 0.5, error_blurring); + const float texel4to5ratio = mix(w5_5off/(w4_5off + w5_5off), 0.5, error_blurring); + // We don't share sample0*, so use the nearest destination fragment: + const float texel0to1ratio_nearest = w1off/(w0off + w1off); + const float texel1to2ratio_nearest = w2off/(w1off + w2off); + // Statically compute texel offsets from the bottom-right fragment to each + // bilinear sample in the bottom-right quadrant: + const vec2 sample0curr_texel_offset = vec2(0.0, 0.0) + vec2(texel0to1ratio_nearest, texel0to1ratio_nearest); + const vec2 sample0adjx_texel_offset = vec2(-1.0, 0.0) + vec2(-texel1to2ratio_nearest, texel0to1ratio_nearest); + const vec2 sample0adjy_texel_offset = vec2(0.0, -1.0) + vec2(texel0to1ratio_nearest, -texel1to2ratio_nearest); + const vec2 sample0diag_texel_offset = vec2(-1.0, -1.0) + vec2(-texel1to2ratio_nearest, -texel1to2ratio_nearest); + const vec2 sample1_texel_offset = vec2(2.0, 0.0) + vec2(texel2to3ratio, texel0to1ratio); + const vec2 sample2_texel_offset = vec2(4.0, 0.0) + vec2(texel4to5ratio, texel0to1ratio); + const vec2 sample3_texel_offset = vec2(0.0, 2.0) + vec2(texel0to1ratio, texel2to3ratio); + const vec2 sample4_texel_offset = vec2(2.0, 2.0) + vec2(texel2to3ratio, texel2to3ratio); + const vec2 sample5_texel_offset = vec2(4.0, 2.0) + vec2(texel4to5ratio, texel2to3ratio); + const vec2 sample6_texel_offset = vec2(0.0, 4.0) + vec2(texel0to1ratio, texel4to5ratio); + const vec2 sample7_texel_offset = vec2(2.0, 4.0) + vec2(texel2to3ratio, texel4to5ratio); + const vec2 sample8_texel_offset = vec2(4.0, 4.0) + vec2(texel4to5ratio, texel4to5ratio); + + // CALCULATE KERNEL WEIGHTS: + // Statically compute bilinear sample weights at each destination fragment + // based on the sum of their 4 underlying texel weights. Assume a same- + // resolution blur, so each symmetrically named sample weight will compute + // the same at every fragment in the pixel quad: We can therefore compute + // texel weights based only on the bottom-right quadrant (fragment at 0d0). + // Too avoid too much boilerplate code, use a macro to get all 4 texel + // weights for a bilinear sample based on the offset of its top-left texel: + #define GET_TEXEL_QUAD_WEIGHTS(xoff, yoff) \ + (exp(-LENGTH_SQ(vec2(xoff, yoff)) * denom_inv) + \ + exp(-LENGTH_SQ(vec2(xoff + 1.0, yoff)) * denom_inv) + \ + exp(-LENGTH_SQ(vec2(xoff, yoff + 1.0)) * denom_inv) + \ + exp(-LENGTH_SQ(vec2(xoff + 1.0, yoff + 1.0)) * denom_inv)) + const float w8diag = GET_TEXEL_QUAD_WEIGHTS(-6.0, -6.0); + const float w7diag = GET_TEXEL_QUAD_WEIGHTS(-4.0, -6.0); + const float w6diag = GET_TEXEL_QUAD_WEIGHTS(-2.0, -6.0); + const float w6adjy = GET_TEXEL_QUAD_WEIGHTS(0.0, -6.0); + const float w7adjy = GET_TEXEL_QUAD_WEIGHTS(2.0, -6.0); + const float w8adjy = GET_TEXEL_QUAD_WEIGHTS(4.0, -6.0); + const float w5diag = GET_TEXEL_QUAD_WEIGHTS(-6.0, -4.0); + const float w4diag = GET_TEXEL_QUAD_WEIGHTS(-4.0, -4.0); + const float w3diag = GET_TEXEL_QUAD_WEIGHTS(-2.0, -4.0); + const float w3adjy = GET_TEXEL_QUAD_WEIGHTS(0.0, -4.0); + const float w4adjy = GET_TEXEL_QUAD_WEIGHTS(2.0, -4.0); + const float w5adjy = GET_TEXEL_QUAD_WEIGHTS(4.0, -4.0); + const float w2diag = GET_TEXEL_QUAD_WEIGHTS(-6.0, -2.0); + const float w1diag = GET_TEXEL_QUAD_WEIGHTS(-4.0, -2.0); + const float w0diag = GET_TEXEL_QUAD_WEIGHTS(-2.0, -2.0); + const float w0adjy = GET_TEXEL_QUAD_WEIGHTS(0.0, -2.0); + const float w1adjy = GET_TEXEL_QUAD_WEIGHTS(2.0, -2.0); + const float w2adjy = GET_TEXEL_QUAD_WEIGHTS(4.0, -2.0); + const float w2adjx = GET_TEXEL_QUAD_WEIGHTS(-6.0, 0.0); + const float w1adjx = GET_TEXEL_QUAD_WEIGHTS(-4.0, 0.0); + const float w0adjx = GET_TEXEL_QUAD_WEIGHTS(-2.0, 0.0); + const float w0curr = GET_TEXEL_QUAD_WEIGHTS(0.0, 0.0); + const float w1curr = GET_TEXEL_QUAD_WEIGHTS(2.0, 0.0); + const float w2curr = GET_TEXEL_QUAD_WEIGHTS(4.0, 0.0); + const float w5adjx = GET_TEXEL_QUAD_WEIGHTS(-6.0, 2.0); + const float w4adjx = GET_TEXEL_QUAD_WEIGHTS(-4.0, 2.0); + const float w3adjx = GET_TEXEL_QUAD_WEIGHTS(-2.0, 2.0); + const float w3curr = GET_TEXEL_QUAD_WEIGHTS(0.0, 2.0); + const float w4curr = GET_TEXEL_QUAD_WEIGHTS(2.0, 2.0); + const float w5curr = GET_TEXEL_QUAD_WEIGHTS(4.0, 2.0); + const float w8adjx = GET_TEXEL_QUAD_WEIGHTS(-6.0, 4.0); + const float w7adjx = GET_TEXEL_QUAD_WEIGHTS(-4.0, 4.0); + const float w6adjx = GET_TEXEL_QUAD_WEIGHTS(-2.0, 4.0); + const float w6curr = GET_TEXEL_QUAD_WEIGHTS(0.0, 4.0); + const float w7curr = GET_TEXEL_QUAD_WEIGHTS(2.0, 4.0); + const float w8curr = GET_TEXEL_QUAD_WEIGHTS(4.0, 4.0); + #undef GET_TEXEL_QUAD_WEIGHTS + // Statically pack weights for runtime: + const vec4 w0 = vec4(w0curr, w0adjx, w0adjy, w0diag); + const vec4 w1 = vec4(w1curr, w1adjx, w1adjy, w1diag); + const vec4 w2 = vec4(w2curr, w2adjx, w2adjy, w2diag); + const vec4 w3 = vec4(w3curr, w3adjx, w3adjy, w3diag); + const vec4 w4 = vec4(w4curr, w4adjx, w4adjy, w4diag); + const vec4 w5 = vec4(w5curr, w5adjx, w5adjy, w5diag); + const vec4 w6 = vec4(w6curr, w6adjx, w6adjy, w6diag); + const vec4 w7 = vec4(w7curr, w7adjx, w7adjy, w7diag); + const vec4 w8 = vec4(w8curr, w8adjx, w8adjy, w8diag); + // Get the weight sum inverse (normalization factor): + const vec4 weight_sum4 = w0 + w1 + w2 + w3 + w4 + w5 + w6 + w7 + w8; + const vec2 weight_sum2 = weight_sum4.xy + weight_sum4.zw; + const float weight_sum = weight_sum2.x + weight_sum2.y; + const float weight_sum_inv = 1.0/(weight_sum); + + // LOAD TEXTURE SAMPLES THIS FRAGMENT IS RESPONSIBLE FOR: + // Get a uv vector from texel 0q0 of this quadrant to texel 0q3: + const vec2 dxdy_curr = dxdy * quad_vector.xy; + // Load bilinear samples for the current quadrant (for this fragment): + const vec3 sample0curr = tex2D_linearize(texture, tex_uv.xy + dxdy_curr * sample0curr_texel_offset).rgb; + const vec3 sample0adjx = tex2D_linearize(texture, tex_uv.xy + dxdy_curr * sample0adjx_texel_offset).rgb; + const vec3 sample0adjy = tex2D_linearize(texture, tex_uv.xy + dxdy_curr * sample0adjy_texel_offset).rgb; + const vec3 sample0diag = tex2D_linearize(texture, tex_uv.xy + dxdy_curr * sample0diag_texel_offset).rgb; + const vec3 sample1curr = tex2Dlod_linearize(texture, tex_uv + uv2_to_uv4(dxdy_curr * sample1_texel_offset)).rgb; + const vec3 sample2curr = tex2Dlod_linearize(texture, tex_uv + uv2_to_uv4(dxdy_curr * sample2_texel_offset)).rgb; + const vec3 sample3curr = tex2Dlod_linearize(texture, tex_uv + uv2_to_uv4(dxdy_curr * sample3_texel_offset)).rgb; + const vec3 sample4curr = tex2Dlod_linearize(texture, tex_uv + uv2_to_uv4(dxdy_curr * sample4_texel_offset)).rgb; + const vec3 sample5curr = tex2Dlod_linearize(texture, tex_uv + uv2_to_uv4(dxdy_curr * sample5_texel_offset)).rgb; + const vec3 sample6curr = tex2Dlod_linearize(texture, tex_uv + uv2_to_uv4(dxdy_curr * sample6_texel_offset)).rgb; + const vec3 sample7curr = tex2Dlod_linearize(texture, tex_uv + uv2_to_uv4(dxdy_curr * sample7_texel_offset)).rgb; + const vec3 sample8curr = tex2Dlod_linearize(texture, tex_uv + uv2_to_uv4(dxdy_curr * sample8_texel_offset)).rgb; + + // GATHER NEIGHBORING SAMPLES AND SUM WEIGHTED SAMPLES: + // Fetch the samples from other fragments in the 2x2 quad: + vec3 sample1adjx, sample1adjy, sample1diag; + vec3 sample2adjx, sample2adjy, sample2diag; + vec3 sample3adjx, sample3adjy, sample3diag; + vec3 sample4adjx, sample4adjy, sample4diag; + vec3 sample5adjx, sample5adjy, sample5diag; + vec3 sample6adjx, sample6adjy, sample6diag; + vec3 sample7adjx, sample7adjy, sample7diag; + vec3 sample8adjx, sample8adjy, sample8diag; + quad_gather(quad_vector, sample1curr, sample1adjx, sample1adjy, sample1diag); + quad_gather(quad_vector, sample2curr, sample2adjx, sample2adjy, sample2diag); + quad_gather(quad_vector, sample3curr, sample3adjx, sample3adjy, sample3diag); + quad_gather(quad_vector, sample4curr, sample4adjx, sample4adjy, sample4diag); + quad_gather(quad_vector, sample5curr, sample5adjx, sample5adjy, sample5diag); + quad_gather(quad_vector, sample6curr, sample6adjx, sample6adjy, sample6diag); + quad_gather(quad_vector, sample7curr, sample7adjx, sample7adjy, sample7diag); + quad_gather(quad_vector, sample8curr, sample8adjx, sample8adjy, sample8diag); + // Statically normalize weights (so total = 1.0), and sum weighted samples. + // Fill each row of a matrix with an rgb sample and pre-multiply by the + // weights to obtain a weighted result: + vec3 sum = vec3(0.0); + sum += (mat4x3(sample0curr, sample0adjx, sample0adjy, sample0diag) * w0); + sum += (mat4x3(sample1curr, sample1adjx, sample1adjy, sample1diag) * w1); + sum += (mat4x3(sample2curr, sample2adjx, sample2adjy, sample2diag) * w2); + sum += (mat4x3(sample3curr, sample3adjx, sample3adjy, sample3diag) * w3); + sum += (mat4x3(sample4curr, sample4adjx, sample4adjy, sample4diag) * w4); + sum += (mat4x3(sample5curr, sample5adjx, sample5adjy, sample5diag) * w5); + sum += (mat4x3(sample6curr, sample6adjx, sample6adjy, sample6diag) * w6); + sum += (mat4x3(sample7curr, sample7adjx, sample7adjy, sample7diag) * w7); + sum += (mat4x3(sample8curr, sample8adjx, sample8adjy, sample8diag) * w8); + return sum * weight_sum_inv; +} + +vec3 tex2Dblur10x10shared(const sampler2D texture, + const vec4 tex_uv, const vec2 dxdy, const vec4 quad_vector, + const float sigma) +{ + // Perform a 1-pass mipmapped blur with shared samples across a pixel quad. + // Requires: Same as tex2Dblur12x12shared() + // Returns: A blurred texture lookup using a "virtual" 10x10 Gaussian + // blur (a 5x5 blur of carefully selected bilinear samples) + // of the given mip level. There will be subtle inaccuracies, + // especially for small or high-frequency detailed sources. + // Description: + // First see the description for tex2Dblur12x12shared(). This + // function shares the same concept and sample placement, but each fragment + // only uses 25 of the 36 samples taken across the pixel quad (to cover a + // 5x5 sample area, or 10x10 texel area), and it uses a lower standard + // deviation to compensate. Thanks to symmetry, the 11 omitted samples + // are always the "same:" + // 8adjx, 2adjx, 5adjx, + // 6adjy, 7adjy, 8adjy, + // 2diag, 5diag, 6diag, 7diag, 8diag + + // COMPUTE COORDS FOR TEXTURE SAMPLES THIS FRAGMENT IS RESPONSIBLE FOR: + // Statically compute bilinear sampling offsets (details in tex2Dblur12x12shared). + const float denom_inv = 0.5/(sigma*sigma); + const float w0off = 1.0; + const float w0_5off = exp(-(0.5*0.5) * denom_inv); + const float w1off = exp(-(1.0*1.0) * denom_inv); + const float w1_5off = exp(-(1.5*1.5) * denom_inv); + const float w2off = exp(-(2.0*2.0) * denom_inv); + const float w2_5off = exp(-(2.5*2.5) * denom_inv); + const float w3_5off = exp(-(3.5*3.5) * denom_inv); + const float w4_5off = exp(-(4.5*4.5) * denom_inv); + const float w5_5off = exp(-(5.5*5.5) * denom_inv); + const float texel0to1ratio = mix(w1_5off/(w0_5off + w1_5off), 0.5, error_blurring); + const float texel2to3ratio = mix(w3_5off/(w2_5off + w3_5off), 0.5, error_blurring); + const float texel4to5ratio = mix(w5_5off/(w4_5off + w5_5off), 0.5, error_blurring); + // We don't share sample0*, so use the nearest destination fragment: + const float texel0to1ratio_nearest = w1off/(w0off + w1off); + const float texel1to2ratio_nearest = w2off/(w1off + w2off); + // Statically compute texel offsets from the bottom-right fragment to each + // bilinear sample in the bottom-right quadrant: + const vec2 sample0curr_texel_offset = vec2(0.0, 0.0) + vec2(texel0to1ratio_nearest, texel0to1ratio_nearest); + const vec2 sample0adjx_texel_offset = vec2(-1.0, 0.0) + vec2(-texel1to2ratio_nearest, texel0to1ratio_nearest); + const vec2 sample0adjy_texel_offset = vec2(0.0, -1.0) + vec2(texel0to1ratio_nearest, -texel1to2ratio_nearest); + const vec2 sample0diag_texel_offset = vec2(-1.0, -1.0) + vec2(-texel1to2ratio_nearest, -texel1to2ratio_nearest); + const vec2 sample1_texel_offset = vec2(2.0, 0.0) + vec2(texel2to3ratio, texel0to1ratio); + const vec2 sample2_texel_offset = vec2(4.0, 0.0) + vec2(texel4to5ratio, texel0to1ratio); + const vec2 sample3_texel_offset = vec2(0.0, 2.0) + vec2(texel0to1ratio, texel2to3ratio); + const vec2 sample4_texel_offset = vec2(2.0, 2.0) + vec2(texel2to3ratio, texel2to3ratio); + const vec2 sample5_texel_offset = vec2(4.0, 2.0) + vec2(texel4to5ratio, texel2to3ratio); + const vec2 sample6_texel_offset = vec2(0.0, 4.0) + vec2(texel0to1ratio, texel4to5ratio); + const vec2 sample7_texel_offset = vec2(2.0, 4.0) + vec2(texel2to3ratio, texel4to5ratio); + const vec2 sample8_texel_offset = vec2(4.0, 4.0) + vec2(texel4to5ratio, texel4to5ratio); + + // CALCULATE KERNEL WEIGHTS: + // Statically compute bilinear sample weights at each destination fragment + // from the sum of their 4 texel weights (details in tex2Dblur12x12shared). + #define GET_TEXEL_QUAD_WEIGHTS(xoff, yoff) \ + (exp(-LENGTH_SQ(vec2(xoff, yoff)) * denom_inv) + \ + exp(-LENGTH_SQ(vec2(xoff + 1.0, yoff)) * denom_inv) + \ + exp(-LENGTH_SQ(vec2(xoff, yoff + 1.0)) * denom_inv) + \ + exp(-LENGTH_SQ(vec2(xoff + 1.0, yoff + 1.0)) * denom_inv)) + // We only need 25 of the 36 sample weights. Skip the following weights: + // 8adjx, 2adjx, 5adjx, + // 6adjy, 7adjy, 8adjy, + // 2diag, 5diag, 6diag, 7diag, 8diag + const float w4diag = GET_TEXEL_QUAD_WEIGHTS(-4.0, -4.0); + const float w3diag = GET_TEXEL_QUAD_WEIGHTS(-2.0, -4.0); + const float w3adjy = GET_TEXEL_QUAD_WEIGHTS(0.0, -4.0); + const float w4adjy = GET_TEXEL_QUAD_WEIGHTS(2.0, -4.0); + const float w5adjy = GET_TEXEL_QUAD_WEIGHTS(4.0, -4.0); + const float w1diag = GET_TEXEL_QUAD_WEIGHTS(-4.0, -2.0); + const float w0diag = GET_TEXEL_QUAD_WEIGHTS(-2.0, -2.0); + const float w0adjy = GET_TEXEL_QUAD_WEIGHTS(0.0, -2.0); + const float w1adjy = GET_TEXEL_QUAD_WEIGHTS(2.0, -2.0); + const float w2adjy = GET_TEXEL_QUAD_WEIGHTS(4.0, -2.0); + const float w1adjx = GET_TEXEL_QUAD_WEIGHTS(-4.0, 0.0); + const float w0adjx = GET_TEXEL_QUAD_WEIGHTS(-2.0, 0.0); + const float w0curr = GET_TEXEL_QUAD_WEIGHTS(0.0, 0.0); + const float w1curr = GET_TEXEL_QUAD_WEIGHTS(2.0, 0.0); + const float w2curr = GET_TEXEL_QUAD_WEIGHTS(4.0, 0.0); + const float w4adjx = GET_TEXEL_QUAD_WEIGHTS(-4.0, 2.0); + const float w3adjx = GET_TEXEL_QUAD_WEIGHTS(-2.0, 2.0); + const float w3curr = GET_TEXEL_QUAD_WEIGHTS(0.0, 2.0); + const float w4curr = GET_TEXEL_QUAD_WEIGHTS(2.0, 2.0); + const float w5curr = GET_TEXEL_QUAD_WEIGHTS(4.0, 2.0); + const float w7adjx = GET_TEXEL_QUAD_WEIGHTS(-4.0, 4.0); + const float w6adjx = GET_TEXEL_QUAD_WEIGHTS(-2.0, 4.0); + const float w6curr = GET_TEXEL_QUAD_WEIGHTS(0.0, 4.0); + const float w7curr = GET_TEXEL_QUAD_WEIGHTS(2.0, 4.0); + const float w8curr = GET_TEXEL_QUAD_WEIGHTS(4.0, 4.0); + #undef GET_TEXEL_QUAD_WEIGHTS + // Get the weight sum inverse (normalization factor): + const float weight_sum_inv = 1.0/(w0curr + w1curr + w2curr + w3curr + + w4curr + w5curr + w6curr + w7curr + w8curr + + w0adjx + w1adjx + w3adjx + w4adjx + w6adjx + w7adjx + + w0adjy + w1adjy + w2adjy + w3adjy + w4adjy + w5adjy + + w0diag + w1diag + w3diag + w4diag); + // Statically pack most weights for runtime. Note the mixed packing: + const vec4 w0 = vec4(w0curr, w0adjx, w0adjy, w0diag); + const vec4 w1 = vec4(w1curr, w1adjx, w1adjy, w1diag); + const vec4 w3 = vec4(w3curr, w3adjx, w3adjy, w3diag); + const vec4 w4 = vec4(w4curr, w4adjx, w4adjy, w4diag); + const vec4 w2and5 = vec4(w2curr, w2adjy, w5curr, w5adjy); + const vec4 w6and7 = vec4(w6curr, w6adjx, w7curr, w7adjx); + + // LOAD TEXTURE SAMPLES THIS FRAGMENT IS RESPONSIBLE FOR: + // Get a uv vector from texel 0q0 of this quadrant to texel 0q3: + const vec2 dxdy_curr = dxdy * quad_vector.xy; + // Load bilinear samples for the current quadrant (for this fragment): + const vec3 sample0curr = tex2D_linearize(texture, tex_uv.xy + dxdy_curr * sample0curr_texel_offset).rgb; + const vec3 sample0adjx = tex2D_linearize(texture, tex_uv.xy + dxdy_curr * sample0adjx_texel_offset).rgb; + const vec3 sample0adjy = tex2D_linearize(texture, tex_uv.xy + dxdy_curr * sample0adjy_texel_offset).rgb; + const vec3 sample0diag = tex2D_linearize(texture, tex_uv.xy + dxdy_curr * sample0diag_texel_offset).rgb; + const vec3 sample1curr = tex2Dlod_linearize(texture, tex_uv + uv2_to_uv4(dxdy_curr * sample1_texel_offset)).rgb; + const vec3 sample2curr = tex2Dlod_linearize(texture, tex_uv + uv2_to_uv4(dxdy_curr * sample2_texel_offset)).rgb; + const vec3 sample3curr = tex2Dlod_linearize(texture, tex_uv + uv2_to_uv4(dxdy_curr * sample3_texel_offset)).rgb; + const vec3 sample4curr = tex2Dlod_linearize(texture, tex_uv + uv2_to_uv4(dxdy_curr * sample4_texel_offset)).rgb; + const vec3 sample5curr = tex2Dlod_linearize(texture, tex_uv + uv2_to_uv4(dxdy_curr * sample5_texel_offset)).rgb; + const vec3 sample6curr = tex2Dlod_linearize(texture, tex_uv + uv2_to_uv4(dxdy_curr * sample6_texel_offset)).rgb; + const vec3 sample7curr = tex2Dlod_linearize(texture, tex_uv + uv2_to_uv4(dxdy_curr * sample7_texel_offset)).rgb; + const vec3 sample8curr = tex2Dlod_linearize(texture, tex_uv + uv2_to_uv4(dxdy_curr * sample8_texel_offset)).rgb; + + // GATHER NEIGHBORING SAMPLES AND SUM WEIGHTED SAMPLES: + // Fetch the samples from other fragments in the 2x2 quad in order of need: + vec3 sample1adjx, sample1adjy, sample1diag; + vec3 sample2adjx, sample2adjy, sample2diag; + vec3 sample3adjx, sample3adjy, sample3diag; + vec3 sample4adjx, sample4adjy, sample4diag; + vec3 sample5adjx, sample5adjy, sample5diag; + vec3 sample6adjx, sample6adjy, sample6diag; + vec3 sample7adjx, sample7adjy, sample7diag; + quad_gather(quad_vector, sample1curr, sample1adjx, sample1adjy, sample1diag); + quad_gather(quad_vector, sample2curr, sample2adjx, sample2adjy, sample2diag); + quad_gather(quad_vector, sample3curr, sample3adjx, sample3adjy, sample3diag); + quad_gather(quad_vector, sample4curr, sample4adjx, sample4adjy, sample4diag); + quad_gather(quad_vector, sample5curr, sample5adjx, sample5adjy, sample5diag); + quad_gather(quad_vector, sample6curr, sample6adjx, sample6adjy, sample6diag); + quad_gather(quad_vector, sample7curr, sample7adjx, sample7adjy, sample7diag); + // Statically normalize weights (so total = 1.0), and sum weighted samples. + // Fill each row of a matrix with an rgb sample and pre-multiply by the + // weights to obtain a weighted result. First do the simple ones: + vec3 sum = vec3(0.0); + sum += (mat4x3(sample0curr, sample0adjx, sample0adjy, sample0diag) * w0); + sum += (mat4x3(sample1curr, sample1adjx, sample1adjy, sample1diag) * w1); + sum += (mat4x3(sample3curr, sample3adjx, sample3adjy, sample3diag) * w2); + sum += (mat4x3(sample4curr, sample4adjx, sample4adjy, sample4diag) * w3); + // Now do the mixed-sample ones: + sum += (mat4x3(sample2curr, sample2adjy, sample5curr, sample5adjy) * w2and5); + sum += (mat4x3(sample6curr, sample6adjx, sample7curr, sample7adjx) * w6and7); + sum += w8curr * sample8curr; + // Normalize the sum (so the weights add to 1.0) and return: + return sum * weight_sum_inv; +} + +vec3 tex2Dblur8x8shared(const sampler2D texture, + const vec4 tex_uv, const vec2 dxdy, const vec4 quad_vector, + const float sigma) +{ + // Perform a 1-pass mipmapped blur with shared samples across a pixel quad. + // Requires: Same as tex2Dblur12x12shared() + // Returns: A blurred texture lookup using a "virtual" 8x8 Gaussian + // blur (a 4x4 blur of carefully selected bilinear samples) + // of the given mip level. There will be subtle inaccuracies, + // especially for small or high-frequency detailed sources. + // Description: + // First see the description for tex2Dblur12x12shared(). This function + // shares the same concept and a similar sample placement, except each + // quadrant contains 4x4 texels and 2x2 samples instead of 6x6 and 3x3 + // respectively. There could be a total of 16 samples, 4 of which each + // fragment is responsible for, but each fragment loads 0a/0b/0c/0d with + // its own offset to reduce shared sample artifacts, bringing the sample + // count for each fragment to 7. Sample placement: + // 3a 2a 2b 3b + // 1a 0a 0b 1b + // 1c 0c 0d 1d + // 3c 2c 2d 3d + // Texel placement: + // 3a3 3a2 2a3 2a2 2b2 2b3 3b2 3b3 + // 3a1 3a0 2a1 2a0 2b0 2b1 3b0 3b1 + // 1a3 1a2 0a3 0a2 0b2 0b3 1b2 1b3 + // 1a1 1a0 0a1 0a0 0b0 0b1 1b0 1b1 + // 1c1 1c0 0c1 0c0 0d0 0d1 1d0 1d1 + // 1c3 1c2 0c3 0c2 0d2 0d3 1d2 1d3 + // 3c1 3c0 2c1 2c0 2d0 2d1 3d0 4d1 + // 3c3 3c2 2c3 2c2 2d2 2d3 3d2 4d3 + + // COMPUTE COORDS FOR TEXTURE SAMPLES THIS FRAGMENT IS RESPONSIBLE FOR: + // Statically compute bilinear sampling offsets (details in tex2Dblur12x12shared). + const float denom_inv = 0.5/(sigma*sigma); + const float w0off = 1.0; + const float w0_5off = exp(-(0.5*0.5) * denom_inv); + const float w1off = exp(-(1.0*1.0) * denom_inv); + const float w1_5off = exp(-(1.5*1.5) * denom_inv); + const float w2off = exp(-(2.0*2.0) * denom_inv); + const float w2_5off = exp(-(2.5*2.5) * denom_inv); + const float w3_5off = exp(-(3.5*3.5) * denom_inv); + const float texel0to1ratio = mix(w1_5off/(w0_5off + w1_5off), 0.5, error_blurring); + const float texel2to3ratio = mix(w3_5off/(w2_5off + w3_5off), 0.5, error_blurring); + // We don't share sample0*, so use the nearest destination fragment: + const float texel0to1ratio_nearest = w1off/(w0off + w1off); + const float texel1to2ratio_nearest = w2off/(w1off + w2off); + // Statically compute texel offsets from the bottom-right fragment to each + // bilinear sample in the bottom-right quadrant: + const vec2 sample0curr_texel_offset = vec2(0.0, 0.0) + vec2(texel0to1ratio_nearest, texel0to1ratio_nearest); + const vec2 sample0adjx_texel_offset = vec2(-1.0, 0.0) + vec2(-texel1to2ratio_nearest, texel0to1ratio_nearest); + const vec2 sample0adjy_texel_offset = vec2(0.0, -1.0) + vec2(texel0to1ratio_nearest, -texel1to2ratio_nearest); + const vec2 sample0diag_texel_offset = vec2(-1.0, -1.0) + vec2(-texel1to2ratio_nearest, -texel1to2ratio_nearest); + const vec2 sample1_texel_offset = vec2(2.0, 0.0) + vec2(texel2to3ratio, texel0to1ratio); + const vec2 sample2_texel_offset = vec2(0.0, 2.0) + vec2(texel0to1ratio, texel2to3ratio); + const vec2 sample3_texel_offset = vec2(2.0, 2.0) + vec2(texel2to3ratio, texel2to3ratio); + + // CALCULATE KERNEL WEIGHTS: + // Statically compute bilinear sample weights at each destination fragment + // from the sum of their 4 texel weights (details in tex2Dblur12x12shared). + #define GET_TEXEL_QUAD_WEIGHTS(xoff, yoff) \ + (exp(-LENGTH_SQ(vec2(xoff, yoff)) * denom_inv) + \ + exp(-LENGTH_SQ(vec2(xoff + 1.0, yoff)) * denom_inv) + \ + exp(-LENGTH_SQ(vec2(xoff, yoff + 1.0)) * denom_inv) + \ + exp(-LENGTH_SQ(vec2(xoff + 1.0, yoff + 1.0)) * denom_inv)) + const float w3diag = GET_TEXEL_QUAD_WEIGHTS(-4.0, -4.0); + const float w2diag = GET_TEXEL_QUAD_WEIGHTS(-2.0, -4.0); + const float w2adjy = GET_TEXEL_QUAD_WEIGHTS(0.0, -4.0); + const float w3adjy = GET_TEXEL_QUAD_WEIGHTS(2.0, -4.0); + const float w1diag = GET_TEXEL_QUAD_WEIGHTS(-4.0, -2.0); + const float w0diag = GET_TEXEL_QUAD_WEIGHTS(-2.0, -2.0); + const float w0adjy = GET_TEXEL_QUAD_WEIGHTS(0.0, -2.0); + const float w1adjy = GET_TEXEL_QUAD_WEIGHTS(2.0, -2.0); + const float w1adjx = GET_TEXEL_QUAD_WEIGHTS(-4.0, 0.0); + const float w0adjx = GET_TEXEL_QUAD_WEIGHTS(-2.0, 0.0); + const float w0curr = GET_TEXEL_QUAD_WEIGHTS(0.0, 0.0); + const float w1curr = GET_TEXEL_QUAD_WEIGHTS(2.0, 0.0); + const float w3adjx = GET_TEXEL_QUAD_WEIGHTS(-4.0, 2.0); + const float w2adjx = GET_TEXEL_QUAD_WEIGHTS(-2.0, 2.0); + const float w2curr = GET_TEXEL_QUAD_WEIGHTS(0.0, 2.0); + const float w3curr = GET_TEXEL_QUAD_WEIGHTS(2.0, 2.0); + #undef GET_TEXEL_QUAD_WEIGHTS + // Statically pack weights for runtime: + const vec4 w0 = vec4(w0curr, w0adjx, w0adjy, w0diag); + const vec4 w1 = vec4(w1curr, w1adjx, w1adjy, w1diag); + const vec4 w2 = vec4(w2curr, w2adjx, w2adjy, w2diag); + const vec4 w3 = vec4(w3curr, w3adjx, w3adjy, w3diag); + // Get the weight sum inverse (normalization factor): + const vec4 weight_sum4 = w0 + w1 + w2 + w3; + const vec2 weight_sum2 = weight_sum4.xy + weight_sum4.zw; + const float weight_sum = weight_sum2.x + weight_sum2.y; + const float weight_sum_inv = 1.0/(weight_sum); + + // LOAD TEXTURE SAMPLES THIS FRAGMENT IS RESPONSIBLE FOR: + // Get a uv vector from texel 0q0 of this quadrant to texel 0q3: + const vec2 dxdy_curr = dxdy * quad_vector.xy; + // Load bilinear samples for the current quadrant (for this fragment): + const vec3 sample0curr = tex2D_linearize(texture, tex_uv.xy + dxdy_curr * sample0curr_texel_offset).rgb; + const vec3 sample0adjx = tex2D_linearize(texture, tex_uv.xy + dxdy_curr * sample0adjx_texel_offset).rgb; + const vec3 sample0adjy = tex2D_linearize(texture, tex_uv.xy + dxdy_curr * sample0adjy_texel_offset).rgb; + const vec3 sample0diag = tex2D_linearize(texture, tex_uv.xy + dxdy_curr * sample0diag_texel_offset).rgb; + const vec3 sample1curr = tex2Dlod_linearize(texture, tex_uv + uv2_to_uv4(dxdy_curr * sample1_texel_offset)).rgb; + const vec3 sample2curr = tex2Dlod_linearize(texture, tex_uv + uv2_to_uv4(dxdy_curr * sample2_texel_offset)).rgb; + const vec3 sample3curr = tex2Dlod_linearize(texture, tex_uv + uv2_to_uv4(dxdy_curr * sample3_texel_offset)).rgb; + + // GATHER NEIGHBORING SAMPLES AND SUM WEIGHTED SAMPLES: + // Fetch the samples from other fragments in the 2x2 quad: + vec3 sample1adjx, sample1adjy, sample1diag; + vec3 sample2adjx, sample2adjy, sample2diag; + vec3 sample3adjx, sample3adjy, sample3diag; + quad_gather(quad_vector, sample1curr, sample1adjx, sample1adjy, sample1diag); + quad_gather(quad_vector, sample2curr, sample2adjx, sample2adjy, sample2diag); + quad_gather(quad_vector, sample3curr, sample3adjx, sample3adjy, sample3diag); + // Statically normalize weights (so total = 1.0), and sum weighted samples. + // Fill each row of a matrix with an rgb sample and pre-multiply by the + // weights to obtain a weighted result: + vec3 sum = vec3(0.0); + sum += (mat4x3(sample0curr, sample0adjx, sample0adjy, sample0diag) * w0); + sum += (mat4x3(sample1curr, sample1adjx, sample1adjy, sample1diag) * w1); + sum += (mat4x3(sample2curr, sample2adjx, sample2adjy, sample2diag) * w2); + sum += (mat4x3(sample3curr, sample3adjx, sample3adjy, sample3diag) * w3); + return sum * weight_sum_inv; +} + +vec3 tex2Dblur6x6shared(const sampler2D texture, + const vec4 tex_uv, const vec2 dxdy, const vec4 quad_vector, + const float sigma) +{ + // Perform a 1-pass mipmapped blur with shared samples across a pixel quad. + // Requires: Same as tex2Dblur12x12shared() + // Returns: A blurred texture lookup using a "virtual" 6x6 Gaussian + // blur (a 3x3 blur of carefully selected bilinear samples) + // of the given mip level. There will be some inaccuracies,subtle inaccuracies, + // especially for small or high-frequency detailed sources. + // Description: + // First see the description for tex2Dblur8x8shared(). This + // function shares the same concept and sample placement, but each fragment + // only uses 9 of the 16 samples taken across the pixel quad (to cover a + // 3x3 sample area, or 6x6 texel area), and it uses a lower standard + // deviation to compensate. Thanks to symmetry, the 7 omitted samples + // are always the "same:" + // 1adjx, 3adjx + // 2adjy, 3adjy + // 1diag, 2diag, 3diag + + // COMPUTE COORDS FOR TEXTURE SAMPLES THIS FRAGMENT IS RESPONSIBLE FOR: + // Statically compute bilinear sampling offsets (details in tex2Dblur12x12shared). + const float denom_inv = 0.5/(sigma*sigma); + const float w0off = 1.0; + const float w0_5off = exp(-(0.5*0.5) * denom_inv); + const float w1off = exp(-(1.0*1.0) * denom_inv); + const float w1_5off = exp(-(1.5*1.5) * denom_inv); + const float w2off = exp(-(2.0*2.0) * denom_inv); + const float w2_5off = exp(-(2.5*2.5) * denom_inv); + const float w3_5off = exp(-(3.5*3.5) * denom_inv); + const float texel0to1ratio = mix(w1_5off/(w0_5off + w1_5off), 0.5, error_blurring); + const float texel2to3ratio = mix(w3_5off/(w2_5off + w3_5off), 0.5, error_blurring); + // We don't share sample0*, so use the nearest destination fragment: + const float texel0to1ratio_nearest = w1off/(w0off + w1off); + const float texel1to2ratio_nearest = w2off/(w1off + w2off); + // Statically compute texel offsets from the bottom-right fragment to each + // bilinear sample in the bottom-right quadrant: + const vec2 sample0curr_texel_offset = vec2(0.0, 0.0) + vec2(texel0to1ratio_nearest, texel0to1ratio_nearest); + const vec2 sample0adjx_texel_offset = vec2(-1.0, 0.0) + vec2(-texel1to2ratio_nearest, texel0to1ratio_nearest); + const vec2 sample0adjy_texel_offset = vec2(0.0, -1.0) + vec2(texel0to1ratio_nearest, -texel1to2ratio_nearest); + const vec2 sample0diag_texel_offset = vec2(-1.0, -1.0) + vec2(-texel1to2ratio_nearest, -texel1to2ratio_nearest); + const vec2 sample1_texel_offset = vec2(2.0, 0.0) + vec2(texel2to3ratio, texel0to1ratio); + const vec2 sample2_texel_offset = vec2(0.0, 2.0) + vec2(texel0to1ratio, texel2to3ratio); + const vec2 sample3_texel_offset = vec2(2.0, 2.0) + vec2(texel2to3ratio, texel2to3ratio); + + // CALCULATE KERNEL WEIGHTS: + // Statically compute bilinear sample weights at each destination fragment + // from the sum of their 4 texel weights (details in tex2Dblur12x12shared). + #define GET_TEXEL_QUAD_WEIGHTS(xoff, yoff) \ + (exp(-LENGTH_SQ(vec2(xoff, yoff)) * denom_inv) + \ + exp(-LENGTH_SQ(vec2(xoff + 1.0, yoff)) * denom_inv) + \ + exp(-LENGTH_SQ(vec2(xoff, yoff + 1.0)) * denom_inv) + \ + exp(-LENGTH_SQ(vec2(xoff + 1.0, yoff + 1.0)) * denom_inv)) + // We only need 9 of the 16 sample weights. Skip the following weights: + // 1adjx, 3adjx + // 2adjy, 3adjy + // 1diag, 2diag, 3diag + const float w0diag = GET_TEXEL_QUAD_WEIGHTS(-2.0, -2.0); + const float w0adjy = GET_TEXEL_QUAD_WEIGHTS(0.0, -2.0); + const float w1adjy = GET_TEXEL_QUAD_WEIGHTS(2.0, -2.0); + const float w0adjx = GET_TEXEL_QUAD_WEIGHTS(-2.0, 0.0); + const float w0curr = GET_TEXEL_QUAD_WEIGHTS(0.0, 0.0); + const float w1curr = GET_TEXEL_QUAD_WEIGHTS(2.0, 0.0); + const float w2adjx = GET_TEXEL_QUAD_WEIGHTS(-2.0, 2.0); + const float w2curr = GET_TEXEL_QUAD_WEIGHTS(0.0, 2.0); + const float w3curr = GET_TEXEL_QUAD_WEIGHTS(2.0, 2.0); + #undef GET_TEXEL_QUAD_WEIGHTS + // Get the weight sum inverse (normalization factor): + const float weight_sum_inv = 1.0/(w0curr + w1curr + w2curr + w3curr + + w0adjx + w2adjx + w0adjy + w1adjy + w0diag); + // Statically pack some weights for runtime: + const vec4 w0 = vec4(w0curr, w0adjx, w0adjy, w0diag); + + // LOAD TEXTURE SAMPLES THIS FRAGMENT IS RESPONSIBLE FOR: + // Get a uv vector from texel 0q0 of this quadrant to texel 0q3: + const vec2 dxdy_curr = dxdy * quad_vector.xy; + // Load bilinear samples for the current quadrant (for this fragment): + const vec3 sample0curr = tex2D_linearize(texture, tex_uv.xy + dxdy_curr * sample0curr_texel_offset).rgb; + const vec3 sample0adjx = tex2D_linearize(texture, tex_uv.xy + dxdy_curr * sample0adjx_texel_offset).rgb; + const vec3 sample0adjy = tex2D_linearize(texture, tex_uv.xy + dxdy_curr * sample0adjy_texel_offset).rgb; + const vec3 sample0diag = tex2D_linearize(texture, tex_uv.xy + dxdy_curr * sample0diag_texel_offset).rgb; + const vec3 sample1curr = tex2Dlod_linearize(texture, tex_uv + uv2_to_uv4(dxdy_curr * sample1_texel_offset)).rgb; + const vec3 sample2curr = tex2Dlod_linearize(texture, tex_uv + uv2_to_uv4(dxdy_curr * sample2_texel_offset)).rgb; + const vec3 sample3curr = tex2Dlod_linearize(texture, tex_uv + uv2_to_uv4(dxdy_curr * sample3_texel_offset)).rgb; + + // GATHER NEIGHBORING SAMPLES AND SUM WEIGHTED SAMPLES: + // Fetch the samples from other fragments in the 2x2 quad: + vec3 sample1adjx, sample1adjy, sample1diag; + vec3 sample2adjx, sample2adjy, sample2diag; + quad_gather(quad_vector, sample1curr, sample1adjx, sample1adjy, sample1diag); + quad_gather(quad_vector, sample2curr, sample2adjx, sample2adjy, sample2diag); + // Statically normalize weights (so total = 1.0), and sum weighted samples. + // Fill each row of a matrix with an rgb sample and pre-multiply by the + // weights to obtain a weighted result for sample1*, and handle the rest + // of the weights more directly/verbosely: + vec3 sum = vec3(0.0); + sum += (mat4x3(sample0curr, sample0adjx, sample0adjy, sample0diag) * w0); + sum += w1curr * sample1curr + w1adjy * sample1adjy + w2curr * sample2curr + + w2adjx * sample2adjx + w3curr * sample3curr; + return sum * weight_sum_inv; +} + + +/////////////////////// MAX OPTIMAL SIGMA BLUR WRAPPERS ////////////////////// + +// The following blurs are static wrappers around the dynamic blurs above. +// HOPEFULLY, the compiler will be smart enough to do constant-folding. + +// Resizable separable blurs: +vec3 tex2Dblur11resize(const sampler2D texture, const vec2 tex_uv, + const vec2 dxdy) +{ + return tex2Dblur11resize(texture, tex_uv, dxdy, blur11_std_dev); +} +vec3 tex2Dblur9resize(const sampler2D texture, const vec2 tex_uv, + const vec2 dxdy) +{ + return tex2Dblur9resize(texture, tex_uv, dxdy, blur9_std_dev); +} +vec3 tex2Dblur7resize(const sampler2D texture, const vec2 tex_uv, + const vec2 dxdy) +{ + return tex2Dblur7resize(texture, tex_uv, dxdy, blur7_std_dev); +} +vec3 tex2Dblur5resize(const sampler2D texture, const vec2 tex_uv, + const vec2 dxdy) +{ + return tex2Dblur5resize(texture, tex_uv, dxdy, blur5_std_dev); +} +vec3 tex2Dblur3resize(const sampler2D texture, const vec2 tex_uv, + const vec2 dxdy) +{ + return tex2Dblur3resize(texture, tex_uv, dxdy, blur3_std_dev); +} +// Fast separable blurs: +vec3 tex2Dblur11fast(const sampler2D texture, const vec2 tex_uv, + const vec2 dxdy) +{ + return tex2Dblur11fast(texture, tex_uv, dxdy, blur11_std_dev); +} +vec3 tex2Dblur9fast(const sampler2D texture, const vec2 tex_uv, + const vec2 dxdy) +{ + return tex2Dblur9fast(texture, tex_uv, dxdy, blur9_std_dev); +} +vec3 tex2Dblur7fast(const sampler2D texture, const vec2 tex_uv, + const vec2 dxdy) +{ + return tex2Dblur7fast(texture, tex_uv, dxdy, blur7_std_dev); +} +vec3 tex2Dblur5fast(const sampler2D texture, const vec2 tex_uv, + const vec2 dxdy) +{ + return tex2Dblur5fast(texture, tex_uv, dxdy, blur5_std_dev); +} +vec3 tex2Dblur3fast(const sampler2D texture, const vec2 tex_uv, + const vec2 dxdy) +{ + return tex2Dblur3fast(texture, tex_uv, dxdy, blur3_std_dev); +} +// Huge, "fast" separable blurs: +vec3 tex2Dblur43fast(const sampler2D texture, const vec2 tex_uv, + const vec2 dxdy) +{ + return tex2Dblur43fast(texture, tex_uv, dxdy, blur43_std_dev); +} +vec3 tex2Dblur31fast(const sampler2D texture, const vec2 tex_uv, + const vec2 dxdy) +{ + return tex2Dblur31fast(texture, tex_uv, dxdy, blur31_std_dev); +} +vec3 tex2Dblur25fast(const sampler2D texture, const vec2 tex_uv, + const vec2 dxdy) +{ + return tex2Dblur25fast(texture, tex_uv, dxdy, blur25_std_dev); +} +vec3 tex2Dblur17fast(const sampler2D texture, const vec2 tex_uv, + const vec2 dxdy) +{ + return tex2Dblur17fast(texture, tex_uv, dxdy, blur17_std_dev); +} +// Resizable one-pass blurs: +vec3 tex2Dblur3x3resize(const sampler2D texture, const vec2 tex_uv, + const vec2 dxdy) +{ + return tex2Dblur3x3resize(texture, tex_uv, dxdy, blur3_std_dev); +} +// "Fast" one-pass blurs: +vec3 tex2Dblur9x9(const sampler2D texture, const vec2 tex_uv, + const vec2 dxdy) +{ + return tex2Dblur9x9(texture, tex_uv, dxdy, blur9_std_dev); +} +vec3 tex2Dblur7x7(const sampler2D texture, const vec2 tex_uv, + const vec2 dxdy) +{ + return tex2Dblur7x7(texture, tex_uv, dxdy, blur7_std_dev); +} +vec3 tex2Dblur5x5(const sampler2D texture, const vec2 tex_uv, + const vec2 dxdy) +{ + return tex2Dblur5x5(texture, tex_uv, dxdy, blur5_std_dev); +} +vec3 tex2Dblur3x3(const sampler2D texture, const vec2 tex_uv, + const vec2 dxdy) +{ + return tex2Dblur3x3(texture, tex_uv, dxdy, blur3_std_dev); +} +// "Fast" shared-sample one-pass blurs: +vec3 tex2Dblur12x12shared(const sampler2D texture, + const vec4 tex_uv, const vec2 dxdy, const vec4 quad_vector) +{ + return tex2Dblur12x12shared(texture, tex_uv, dxdy, quad_vector, blur12_std_dev); +} +vec3 tex2Dblur10x10shared(const sampler2D texture, + const vec4 tex_uv, const vec2 dxdy, const vec4 quad_vector) +{ + return tex2Dblur10x10shared(texture, tex_uv, dxdy, quad_vector, blur10_std_dev); +} +vec3 tex2Dblur8x8shared(const sampler2D texture, + const vec4 tex_uv, const vec2 dxdy, const vec4 quad_vector) +{ + return tex2Dblur8x8shared(texture, tex_uv, dxdy, quad_vector, blur8_std_dev); +} +vec3 tex2Dblur6x6shared(const sampler2D texture, + const vec4 tex_uv, const vec2 dxdy, const vec4 quad_vector) +{ + return tex2Dblur6x6shared(texture, tex_uv, dxdy, quad_vector, blur6_std_dev); +} + + +#endif // BLUR_FUNCTIONS_H + diff --git a/include/blur-functions.h b/include/blur-functions.h new file mode 100644 index 0000000..25b83dc --- /dev/null +++ b/include/blur-functions.h @@ -0,0 +1,281 @@ +#define BLUR_FUNCTIONS + +///////////////////////////////// MIT LICENSE //////////////////////////////// + +// Copyright (C) 2014 TroggleMonkey +// +// 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. + +///////////////////////////////// DESCRIPTION //////////////////////////////// + +// This file provides reusable one-pass and separable (two-pass) blurs. +// Requires: All blurs share these requirements (dxdy requirement is split): +// 1.) All requirements of gamma-management.h must be satisfied! +// 2.) filter_linearN must == "true" in your .cgp preset unless +// you're using tex2DblurNresize at 1x scale. +// 3.) mipmap_inputN must == "true" in your .cgp preset if +// IN.output_size < IN.video_size. +// 4.) IN.output_size == IN.video_size / pow(2, M), where M is some +// positive integer. tex2Dblur*resize can resize arbitrarily +// (and the blur will be done after resizing), but arbitrary +// resizes "fail" with other blurs due to the way they mix +// static weights with bilinear sample exploitation. +// 5.) In general, dxdy should contain the uv pixel spacing: +// dxdy = (IN.video_size/IN.output_size)/IN.texture_size +// 6.) For separable blurs (tex2DblurNresize and tex2DblurNfast), +// zero out the dxdy component in the unblurred dimension: +// dxdy = vec2(dxdy.x, 0.0) or vec2(0.0, dxdy.y) +// Many blurs share these requirements: +// 1.) One-pass blurs require scale_xN == scale_yN or scales > 1.0, +// or they will blur more in the lower-scaled dimension. +// 2.) One-pass shared sample blurs require ddx(), ddy(), and +// tex2Dlod() to be supported by the current Cg profile, and +// the drivers must support high-quality derivatives. +// 3.) One-pass shared sample blurs require: +// tex_uv.w == log2(IN.video_size/IN.output_size).y; +// Non-wrapper blurs share this requirement: +// 1.) sigma is the intended standard deviation of the blur +// Wrapper blurs share this requirement, which is automatically +// met (unless OVERRIDE_BLUR_STD_DEVS is #defined; see below): +// 1.) blurN_std_dev must be global static const float values +// specifying standard deviations for Nx blurs in units +// of destination pixels +// Optional: 1.) The including file (or an earlier included file) may +// optionally #define USE_BINOMIAL_BLUR_STD_DEVS to replace +// default standard deviations with those matching a binomial +// distribution. (See below for details/properties.) +// 2.) The including file (or an earlier included file) may +// optionally #define OVERRIDE_BLUR_STD_DEVS and override: +// static const float blur3_std_dev +// static const float blur4_std_dev +// static const float blur5_std_dev +// static const float blur6_std_dev +// static const float blur7_std_dev +// static const float blur8_std_dev +// static const float blur9_std_dev +// static const float blur10_std_dev +// static const float blur11_std_dev +// static const float blur12_std_dev +// static const float blur17_std_dev +// static const float blur25_std_dev +// static const float blur31_std_dev +// static const float blur43_std_dev +// 3.) The including file (or an earlier included file) may +// optionally #define OVERRIDE_ERROR_BLURRING and override: +// static const float error_blurring +// This tuning value helps mitigate weighting errors from one- +// pass shared-sample blurs sharing bilinear samples between +// fragments. Values closer to 0.0 have "correct" blurriness +// but allow more artifacts, and values closer to 1.0 blur away +// artifacts by sampling closer to halfway between texels. +// UPDATE 6/21/14: The above static constants may now be overridden +// by non-static uniform constants. This permits exposing blur +// standard deviations as runtime GUI shader parameters. However, +// using them keeps weights from being statically computed, and the +// speed hit depends on the blur: On my machine, uniforms kill over +// 53% of the framerate with tex2Dblur12x12shared, but they only +// drop the framerate by about 18% with tex2Dblur11fast. +// Quality and Performance Comparisons: +// For the purposes of the following discussion, "no sRGB" means +// GAMMA_ENCODE_EVERY_FBO is #defined, and "sRGB" means it isn't. +// 1.) tex2DblurNfast is always faster than tex2DblurNresize. +// 2.) tex2DblurNresize functions are the only ones that can arbitrarily resize +// well, because they're the only ones that don't exploit bilinear samples. +// This also means they're the only functions which can be truly gamma- +// correct without linear (or sRGB FBO) input, but only at 1x scale. +// 3.) One-pass shared sample blurs only have a speed advantage without sRGB. +// They also have some inaccuracies due to their shared-[bilinear-]sample +// design, which grow increasingly bothersome for smaller blurs and higher- +// frequency source images (relative to their resolution). I had high +// hopes for them, but their most realistic use case is limited to quickly +// reblurring an already blurred input at full resolution. Otherwise: +// a.) If you're blurring a low-resolution source, you want a better blur. +// b.) If you're blurring a lower mipmap, you want a better blur. +// c.) If you're blurring a high-resolution, high-frequency source, you +// want a better blur. +// 4.) The one-pass blurs without shared samples grow slower for larger blurs, +// but they're competitive with separable blurs at 5x5 and smaller, and +// even tex2Dblur7x7 isn't bad if you're wanting to conserve passes. +// Here are some framerates from a GeForce 8800GTS. The first pass resizes to +// viewport size (4x in this test) and linearizes for sRGB codepaths, and the +// remaining passes perform 6 full blurs. Mipmapped tests are performed at the +// same scale, so they just measure the cost of mipmapping each FBO (only every +// other FBO is mipmapped for separable blurs, to mimic realistic usage). +// Mipmap Neither sRGB+Mipmap sRGB Function +// 76.0 92.3 131.3 193.7 tex2Dblur3fast +// 63.2 74.4 122.4 175.5 tex2Dblur3resize +// 93.7 121.2 159.3 263.2 tex2Dblur3x3 +// 59.7 68.7 115.4 162.1 tex2Dblur3x3resize +// 63.2 74.4 122.4 175.5 tex2Dblur5fast +// 49.3 54.8 100.0 132.7 tex2Dblur5resize +// 59.7 68.7 115.4 162.1 tex2Dblur5x5 +// 64.9 77.2 99.1 137.2 tex2Dblur6x6shared +// 55.8 63.7 110.4 151.8 tex2Dblur7fast +// 39.8 43.9 83.9 105.8 tex2Dblur7resize +// 40.0 44.2 83.2 104.9 tex2Dblur7x7 +// 56.4 65.5 71.9 87.9 tex2Dblur8x8shared +// 49.3 55.1 99.9 132.5 tex2Dblur9fast +// 33.3 36.2 72.4 88.0 tex2Dblur9resize +// 27.8 29.7 61.3 72.2 tex2Dblur9x9 +// 37.2 41.1 52.6 60.2 tex2Dblur10x10shared +// 44.4 49.5 91.3 117.8 tex2Dblur11fast +// 28.8 30.8 63.6 75.4 tex2Dblur11resize +// 33.6 36.5 40.9 45.5 tex2Dblur12x12shared +// TODO: Fill in benchmarks for new untested blurs. +// tex2Dblur17fast +// tex2Dblur25fast +// tex2Dblur31fast +// tex2Dblur43fast +// tex2Dblur3x3resize + +///////////////////////////// SETTINGS MANAGEMENT //////////////////////////// + +// Set static standard deviations, but allow users to override them with their +// own constants (even non-static uniforms if they're okay with the speed hit): +#ifndef OVERRIDE_BLUR_STD_DEVS + // blurN_std_dev values are specified in terms of dxdy strides. + #ifdef USE_BINOMIAL_BLUR_STD_DEVS + // By request, we can define standard deviations corresponding to a + // binomial distribution with p = 0.5 (related to Pascal's triangle). + // This distribution works such that blurring multiple times should + // have the same result as a single larger blur. These values are + // larger than default for blurs up to 6x and smaller thereafter. + const float blur3_std_dev = 0.84931640625; + const float blur4_std_dev = 0.84931640625; + const float blur5_std_dev = 1.0595703125; + const float blur6_std_dev = 1.06591796875; + const float blur7_std_dev = 1.17041015625; + const float blur8_std_dev = 1.1720703125; + const float blur9_std_dev = 1.2259765625; + const float blur10_std_dev = 1.21982421875; + const float blur11_std_dev = 1.25361328125; + const float blur12_std_dev = 1.2423828125; + const float blur17_std_dev = 1.27783203125; + const float blur25_std_dev = 1.2810546875; + const float blur31_std_dev = 1.28125; + const float blur43_std_dev = 1.28125; + #else + // The defaults are the largest values that keep the largest unused + // blur term on each side <= 1.0/256.0. (We could get away with more + // or be more conservative, but this compromise is pretty reasonable.) + const float blur3_std_dev = 0.62666015625; + const float blur4_std_dev = 0.66171875; + const float blur5_std_dev = 0.9845703125; + const float blur6_std_dev = 1.02626953125; + const float blur7_std_dev = 1.36103515625; + const float blur8_std_dev = 1.4080078125; + const float blur9_std_dev = 1.7533203125; + const float blur10_std_dev = 1.80478515625; + const float blur11_std_dev = 2.15986328125; + const float blur12_std_dev = 2.215234375; + const float blur17_std_dev = 3.45535583496; + const float blur25_std_dev = 5.3409576416; + const float blur31_std_dev = 6.86488037109; + const float blur43_std_dev = 10.1852050781; + #endif // USE_BINOMIAL_BLUR_STD_DEVS +#endif // OVERRIDE_BLUR_STD_DEVS + +#ifndef OVERRIDE_ERROR_BLURRING + // error_blurring should be in [0.0, 1.0]. Higher values reduce ringing + // in shared-sample blurs but increase blurring and feature shifting. + const float error_blurring = 0.5; +#endif + +// Make a length squared helper macro (for usage with static constants): +#define LENGTH_SQ(vec) (dot(vec, vec)) + +/////////////////////////////////// HELPERS ////////////////////////////////// + +vec4 uv2_to_uv4(vec2 tex_uv) +{ + // Make a vec2 uv offset safe for adding to vec4 tex2Dlod coords: + return vec4(tex_uv, 0.0, 0.0); +} + +// Make a length squared helper macro (for usage with static constants): +#define LENGTH_SQ(vec) (dot(vec, vec)) + +float get_fast_gaussian_weight_sum_inv(const float sigma) +{ + // We can use the Gaussian integral to calculate the asymptotic weight for + // the center pixel. Since the unnormalized center pixel weight is 1.0, + // the normalized weight is the same as the weight sum inverse. Given a + // large enough blur (9+), the asymptotic weight sum is close and faster: + // center_weight = 0.5 * + // (erf(0.5/(sigma*sqrt(2.0))) - erf(-0.5/(sigma*sqrt(2.0)))) + // erf(-x) == -erf(x), so we get 0.5 * (2.0 * erf(blah blah)): + // However, we can get even faster results with curve-fitting. These are + // also closer than the asymptotic results, because they were constructed + // from 64 blurs sizes from [3, 131) and 255 equally-spaced sigmas from + // (0, blurN_std_dev), so the results for smaller sigmas are biased toward + // smaller blurs. The max error is 0.0031793913. + // Relative FPS: 134.3 with erf, 135.8 with curve-fitting. + //static const float temp = 0.5/sqrt(2.0); + //return erf(temp/sigma); + return min(exp(exp(0.348348412457428/ + (sigma - 0.0860587260734721))), 0.399334576340352/sigma); +} + +//////////////////// ARBITRARILY RESIZABLE ONE-PASS BLURS //////////////////// + +vec3 tex2Dblur3x3resize(const sampler2D tex, const vec2 tex_uv, + const vec2 dxdy, const float sigma) +{ + // Requires: Global requirements must be met (see file description). + // Returns: A 3x3 Gaussian blurred mipmapped texture lookup of the + // resized input. + // Description: + // This is the only arbitrarily resizable one-pass blur; tex2Dblur5x5resize + // would perform like tex2Dblur9x9, MUCH slower than tex2Dblur5resize. + const float denom_inv = 0.5/(sigma*sigma); + // Load each sample. We need all 3x3 samples. Quad-pixel communication + // won't help either: This should perform like tex2Dblur5x5, but sharing a + // 4x4 sample field would perform more like tex2Dblur8x8shared (worse). + const vec2 sample4_uv = tex_uv; + const vec2 dx = vec2(dxdy.x, 0.0); + const vec2 dy = vec2(0.0, dxdy.y); + const vec2 sample1_uv = sample4_uv - dy; + const vec2 sample7_uv = sample4_uv + dy; + const vec3 sample0 = tex2D_linearize(tex, sample1_uv - dx).rgb; + const vec3 sample1 = tex2D_linearize(tex, sample1_uv).rgb; + const vec3 sample2 = tex2D_linearize(tex, sample1_uv + dx).rgb; + const vec3 sample3 = tex2D_linearize(tex, sample4_uv - dx).rgb; + const vec3 sample4 = tex2D_linearize(tex, sample4_uv).rgb; + const vec3 sample5 = tex2D_linearize(tex, sample4_uv + dx).rgb; + const vec3 sample6 = tex2D_linearize(tex, sample7_uv - dx).rgb; + const vec3 sample7 = tex2D_linearize(tex, sample7_uv).rgb; + const vec3 sample8 = tex2D_linearize(tex, sample7_uv + dx).rgb; + // Statically compute Gaussian sample weights: + const float w4 = 1.0; + const float w1_3_5_7 = exp(-LENGTH_SQ(vec2(1.0, 0.0)) * denom_inv); + const float w0_2_6_8 = exp(-LENGTH_SQ(vec2(1.0, 1.0)) * denom_inv); + const float weight_sum_inv = 1.0/(w4 + 4.0 * (w1_3_5_7 + w0_2_6_8)); + // Weight and sum the samples: + const vec3 sum = w4 * sample4 + + w1_3_5_7 * (sample1 + sample3 + sample5 + sample7) + + w0_2_6_8 * (sample0 + sample2 + sample6 + sample8); + return sum * weight_sum_inv; +} + +// Resizable one-pass blurs: +vec3 tex2Dblur3x3resize(const sampler2D texture, const vec2 tex_uv, + const vec2 dxdy) +{ + return tex2Dblur3x3resize(texture, tex_uv, dxdy, blur3_std_dev); +} \ No newline at end of file diff --git a/include/gamma-management-old.h b/include/gamma-management-old.h new file mode 100644 index 0000000..18963c7 --- /dev/null +++ b/include/gamma-management-old.h @@ -0,0 +1,547 @@ +#ifndef GAMMA_MANAGEMENT_H +#define GAMMA_MANAGEMENT_H + +///////////////////////////////// MIT LICENSE //////////////////////////////// + +// Copyright (C) 2014 TroggleMonkey +// +// 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. + +///////////////////////////////// DESCRIPTION //////////////////////////////// + +// This file provides gamma-aware tex*D*() and encode_output() functions. +// Requires: Before #include-ing this file, the including file must #define +// the following macros when applicable and follow their rules: +// 1.) #define FIRST_PASS if this is the first pass. +// 2.) #define LAST_PASS if this is the last pass. +// 3.) If sRGB is available, set srgb_framebufferN = "true" for +// every pass except the last in your .cgp preset. +// 4.) If sRGB isn't available but you want gamma-correctness with +// no banding, #define GAMMA_ENCODE_EVERY_FBO each pass. +// 5.) #define SIMULATE_CRT_ON_LCD if desired (precedence over 5-7) +// 6.) #define SIMULATE_GBA_ON_LCD if desired (precedence over 6-7) +// 7.) #define SIMULATE_LCD_ON_CRT if desired (precedence over 7) +// 8.) #define SIMULATE_GBA_ON_CRT if desired (precedence over -) +// If an option in [5, 8] is #defined in the first or last pass, it +// should be #defined for both. It shouldn't make a difference +// whether it's #defined for intermediate passes or not. +// Optional: The including file (or an earlier included file) may optionally +// #define a number of macros indicating it will override certain +// macros and associated constants are as follows: +// static constants with either static or uniform constants. The +// 1.) OVERRIDE_STANDARD_GAMMA: The user must first define: +// static const float ntsc_gamma +// static const float pal_gamma +// static const float crt_reference_gamma_high +// static const float crt_reference_gamma_low +// static const float lcd_reference_gamma +// static const float crt_office_gamma +// static const float lcd_office_gamma +// 2.) OVERRIDE_DEVICE_GAMMA: The user must first define: +// static const float crt_gamma +// static const float gba_gamma +// static const float lcd_gamma +// 3.) OVERRIDE_FINAL_GAMMA: The user must first define: +// static const float input_gamma +// static const float intermediate_gamma +// static const float output_gamma +// (intermediate_gamma is for GAMMA_ENCODE_EVERY_FBO.) +// 4.) OVERRIDE_ALPHA_ASSUMPTIONS: The user must first define: +// static const bool assume_opaque_alpha +// The gamma constant overrides must be used in every pass or none, +// and OVERRIDE_FINAL_GAMMA bypasses all of the SIMULATE* macros. +// OVERRIDE_ALPHA_ASSUMPTIONS may be set on a per-pass basis. +// Usage: After setting macros appropriately, ignore gamma correction and +// replace all tex*D*() calls with equivalent gamma-aware +// tex*D*_linearize calls, except: +// 1.) When you read an LUT, use regular tex*D or a gamma-specified +// function, depending on its gamma encoding: +// tex*D*_linearize_gamma (takes a runtime gamma parameter) +// 2.) If you must read pass0's original input in a later pass, use +// tex2D_linearize_ntsc_gamma. If you want to read pass0's +// input with gamma-corrected bilinear filtering, consider +// creating a first linearizing pass and reading from the input +// of pass1 later. +// Then, return encode_output(color) from every fragment shader. +// Finally, use the global gamma_aware_bilinear boolean if you want +// to statically branch based on whether bilinear filtering is +// gamma-correct or not (e.g. for placing Gaussian blur samples). +// +// Detailed Policy: +// tex*D*_linearize() functions enforce a consistent gamma-management policy +// based on the FIRST_PASS and GAMMA_ENCODE_EVERY_FBO settings. They assume +// their input texture has the same encoding characteristics as the input for +// the current pass (which doesn't apply to the exceptions listed above). +// Similarly, encode_output() enforces a policy based on the LAST_PASS and +// GAMMA_ENCODE_EVERY_FBO settings. Together, they result in one of the +// following two pipelines. +// Typical pipeline with intermediate sRGB framebuffers: +// linear_color = pow(pass0_encoded_color, input_gamma); +// intermediate_output = linear_color; // Automatic sRGB encoding +// linear_color = intermediate_output; // Automatic sRGB decoding +// final_output = pow(intermediate_output, 1.0/output_gamma); +// Typical pipeline without intermediate sRGB framebuffers: +// linear_color = pow(pass0_encoded_color, input_gamma); +// intermediate_output = pow(linear_color, 1.0/intermediate_gamma); +// linear_color = pow(intermediate_output, intermediate_gamma); +// final_output = pow(intermediate_output, 1.0/output_gamma); +// Using GAMMA_ENCODE_EVERY_FBO is much slower, but it's provided as a way to +// easily get gamma-correctness without banding on devices where sRGB isn't +// supported. +// +// Use This Header to Maximize Code Reuse: +// The purpose of this header is to provide a consistent interface for texture +// reads and output gamma-encoding that localizes and abstracts away all the +// annoying details. This greatly reduces the amount of code in each shader +// pass that depends on the pass number in the .cgp preset or whether sRGB +// FBO's are being used: You can trivially change the gamma behavior of your +// whole pass by commenting or uncommenting 1-3 #defines. To reuse the same +// code in your first, Nth, and last passes, you can even put it all in another +// header file and #include it from skeleton .cg files that #define the +// appropriate pass-specific settings. +// +// Rationale for Using Three Macros: +// This file uses GAMMA_ENCODE_EVERY_FBO instead of an opposite macro like +// SRGB_PIPELINE to ensure sRGB is assumed by default, which hopefully imposes +// a lower maintenance burden on each pass. At first glance it seems we could +// accomplish everything with two macros: GAMMA_CORRECT_IN / GAMMA_CORRECT_OUT. +// This works for simple use cases where input_gamma == output_gamma, but it +// breaks down for more complex scenarios like CRT simulation, where the pass +// number determines the gamma encoding of the input and output. + + +/////////////////////////////// BASE CONSTANTS /////////////////////////////// + +// Set standard gamma constants, but allow users to override them: +#ifndef OVERRIDE_STANDARD_GAMMA + // Standard encoding gammas: + const float ntsc_gamma = 2.2; // Best to use NTSC for PAL too? + const float pal_gamma = 2.8; // Never actually 2.8 in practice + // Typical device decoding gammas (only use for emulating devices): + // CRT/LCD reference gammas are higher than NTSC and Rec.709 video standard + // gammas: The standards purposely undercorrected for an analog CRT's + // assumed 2.5 reference display gamma to maintain contrast in assumed + // [dark] viewing conditions: http://www.poynton.com/PDFs/GammaFAQ.pdf + // These unstated assumptions about display gamma and perceptual rendering + // intent caused a lot of confusion, and more modern CRT's seemed to target + // NTSC 2.2 gamma with circuitry. LCD displays seem to have followed suit + // (they struggle near black with 2.5 gamma anyway), especially PC/laptop + // displays designed to view sRGB in bright environments. (Standards are + // also in flux again with BT.1886, but it's underspecified for displays.) + const float crt_reference_gamma_high = 2.5; // In (2.35, 2.55) + const float crt_reference_gamma_low = 2.35; // In (2.35, 2.55) + const float lcd_reference_gamma = 2.5; // To match CRT + const float crt_office_gamma = 2.2; // Circuitry-adjusted for NTSC + const float lcd_office_gamma = 2.2; // Approximates sRGB +#endif // OVERRIDE_STANDARD_GAMMA + +// Assuming alpha == 1.0 might make it easier for users to avoid some bugs, +// but only if they're aware of it. +#ifndef OVERRIDE_ALPHA_ASSUMPTIONS + const bool assume_opaque_alpha = false; +#endif + + +/////////////////////// DERIVED CONSTANTS AS FUNCTIONS /////////////////////// + +// gamma-management.h should be compatible with overriding gamma values with +// runtime user parameters, but we can only define other global constants in +// terms of static constants, not uniform user parameters. To get around this +// limitation, we need to define derived constants using functions. + +// Set device gamma constants, but allow users to override them: +#ifdef OVERRIDE_DEVICE_GAMMA + // The user promises to globally define the appropriate constants: + float get_crt_gamma() { return crt_gamma; } + float get_gba_gamma() { return gba_gamma; } + float get_lcd_gamma() { return lcd_gamma; } +#else + float get_crt_gamma() { return crt_reference_gamma_high; } + float get_gba_gamma() { return 3.5; } // Game Boy Advance; in (3.0, 4.0) + float get_lcd_gamma() { return lcd_office_gamma; } +#endif // OVERRIDE_DEVICE_GAMMA + +// Set decoding/encoding gammas for the first/lass passes, but allow overrides: +#ifdef OVERRIDE_FINAL_GAMMA + // The user promises to globally define the appropriate constants: + float get_intermediate_gamma() { return intermediate_gamma; } + float get_input_gamma() { return input_gamma; } + float get_output_gamma() { return output_gamma; } +#else + // If we gamma-correct every pass, always use ntsc_gamma between passes to + // ensure middle passes don't need to care if anything is being simulated: + float get_intermediate_gamma() { return ntsc_gamma; } + #ifdef SIMULATE_CRT_ON_LCD + float get_input_gamma() { return get_crt_gamma(); } + float get_output_gamma() { return get_lcd_gamma(); } + #else + #ifdef SIMULATE_GBA_ON_LCD + float get_input_gamma() { return get_gba_gamma(); } + float get_output_gamma() { return get_lcd_gamma(); } + #else + #ifdef SIMULATE_LCD_ON_CRT + float get_input_gamma() { return get_lcd_gamma(); } + float get_output_gamma() { return get_crt_gamma(); } + #else + #ifdef SIMULATE_GBA_ON_CRT + float get_input_gamma() { return get_gba_gamma(); } + float get_output_gamma() { return get_crt_gamma(); } + #else // Don't simulate anything: + float get_input_gamma() { return ntsc_gamma; } + float get_output_gamma() { return ntsc_gamma; } + #endif // SIMULATE_GBA_ON_CRT + #endif // SIMULATE_LCD_ON_CRT + #endif // SIMULATE_GBA_ON_LCD + #endif // SIMULATE_CRT_ON_LCD +#endif // OVERRIDE_FINAL_GAMMA + +// Set decoding/encoding gammas for the current pass. Use static constants for +// linearize_input and gamma_encode_output, because they aren't derived, and +// they let the compiler do dead-code elimination. +#ifndef GAMMA_ENCODE_EVERY_FBO + #ifdef FIRST_PASS + const bool linearize_input = true; + float get_pass_input_gamma() { return get_input_gamma(); } + #else + const bool linearize_input = false; + float get_pass_input_gamma() { return 1.0; } + #endif + #ifdef LAST_PASS + const bool gamma_encode_output = true; + float get_pass_output_gamma() { return get_output_gamma(); } + #else + const bool gamma_encode_output = false; + float get_pass_output_gamma() { return 1.0; } + #endif +#else + const bool linearize_input = true; + const bool gamma_encode_output = true; + #ifdef FIRST_PASS + float get_pass_input_gamma() { return get_input_gamma(); } + #else + float get_pass_input_gamma() { return get_intermediate_gamma(); } + #endif + #ifdef LAST_PASS + float get_pass_output_gamma() { return get_output_gamma(); } + #else + float get_pass_output_gamma() { return get_intermediate_gamma(); } + #endif +#endif + +// Users might want to know if bilinear filtering will be gamma-correct: +const bool gamma_aware_bilinear = !linearize_input; + + +////////////////////// COLOR ENCODING/DECODING FUNCTIONS ///////////////////// + +vec4 encode_output(const vec4 color) +{ + if(gamma_encode_output) + { + if(assume_opaque_alpha) + { + return vec4(pow(color.rgb, vec3(1.0/get_pass_output_gamma())), 1.0); + } + else + { + return vec4(pow(color.rgb, vec3(1.0/get_pass_output_gamma())), color.a); + } + } + else + { + return color; + } +} + +vec4 decode_input(const vec4 color) +{ + if(linearize_input) + { + if(assume_opaque_alpha) + { + return vec4(pow(color.rgb, vec3(get_pass_input_gamma())), 1.0); + } + else + { + return vec4(pow(color.rgb, vec3(get_pass_input_gamma())), color.a); + } + } + else + { + return color; + } +} + +vec4 decode_gamma_input(const vec4 color, const vec3 gamma) +{ + if(assume_opaque_alpha) + { + return vec4(pow(color.rgb, vec3(gamma)), 1.0); + } + else + { + return vec4(pow(color.rgb, vec3(gamma)), color.a); + } +} + + +/////////////////////////// TEXTURE LOOKUP WRAPPERS ////////////////////////// + +// "SMART" LINEARIZING TEXTURE LOOKUP FUNCTIONS: +// Provide a wide array of linearizing texture lookup wrapper functions. The +// Cg shader spec Retroarch uses only allows for 2D textures, but 1D and 3D +// lookups are provided for completeness in case that changes someday. Nobody +// is likely to use the *fetch and *proj functions, but they're included just +// in case. The only tex*D texture sampling functions omitted are: +// - tex*Dcmpbias +// - tex*Dcmplod +// - tex*DARRAY* +// - tex*DMS* +// - Variants returning integers +// Standard line length restrictions are ignored below for vertical brevity. + +/* +// tex1D: +vec4 tex1D_linearize(const sampler1D texture, const float tex_coords) +{ return decode_input(tex1D(texture, tex_coords)); } + +vec4 tex1D_linearize(const sampler1D texture, const vec2 tex_coords) +{ return decode_input(tex1D(texture, tex_coords)); } + +vec4 tex1D_linearize(const sampler1D texture, const float tex_coords, const int texel_off) +{ return decode_input(tex1D(texture, tex_coords, texel_off)); } + +vec4 tex1D_linearize(const sampler1D texture, const vec2 tex_coords, const int texel_off) +{ return decode_input(tex1D(texture, tex_coords, texel_off)); } + +vec4 tex1D_linearize(const sampler1D texture, const float tex_coords, const float dx, const float dy) +{ return decode_input(tex1D(texture, tex_coords, dx, dy)); } + +vec4 tex1D_linearize(const sampler1D texture, const vec2 tex_coords, const float dx, const float dy) +{ return decode_input(tex1D(texture, tex_coords, dx, dy)); } + +vec4 tex1D_linearize(const sampler1D texture, const float tex_coords, const float dx, const float dy, const int texel_off) +{ return decode_input(tex1D(texture, tex_coords, dx, dy, texel_off)); } + +vec4 tex1D_linearize(const sampler1D texture, const vec2 tex_coords, const float dx, const float dy, const int texel_off) +{ return decode_input(tex1D(texture, tex_coords, dx, dy, texel_off)); } + +// tex1Dbias: +vec4 tex1Dbias_linearize(const sampler1D texture, const vec4 tex_coords) +{ return decode_input(tex1Dbias(texture, tex_coords)); } + +vec4 tex1Dbias_linearize(const sampler1D texture, const vec4 tex_coords, const int texel_off) +{ return decode_input(tex1Dbias(texture, tex_coords, texel_off)); } + +// tex1Dfetch: +vec4 tex1Dfetch_linearize(const sampler1D texture, const int4 tex_coords) +{ return decode_input(tex1Dfetch(texture, tex_coords)); } + +vec4 tex1Dfetch_linearize(const sampler1D texture, const int4 tex_coords, const int texel_off) +{ return decode_input(tex1Dfetch(texture, tex_coords, texel_off)); } + +// tex1Dlod: +vec4 tex1Dlod_linearize(const sampler1D texture, const vec4 tex_coords) +{ return decode_input(tex1Dlod(texture, tex_coords)); } + +vec4 tex1Dlod_linearize(const sampler1D texture, const vec4 tex_coords, const int texel_off) +{ return decode_input(tex1Dlod(texture, tex_coords, texel_off)); } + +// tex1Dproj: +vec4 tex1Dproj_linearize(const sampler1D texture, const vec2 tex_coords) +{ return decode_input(tex1Dproj(texture, tex_coords)); } + +vec4 tex1Dproj_linearize(const sampler1D texture, const vec3 tex_coords) +{ return decode_input(tex1Dproj(texture, tex_coords)); } + +vec4 tex1Dproj_linearize(const sampler1D texture, const vec2 tex_coords, const int texel_off) +{ return decode_input(tex1Dproj(texture, tex_coords, texel_off)); } + +vec4 tex1Dproj_linearize(const sampler1D texture, const vec3 tex_coords, const int texel_off) +{ return decode_input(tex1Dproj(texture, tex_coords, texel_off)); } +*/ + +// tex2D: +vec4 tex2D_linearize(const sampler2D tex, const vec2 tex_coords) +{ return decode_input(vec4(texture(tex, tex_coords))); } + +vec4 tex2D_linearize(const sampler2D tex, const vec3 tex_coords) +{ return decode_input(vec4(texture(tex, tex_coords))); } + +vec4 tex2D_linearize(const sampler2D tex, const vec2 tex_coords, const int texel_off) +{ return decode_input(vec4(texture(tex, tex_coords, texel_off))); } + +vec4 tex2D_linearize(const sampler2D tex, const vec3 tex_coords, const int texel_off) +{ return decode_input(vec4(texture(tex, tex_coords, texel_off))); } + +vec4 tex2D_linearize(const sampler2D tex, const vec2 tex_coords, const vec2 dx, const vec2 dy) +{ return decode_input(vec4(texture(tex, tex_coords, dx, dy))); } + +vec4 tex2D_linearize(const sampler2D tex, const vec3 tex_coords, const vec2 dx, const vec2 dy) +{ return decode_input(vec4(texture(tex, tex_coords, dx, dy))); } + +vec4 tex2D_linearize(const sampler2D tex, const vec2 tex_coords, const vec2 dx, const vec2 dy, const int texel_off) +{ return decode_input(vec4(texture(tex, tex_coords, dx, dy, texel_off))); } + +vec4 tex2D_linearize(const sampler2D tex, const vec3 tex_coords, const vec2 dx, const vec2 dy, const int texel_off) +{ return decode_input(vec4(texture(tex, tex_coords, dx, dy, texel_off))); } + +// tex2Dbias: +vec4 tex2Dbias_linearize(const sampler2D tex, const vec4 tex_coords) +{ return decode_input(vec4(tex2Dbias(tex, tex_coords))); } + +vec4 tex2Dbias_linearize(const sampler2D tex, const vec4 tex_coords, const int texel_off) +{ return decode_input(vec4(tex2Dbias(tex, tex_coords, texel_off))); } + +// tex2Dfetch: +vec4 tex2Dfetch_linearize(const sampler2D tex, const ivec4 tex_coords) +{ return decode_input(vec4(texture2Dfetch(tex, tex_coords))); } + +vec4 tex2Dfetch_linearize(const sampler2D tex, const ivec4 tex_coords, const int texel_off) +{ return decode_input(vec4(texture2Dfetch(tex, tex_coords, texel_off))); } + +// tex2Dlod: +vec4 tex2Dlod_linearize(const sampler2D tex, const vec4 tex_coords) +{ return decode_input(vec4(texture2Dlod(tex, tex_coords))); } + +vec4 tex2Dlod_linearize(const sampler2D tex, const vec4 tex_coords, const int texel_off) +{ return decode_input(vec4(texture2Dlod(tex, tex_coords, texel_off))); } + +// tex2Dproj: +vec4 tex2Dproj_linearize(const sampler2D tex, const vec3 tex_coords) +{ return decode_input(vec4(tex2Dproj(tex, tex_coords))); } + +vec4 tex2Dproj_linearize(const sampler2D tex, const vec4 tex_coords) +{ return decode_input(vec4(tex2Dproj(tex, tex_coords))); } + +vec4 tex2Dproj_linearize(const sampler2D tex, const vec3 tex_coords, const int texel_off) +{ return decode_input(vec4(tex2Dproj(tex, tex_coords, texel_off))); } + +vec4 tex2Dproj_linearize(const sampler2D tex, const vec4 tex_coords, const int texel_off) +{ return decode_input(vec4(tex2Dproj(tex, tex_coords, texel_off))); } + +/* +// tex3D: +vec4 tex3D_linearize(const sampler3D texture, const vec3 tex_coords) +{ return decode_input(tex3D(texture, tex_coords)); } + +vec4 tex3D_linearize(const sampler3D texture, const vec3 tex_coords, const int texel_off) +{ return decode_input(tex3D(texture, tex_coords, texel_off)); } + +vec4 tex3D_linearize(const sampler3D texture, const vec3 tex_coords, const vec3 dx, const vec3 dy) +{ return decode_input(tex3D(texture, tex_coords, dx, dy)); } + +vec4 tex3D_linearize(const sampler3D texture, const vec3 tex_coords, const vec3 dx, const vec3 dy, const int texel_off) +{ return decode_input(tex3D(texture, tex_coords, dx, dy, texel_off)); } + +// tex3Dbias: +vec4 tex3Dbias_linearize(const sampler3D texture, const vec4 tex_coords) +{ return decode_input(tex3Dbias(texture, tex_coords)); } + +vec4 tex3Dbias_linearize(const sampler3D texture, const vec4 tex_coords, const int texel_off) +{ return decode_input(tex3Dbias(texture, tex_coords, texel_off)); } + +// tex3Dfetch: +vec4 tex3Dfetch_linearize(const sampler3D texture, const int4 tex_coords) +{ return decode_input(tex3Dfetch(texture, tex_coords)); } + +vec4 tex3Dfetch_linearize(const sampler3D texture, const int4 tex_coords, const int texel_off) +{ return decode_input(tex3Dfetch(texture, tex_coords, texel_off)); } + +// tex3Dlod: +vec4 tex3Dlod_linearize(const sampler3D texture, const vec4 tex_coords) +{ return decode_input(tex3Dlod(texture, tex_coords)); } + +vec4 tex3Dlod_linearize(const sampler3D texture, const vec4 tex_coords, const int texel_off) +{ return decode_input(tex3Dlod(texture, tex_coords, texel_off)); } + +// tex3Dproj: +vec4 tex3Dproj_linearize(const sampler3D texture, const vec4 tex_coords) +{ return decode_input(tex3Dproj(texture, tex_coords)); } + +vec4 tex3Dproj_linearize(const sampler3D texture, const vec4 tex_coords, const int texel_off) +{ return decode_input(tex3Dproj(texture, tex_coords, texel_off)); } +*/ + + +// NONSTANDARD "SMART" LINEARIZING TEXTURE LOOKUP FUNCTIONS: +// This narrow selection of nonstandard tex2D* functions can be useful: + +// tex2Dlod0: Automatically fill in the tex2D LOD parameter for mip level 0. +vec4 tex2Dlod0_linearize(const sampler2D texture, const vec2 tex_coords) +{ return decode_input(vec4(texture2Dlod(texture, vec4(tex_coords, 0.0, 0.0)))); } + +vec4 tex2Dlod0_linearize(const sampler2D texture, const vec2 tex_coords, const int texel_off) +{ return decode_input(vec4(texture2Dlod(texture, vec4(tex_coords, 0.0, 0.0), texel_off))); } + + +// MANUALLY LINEARIZING TEXTURE LOOKUP FUNCTIONS: +// Provide a narrower selection of tex2D* wrapper functions that decode an +// input sample with a specified gamma value. These are useful for reading +// LUT's and for reading the input of pass0 in a later pass. + +// tex2D: +vec4 tex2D_linearize_gamma(const sampler2D tex, const vec2 tex_coords, const vec3 gamma) +{ return decode_gamma_input(vec4(texture(tex, tex_coords), vec3(gamma))); } + +vec4 tex2D_linearize_gamma(const sampler2D tex, const vec3 tex_coords, const vec3 gamma) +{ return decode_gamma_input(vec4(texture(tex, tex_coords), vec3(gamma))); } + +vec4 tex2D_linearize_gamma(const sampler2D tex, const vec2 tex_coords, const int texel_off, const vec3 gamma) +{ return decode_gamma_input(vec4(texture(tex, tex_coords, texel_off), vec3(gamma))); } + +vec4 tex2D_linearize_gamma(const sampler2D tex, const vec3 tex_coords, const int texel_off, const vec3 gamma) +{ return decode_gamma_input(vec4(texture(tex, tex_coords, texel_off), vec3(gamma))); } + +vec4 tex2D_linearize_gamma(const sampler2D tex, const vec2 tex_coords, const vec2 dx, const vec2 dy, const vec3 gamma) +{ return decode_gamma_input(vec4(texture(tex, tex_coords, dx, dy), vec3(gamma))); } + +vec4 tex2D_linearize_gamma(const sampler2D tex, const vec3 tex_coords, const vec2 dx, const vec2 dy, const vec3 gamma) +{ return decode_gamma_input(vec4(texture(tex, tex_coords, dx, dy), vec3(gamma))); } + +vec4 tex2D_linearize_gamma(const sampler2D tex, const vec2 tex_coords, const vec2 dx, const vec2 dy, const int texel_off, const vec3 gamma) +{ return decode_gamma_input(vec4(texture(tex, tex_coords, dx, dy, texel_off), vec3(gamma))); } + +vec4 tex2D_linearize_gamma(const sampler2D tex, const vec3 tex_coords, const vec2 dx, const vec2 dy, const int texel_off, const vec3 gamma) +{ return decode_gamma_input(vec4(texture(tex, tex_coords, dx, dy, texel_off), vec3(gamma))); } + +// tex2Dbias: +vec4 tex2Dbias_linearize_gamma(const sampler2D tex, const vec4 tex_coords, const vec3 gamma) +{ return decode_gamma_input(vec4(tex2Dbias(tex, tex_coords), vec3(gamma))); } + +vec4 tex2Dbias_linearize_gamma(const sampler2D tex, const vec4 tex_coords, const int texel_off, const vec3 gamma) +{ return decode_gamma_input(vec4(tex2Dbias(tex, tex_coords, texel_off), vec3(gamma))); } + +// tex2Dfetch: +vec4 tex2Dfetch_linearize_gamma(const sampler2D tex, const int4 tex_coords, const vec3 gamma) +{ return decode_gamma_input(vec4(tex2Dfetch(tex, tex_coords), vec3(gamma))); } + +vec4 tex2Dfetch_linearize_gamma(const sampler2D tex, const int4 tex_coords, const int texel_off, const vec3 gamma) +{ return decode_gamma_input(vec4(tex2Dfetch(tex, tex_coords, texel_off), vec3(gamma))); } + +// tex2Dlod: +vec4 tex2Dlod_linearize_gamma(const sampler2D tex, const vec4 tex_coords, const vec3 gamma) +{ return decode_gamma_input(vec4(tex2Dlod(tex, tex_coords), vec3(gamma))); } + +vec4 tex2Dlod_linearize_gamma(const sampler2D tex, const vec4 tex_coords, const int texel_off, const vec3 gamma) +{ return decode_gamma_input(vec4(tex2Dlod(tex, tex_coords, texel_off), vec3(gamma))); } + + +#endif // GAMMA_MANAGEMENT_H + diff --git a/include/gamma-management.h b/include/gamma-management.h new file mode 100644 index 0000000..4236bb3 --- /dev/null +++ b/include/gamma-management.h @@ -0,0 +1,160 @@ +/////////////////////////////// BASE CONSTANTS /////////////////////////////// + +// Set standard gamma constants, but allow users to override them: +#ifndef OVERRIDE_STANDARD_GAMMA + // Standard encoding gammas: + const float ntsc_gamma = 2.2; // Best to use NTSC for PAL too? + const float pal_gamma = 2.8; // Never actually 2.8 in practice + // Typical device decoding gammas (only use for emulating devices): + // CRT/LCD reference gammas are higher than NTSC and Rec.709 video standard + // gammas: The standards purposely undercorrected for an analog CRT's + // assumed 2.5 reference display gamma to maintain contrast in assumed + // [dark] viewing conditions: http://www.poynton.com/PDFs/GammaFAQ.pdf + // These unstated assumptions about display gamma and perceptual rendering + // intent caused a lot of confusion, and more modern CRT's seemed to target + // NTSC 2.2 gamma with circuitry. LCD displays seem to have followed suit + // (they struggle near black with 2.5 gamma anyway), especially PC/laptop + // displays designed to view sRGB in bright environments. (Standards are + // also in flux again with BT.1886, but it's underspecified for displays.) + const float crt_reference_gamma_high = 2.5; // In (2.35, 2.55) + const float crt_reference_gamma_low = 2.35; // In (2.35, 2.55) + const float lcd_reference_gamma = 2.5; // To match CRT + const float crt_office_gamma = 2.2; // Circuitry-adjusted for NTSC + const float lcd_office_gamma = 2.2; // Approximates sRGB +#endif // OVERRIDE_STANDARD_GAMMA + +// Assuming alpha == 1.0 might make it easier for users to avoid some bugs, +// but only if they're aware of it. +#ifndef OVERRIDE_ALPHA_ASSUMPTIONS + const bool assume_opaque_alpha = false; +#endif + + +/////////////////////// DERIVED CONSTANTS AS FUNCTIONS /////////////////////// + +// gamma-management.h should be compatible with overriding gamma values with +// runtime user parameters, but we can only define other global constants in +// terms of static constants, not uniform user parameters. To get around this +// limitation, we need to define derived constants using functions. + +// Set device gamma constants, but allow users to override them: +#ifdef OVERRIDE_DEVICE_GAMMA + // The user promises to globally define the appropriate constants: + float get_crt_gamma() { return crt_gamma; } + float get_gba_gamma() { return gba_gamma; } + float get_lcd_gamma() { return lcd_gamma; } +#else + float get_crt_gamma() { return crt_reference_gamma_high; } + float get_gba_gamma() { return 3.5; } // Game Boy Advance; in (3.0, 4.0) + float get_lcd_gamma() { return lcd_office_gamma; } +#endif // OVERRIDE_DEVICE_GAMMA + +// Set decoding/encoding gammas for the first/lass passes, but allow overrides: +#ifdef OVERRIDE_FINAL_GAMMA + // The user promises to globally define the appropriate constants: + float get_intermediate_gamma() { return intermediate_gamma; } + float get_input_gamma() { return input_gamma; } + float get_output_gamma() { return output_gamma; } +#else + // If we gamma-correct every pass, always use ntsc_gamma between passes to + // ensure middle passes don't need to care if anything is being simulated: + float get_intermediate_gamma() { return ntsc_gamma; } + #ifdef SIMULATE_CRT_ON_LCD + float get_input_gamma() { return get_crt_gamma(); } + float get_output_gamma() { return get_lcd_gamma(); } + #else + #ifdef SIMULATE_GBA_ON_LCD + float get_input_gamma() { return get_gba_gamma(); } + float get_output_gamma() { return get_lcd_gamma(); } + #else + #ifdef SIMULATE_LCD_ON_CRT + float get_input_gamma() { return get_lcd_gamma(); } + float get_output_gamma() { return get_crt_gamma(); } + #else + #ifdef SIMULATE_GBA_ON_CRT + float get_input_gamma() { return get_gba_gamma(); } + float get_output_gamma() { return get_crt_gamma(); } + #else // Don't simulate anything: + float get_input_gamma() { return ntsc_gamma; } + float get_output_gamma() { return ntsc_gamma; } + #endif // SIMULATE_GBA_ON_CRT + #endif // SIMULATE_LCD_ON_CRT + #endif // SIMULATE_GBA_ON_LCD + #endif // SIMULATE_CRT_ON_LCD +#endif // OVERRIDE_FINAL_GAMMA + +#ifndef GAMMA_ENCODE_EVERY_FBO + #ifdef FIRST_PASS + const bool linearize_input = true; + float get_pass_input_gamma() { return get_input_gamma(); } + #else + const bool linearize_input = false; + float get_pass_input_gamma() { return 1.0; } + #endif + #ifdef LAST_PASS + const bool gamma_encode_output = true; + float get_pass_output_gamma() { return get_output_gamma(); } + #else + const bool gamma_encode_output = false; + float get_pass_output_gamma() { return 1.0; } + #endif +#else + const bool linearize_input = true; + const bool gamma_encode_output = true; + #ifdef FIRST_PASS + float get_pass_input_gamma() { return get_input_gamma(); } + #else + float get_pass_input_gamma() { return get_intermediate_gamma(); } + #endif + #ifdef LAST_PASS + float get_pass_output_gamma() { return get_output_gamma(); } + #else + float get_pass_output_gamma() { return get_intermediate_gamma(); } + #endif +#endif + +vec4 decode_input(const vec4 color) +{ + if(linearize_input) + { + if(assume_opaque_alpha) + { + return vec4(pow(color.rgb, vec3(get_pass_input_gamma())), 1.0); + } + else + { + return vec4(pow(color.rgb, vec3(get_pass_input_gamma())), color.a); + } + } + else + { + return color; + } +} + +vec4 encode_output(const vec4 color) +{ + if(gamma_encode_output) + { + if(assume_opaque_alpha) + { + return vec4(pow(color.rgb, vec3(1.0/get_pass_output_gamma())), 1.0); + } + else + { + return vec4(pow(color.rgb, vec3(1.0/get_pass_output_gamma())), color.a); + } + } + else + { + return color; + } +} + +#define tex2D_linearize(C, D) decode_input(vec4(texture(C, D))) +//vec4 tex2D_linearize(const sampler2D tex, const vec2 tex_coords) +//{ return decode_input(vec4(texture(tex, tex_coords))); } + +//#define tex2D_linearize(C, D, E) decode_input(vec4(texture(C, D, E))) +//vec4 tex2D_linearize(const sampler2D tex, const vec2 tex_coords, const int texel_off) +//{ return decode_input(vec4(texture(tex, tex_coords, texel_off))); } \ No newline at end of file diff --git a/include/quad-pixel-communication.h b/include/quad-pixel-communication.h new file mode 100644 index 0000000..4c3f1cb --- /dev/null +++ b/include/quad-pixel-communication.h @@ -0,0 +1,243 @@ +#ifndef QUAD_PIXEL_COMMUNICATION_H +#define QUAD_PIXEL_COMMUNICATION_H + +///////////////////////////////// MIT LICENSE //////////////////////////////// + +// Copyright (C) 2014 TroggleMonkey* +// +// 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. + +///////////////////////////////// DISCLAIMER ///////////////////////////////// + +// *This code was inspired by "Shader Amortization using Pixel Quad Message +// Passing" by Eric Penner, published in GPU Pro 2, Chapter VI.2. My intent +// is not to plagiarize his fundamentally similar code and assert my own +// copyright, but the algorithmic helper functions require so little code that +// implementations can't vary by much except bugfixes and conventions. I just +// wanted to license my own particular code here to avoid ambiguity and make it +// clear that as far as I'm concerned, people can do as they please with it. + +///////////////////////////////// DESCRIPTION //////////////////////////////// + +// Given screen pixel numbers, derive a "quad vector" describing a fragment's +// position in its 2x2 pixel quad. Given that vector, obtain the values of any +// variable at neighboring fragments. +// Requires: Using this file in general requires: +// 1.) ddx() and ddy() are present in the current Cg profile. +// 2.) The GPU driver is using fine/high-quality derivatives. +// Functions will give incorrect results if this is not true, +// so a test function is included. + + +///////////////////// QUAD-PIXEL COMMUNICATION PRIMITIVES //////////////////// + +vec4 get_quad_vector_naive(const vec4 output_pixel_num_wrt_uvxy) +{ + // Requires: Two measures of the current fragment's output pixel number + // in the range ([0, IN.output_size.x), [0, IN.output_size.y)): + // 1.) output_pixel_num_wrt_uvxy.xy increase with uv coords. + // 2.) output_pixel_num_wrt_uvxy.zw increase with screen xy. + // Returns: Two measures of the fragment's position in its 2x2 quad: + // 1.) The .xy components are its 2x2 placement with respect to + // uv direction (the origin (0, 0) is at the top-left): + // top-left = (-1.0, -1.0) top-right = ( 1.0, -1.0) + // bottom-left = (-1.0, 1.0) bottom-right = ( 1.0, 1.0) + // You need this to arrange/weight shared texture samples. + // 2.) The .zw components are its 2x2 placement with respect to + // screen xy direction (IN.position); the origin varies. + // quad_gather needs this measure to work correctly. + // Note: quad_vector.zw = quad_vector.xy * vec2( + // ddx(output_pixel_num_wrt_uvxy.x), + // ddy(output_pixel_num_wrt_uvxy.y)); + // Caveats: This function assumes the GPU driver always starts 2x2 pixel + // quads at even pixel numbers. This assumption can be wrong + // for odd output resolutions (nondeterministically so). + const vec4 pixel_odd = frac(output_pixel_num_wrt_uvxy * 0.5) * 2.0; + const vec4 quad_vector = pixel_odd * 2.0 - vec4(1.0); + return quad_vector; +} + +vec4 get_quad_vector(const vec4 output_pixel_num_wrt_uvxy) +{ + // Requires: Same as get_quad_vector_naive() (see that first). + // Returns: Same as get_quad_vector_naive() (see that first), but it's + // correct even if the 2x2 pixel quad starts at an odd pixel, + // which can occur at odd resolutions. + const vec4 quad_vector_guess = + get_quad_vector_naive(output_pixel_num_wrt_uvxy); + // If quad_vector_guess.zw doesn't increase with screen xy, we know + // the 2x2 pixel quad starts at an odd pixel: + const vec2 odd_start_mirror = 0.5 * vec2(ddx(quad_vector_guess.z), + ddy(quad_vector_guess.w)); + return quad_vector_guess * odd_start_mirror.xyxy; +} + +vec4 get_quad_vector(const vec2 output_pixel_num_wrt_uv) +{ + // Requires: 1.) ddx() and ddy() are present in the current Cg profile. + // 2.) output_pixel_num_wrt_uv must increase with uv coords and + // measure the current fragment's output pixel number in: + // ([0, IN.output_size.x), [0, IN.output_size.y)) + // Returns: Same as get_quad_vector_naive() (see that first), but it's + // correct even if the 2x2 pixel quad starts at an odd pixel, + // which can occur at odd resolutions. + // Caveats: This function requires less information than the version + // taking a vec4, but it's potentially slower. + // Do screen coords increase with or against uv? Get the direction + // with respect to (uv.x, uv.y) for (screen.x, screen.y) in {-1, 1}. + const vec2 screen_uv_mirror = vec2(ddx(output_pixel_num_wrt_uv.x), + ddy(output_pixel_num_wrt_uv.y)); + const vec2 pixel_odd_wrt_uv = frac(output_pixel_num_wrt_uv * 0.5) * 2.0; + const vec2 quad_vector_uv_guess = (pixel_odd_wrt_uv - vec2(0.5)) * 2.0; + const vec2 quad_vector_screen_guess = quad_vector_uv_guess * screen_uv_mirror; + // If quad_vector_screen_guess doesn't increase with screen xy, we know + // the 2x2 pixel quad starts at an odd pixel: + const vec2 odd_start_mirror = 0.5 * vec2(ddx(quad_vector_screen_guess.x), + ddy(quad_vector_screen_guess.y)); + const vec4 quad_vector_guess = vec4( + quad_vector_uv_guess, quad_vector_screen_guess); + return quad_vector_guess * odd_start_mirror.xyxy; +} + +void quad_gather(const vec4 quad_vector, const vec4 curr, + out vec4 adjx, out vec4 adjy, out vec4 diag) +{ + // Requires: 1.) ddx() and ddy() are present in the current Cg profile. + // 2.) The GPU driver is using fine/high-quality derivatives. + // 3.) quad_vector describes the current fragment's location in + // its 2x2 pixel quad using get_quad_vector()'s conventions. + // 4.) curr is any vector you wish to get neighboring values of. + // Returns: Values of an input vector (curr) at neighboring fragments + // adjacent x, adjacent y, and diagonal (via out parameters). + adjx = curr - ddx(curr) * quad_vector.z; + adjy = curr - ddy(curr) * quad_vector.w; + diag = adjx - ddy(adjx) * quad_vector.w; +} + +void quad_gather(const vec4 quad_vector, const vec3 curr, + out vec3 adjx, out vec3 adjy, out vec3 diag) +{ + // vec3 version + adjx = curr - ddx(curr) * quad_vector.z; + adjy = curr - ddy(curr) * quad_vector.w; + diag = adjx - ddy(adjx) * quad_vector.w; +} + +void quad_gather(const vec4 quad_vector, const vec2 curr, + out vec2 adjx, out vec2 adjy, out vec2 diag) +{ + // vec2 version + adjx = curr - ddx(curr) * quad_vector.z; + adjy = curr - ddy(curr) * quad_vector.w; + diag = adjx - ddy(adjx) * quad_vector.w; +} + +vec4 quad_gather(const vec4 quad_vector, const float curr) +{ + // Float version: + // Returns: return.x == current + // return.y == adjacent x + // return.z == adjacent y + // return.w == diagonal + vec4 all = vec4(curr); + all.y = all.x - ddx(all.x) * quad_vector.z; + all.zw = all.xy - ddy(all.xy) * quad_vector.w; + return all; +} + +vec4 quad_gather_sum(const vec4 quad_vector, const vec4 curr) +{ + // Requires: Same as quad_gather() + // Returns: Sum of an input vector (curr) at all fragments in a quad. + vec4 adjx, adjy, diag; + quad_gather(quad_vector, curr, adjx, adjy, diag); + return (curr + adjx + adjy + diag); +} + +vec3 quad_gather_sum(const vec4 quad_vector, const vec3 curr) +{ + // vec3 version: + vec3 adjx, adjy, diag; + quad_gather(quad_vector, curr, adjx, adjy, diag); + return (curr + adjx + adjy + diag); +} + +vec2 quad_gather_sum(const vec4 quad_vector, const vec2 curr) +{ + // vec2 version: + vec2 adjx, adjy, diag; + quad_gather(quad_vector, curr, adjx, adjy, diag); + return (curr + adjx + adjy + diag); +} + +float quad_gather_sum(const vec4 quad_vector, const float curr) +{ + // Float version: + const vec4 all_values = quad_gather(quad_vector, curr); + return (all_values.x + all_values.y + all_values.z + all_values.w); +} + +bool fine_derivatives_working(const vec4 quad_vector, vec4 curr) +{ + // Requires: 1.) ddx() and ddy() are present in the current Cg profile. + // 2.) quad_vector describes the current fragment's location in + // its 2x2 pixel quad using get_quad_vector()'s conventions. + // 3.) curr must be a test vector with non-constant derivatives + // (its value should change nonlinearly across fragments). + // Returns: true if fine/hybrid/high-quality derivatives are used, or + // false if coarse derivatives are used or inconclusive + // Usage: Test whether quad-pixel communication is working! + // Method: We can confirm fine derivatives are used if the following + // holds (ever, for any value at any fragment): + // (ddy(curr) != ddy(adjx)) or (ddx(curr) != ddx(adjy)) + // The more values we test (e.g. test a vec4 two ways), the + // easier it is to demonstrate fine derivatives are working. + // TODO: Check for floating point exact comparison issues! + vec4 ddx_curr = ddx(curr); + vec4 ddy_curr = ddy(curr); + vec4 adjx = curr - ddx_curr * quad_vector.z; + vec4 adjy = curr - ddy_curr * quad_vector.w; + bool ddy_different = any(ddy_curr != ddy(adjx)); + bool ddx_different = any(ddx_curr != ddx(adjy)); + return any(bool2(ddy_different, ddx_different)); +} + +bool fine_derivatives_working_fast(const vec4 quad_vector, float curr) +{ + // Requires: Same as fine_derivatives_working() + // Returns: Same as fine_derivatives_working() + // Usage: This is faster than fine_derivatives_working() but more + // likely to return false negatives, so it's less useful for + // offline testing/debugging. It's also useless as the basis + // for dynamic runtime branching as of May 2014: Derivatives + // (and quad-pixel communication) are currently disallowed in + // branches. However, future GPU's may allow you to use them + // in dynamic branches if you promise the branch condition + // evaluates the same for every fragment in the quad (and/or if + // the driver enforces that promise by making a single fragment + // control branch decisions). If that ever happens, this + // version may become a more economical choice. + float ddx_curr = ddx(curr); + float ddy_curr = ddy(curr); + float adjx = curr - ddx_curr * quad_vector.z; + return (ddy_curr != ddy(adjx)); +} + +#endif // QUAD_PIXEL_COMMUNICATION_H + diff --git a/include/special-functions-old.h b/include/special-functions-old.h new file mode 100644 index 0000000..839267a --- /dev/null +++ b/include/special-functions-old.h @@ -0,0 +1,498 @@ +#ifndef SPECIAL_FUNCTIONS_H +#define SPECIAL_FUNCTIONS_H + +///////////////////////////////// MIT LICENSE //////////////////////////////// + +// Copyright (C) 2014 TroggleMonkey +// +// 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. + + +///////////////////////////////// DESCRIPTION //////////////////////////////// + +// This file implements the following mathematical special functions: +// 1.) erf() = 2/sqrt(pi) * indefinite_integral(e**(-x**2)) +// 2.) gamma(s), a real-numbered extension of the integer factorial function +// It also implements normalized_ligamma(s, z), a normalized lower incomplete +// gamma function for s < 0.5 only. Both gamma() and normalized_ligamma() can +// be called with an _impl suffix to use an implementation version with a few +// extra precomputed parameters (which may be useful for the caller to reuse). +// See below for details. +// +// Design Rationale: +// Pretty much every line of code in this file is duplicated four times for +// different input types (vec4/vec3/vec2/float). This is unfortunate, +// but Cg doesn't allow function templates. Macros would be far less verbose, +// but they would make the code harder to document and read. I don't expect +// these functions will require a whole lot of maintenance changes unless +// someone ever has need for more robust incomplete gamma functions, so code +// duplication seems to be the lesser evil in this case. + + +/////////////////////////// GAUSSIAN ERROR FUNCTION ////////////////////////// + +vec4 erf6(vec4 x) +{ + // Requires: x is the standard parameter to erf(). + // Returns: Return an Abramowitz/Stegun approximation of erf(), where: + // erf(x) = 2/sqrt(pi) * integral(e**(-x**2)) + // This approximation has a max absolute error of 2.5*10**-5 + // with solid numerical robustness and efficiency. See: + // https://en.wikipedia.org/wiki/Error_function#Approximation_with_elementary_functions + const vec4 one = vec4(1.0); + const vec4 sign_x = sign(x); + const vec4 t = one/(one + 0.47047*abs(x)); + const vec4 result = one - t*(0.3480242 + t*(-0.0958798 + t*0.7478556))* + exp(-(x*x)); + return result * sign_x; +} + +vec3 erf6(const vec3 x) +{ + // vec3 version: + const vec3 one = vec3(1.0); + const vec3 sign_x = sign(x); + const vec3 t = one/(one + 0.47047*abs(x)); + const vec3 result = one - t*(0.3480242 + t*(-0.0958798 + t*0.7478556))* + exp(-(x*x)); + return result * sign_x; +} + +vec2 erf6(const vec2 x) +{ + // vec2 version: + const vec2 one = vec2(1.0); + const vec2 sign_x = sign(x); + const vec2 t = one/(one + 0.47047*abs(x)); + const vec2 result = one - t*(0.3480242 + t*(-0.0958798 + t*0.7478556))* + exp(-(x*x)); + return result * sign_x; +} + +float erf6(const float x) +{ + // Float version: + const float sign_x = sign(x); + const float t = 1.0/(1.0 + 0.47047*abs(x)); + const float result = 1.0 - t*(0.3480242 + t*(-0.0958798 + t*0.7478556))* + exp(-(x*x)); + return result * sign_x; +} + +vec4 erft(const vec4 x) +{ + // Requires: x is the standard parameter to erf(). + // Returns: Approximate erf() with the hyperbolic tangent. The error is + // visually noticeable, but it's blazing fast and perceptually + // close...at least on ATI hardware. See: + // http://www.maplesoft.com/applications/view.aspx?SID=5525&view=html + // Warning: Only use this if your hardware drivers correctly implement + // tanh(): My nVidia 8800GTS returns garbage output. + return tanh(1.202760580 * x); +} + +vec3 erft(const vec3 x) +{ + // vec3 version: + return tanh(1.202760580 * x); +} + +vec2 erft(const vec2 x) +{ + // vec2 version: + return tanh(1.202760580 * x); +} + +float erft(const float x) +{ + // Float version: + return tanh(1.202760580 * x); +} + +vec4 erf(const vec4 x) +{ + // Requires: x is the standard parameter to erf(). + // Returns: Some approximation of erf(x), depending on user settings. + #ifdef ERF_FAST_APPROXIMATION + return erft(x); + #else + return erf6(x); + #endif +} + +vec3 erf(const vec3 x) +{ + // vec3 version: + #ifdef ERF_FAST_APPROXIMATION + return erft(x); + #else + return erf6(x); + #endif +} + +vec2 erf(const vec2 x) +{ + // vec2 version: + #ifdef ERF_FAST_APPROXIMATION + return erft(x); + #else + return erf6(x); + #endif +} + +float erf(const float x) +{ + // Float version: + #ifdef ERF_FAST_APPROXIMATION + return erft(x); + #else + return erf6(x); + #endif +} + + +/////////////////////////// COMPLETE GAMMA FUNCTION ////////////////////////// + +vec4 gamma_impl(const vec4 s, const vec4 s_inv) +{ + // Requires: 1.) s is the standard parameter to the gamma function, and + // it should lie in the [0, 36] range. + // 2.) s_inv = 1.0/s. This implementation function requires + // the caller to precompute this value, giving users the + // opportunity to reuse it. + // Returns: Return approximate gamma function (real-numbered factorial) + // output using the Lanczos approximation with two coefficients + // calculated using Paul Godfrey's method here: + // http://my.fit.edu/~gabdo/gamma.txt + // An optimal g value for s in [0, 36] is ~1.12906830989, with + // a maximum relative error of 0.000463 for 2**16 equally + // evals. We could use three coeffs (0.0000346 error) without + // hurting latency, but this allows more parallelism with + // outside instructions. + const vec4 g = vec4(1.12906830989); + const vec4 c0 = vec4(0.8109119309638332633713423362694399653724431); + const vec4 c1 = vec4(0.4808354605142681877121661197951496120000040); + const vec4 e = vec4(2.71828182845904523536028747135266249775724709); + const vec4 sph = s + vec4(0.5); + const vec4 lanczos_sum = c0 + c1/(s + vec4(1.0)); + const vec4 base = (sph + g)/e; // or (s + g + vec4(0.5))/e + // gamma(s + 1) = base**sph * lanczos_sum; divide by s for gamma(s). + // This has less error for small s's than (s -= 1.0) at the beginning. + return (pow(base, sph) * lanczos_sum) * s_inv; +} + +vec3 gamma_impl(const vec3 s, const vec3 s_inv) +{ + // vec3 version: + const vec3 g = vec3(1.12906830989); + const vec3 c0 = vec3(0.8109119309638332633713423362694399653724431); + const vec3 c1 = vec3(0.4808354605142681877121661197951496120000040); + const vec3 e = vec3(2.71828182845904523536028747135266249775724709); + const vec3 sph = s + vec3(0.5); + const vec3 lanczos_sum = c0 + c1/(s + vec3(1.0)); + const vec3 base = (sph + g)/e; + return (pow(base, sph) * lanczos_sum) * s_inv; +} + +vec2 gamma_impl(const vec2 s, const vec2 s_inv) +{ + // vec2 version: + const vec2 g = vec2(1.12906830989); + const vec2 c0 = vec2(0.8109119309638332633713423362694399653724431); + const vec2 c1 = vec2(0.4808354605142681877121661197951496120000040); + const vec2 e = vec2(2.71828182845904523536028747135266249775724709); + const vec2 sph = s + vec2(0.5); + const vec2 lanczos_sum = c0 + c1/(s + vec2(1.0)); + const vec2 base = (sph + g)/e; + return (pow(base, sph) * lanczos_sum) * s_inv; +} + +float gamma_impl(const float s, const float s_inv) +{ + // Float version: + const float g = 1.12906830989; + const float c0 = 0.8109119309638332633713423362694399653724431; + const float c1 = 0.4808354605142681877121661197951496120000040; + const float e = 2.71828182845904523536028747135266249775724709; + const float sph = s + 0.5; + const float lanczos_sum = c0 + c1/(s + 1.0); + const float base = (sph + g)/e; + return (pow(base, sph) * lanczos_sum) * s_inv; +} + +vec4 gamma(const vec4 s) +{ + // Requires: s is the standard parameter to the gamma function, and it + // should lie in the [0, 36] range. + // Returns: Return approximate gamma function output with a maximum + // relative error of 0.000463. See gamma_impl for details. + return gamma_impl(s, vec4(1.0)/s); +} + +vec3 gamma(const vec3 s) +{ + // vec3 version: + return gamma_impl(s, vec3(1.0)/s); +} + +vec2 gamma(const vec2 s) +{ + // vec2 version: + return gamma_impl(s, vec2(1.0)/s); +} + +float gamma(const float s) +{ + // Float version: + return gamma_impl(s, 1.0/s); +} + + +//////////////// INCOMPLETE GAMMA FUNCTIONS (RESTRICTED INPUT) /////////////// + +// Lower incomplete gamma function for small s and z (implementation): +vec4 ligamma_small_z_impl(const vec4 s, const vec4 z, const vec4 s_inv) +{ + // Requires: 1.) s < ~0.5 + // 2.) z <= ~0.775075 + // 3.) s_inv = 1.0/s (precomputed for outside reuse) + // Returns: A series representation for the lower incomplete gamma + // function for small s and small z (4 terms). + // The actual "rolled up" summation looks like: + // last_sign = 1.0; last_pow = 1.0; last_factorial = 1.0; + // sum = last_sign * last_pow / ((s + k) * last_factorial) + // for(int i = 0; i < 4; ++i) + // { + // last_sign *= -1.0; last_pow *= z; last_factorial *= i; + // sum += last_sign * last_pow / ((s + k) * last_factorial); + // } + // Unrolled, constant-unfolded and arranged for madds and parallelism: + const vec4 scale = pow(z, s); + vec4 sum = s_inv; // Summation iteration 0 result + // Summation iterations 1, 2, and 3: + const vec4 z_sq = z*z; + const vec4 denom1 = s + vec4(1.0); + const vec4 denom2 = 2.0*s + vec4(4.0); + const vec4 denom3 = 6.0*s + vec4(18.0); + //vec4 denom4 = 24.0*s + vec4(96.0); + sum -= z/denom1; + sum += z_sq/denom2; + sum -= z * z_sq/denom3; + //sum += z_sq * z_sq / denom4; + // Scale and return: + return scale * sum; +} + +vec3 ligamma_small_z_impl(const vec3 s, const vec3 z, const vec3 s_inv) +{ + // vec3 version: + const vec3 scale = pow(z, s); + vec3 sum = s_inv; + const vec3 z_sq = z*z; + const vec3 denom1 = s + vec3(1.0); + const vec3 denom2 = 2.0*s + vec3(4.0); + const vec3 denom3 = 6.0*s + vec3(18.0); + sum -= z/denom1; + sum += z_sq/denom2; + sum -= z * z_sq/denom3; + return scale * sum; +} + +vec2 ligamma_small_z_impl(const vec2 s, const vec2 z, const vec2 s_inv) +{ + // vec2 version: + const vec2 scale = pow(z, s); + vec2 sum = s_inv; + const vec2 z_sq = z*z; + const vec2 denom1 = s + vec2(1.0); + const vec2 denom2 = 2.0*s + vec2(4.0); + const vec2 denom3 = 6.0*s + vec2(18.0); + sum -= z/denom1; + sum += z_sq/denom2; + sum -= z * z_sq/denom3; + return scale * sum; +} + +float ligamma_small_z_impl(const float s, const float z, const float s_inv) +{ + // Float version: + const float scale = pow(z, s); + float sum = s_inv; + const float z_sq = z*z; + const float denom1 = s + 1.0; + const float denom2 = 2.0*s + 4.0; + const float denom3 = 6.0*s + 18.0; + sum -= z/denom1; + sum += z_sq/denom2; + sum -= z * z_sq/denom3; + return scale * sum; +} + +// Upper incomplete gamma function for small s and large z (implementation): +vec4 uigamma_large_z_impl(const vec4 s, const vec4 z) +{ + // Requires: 1.) s < ~0.5 + // 2.) z > ~0.775075 + // Returns: Gauss's continued fraction representation for the upper + // incomplete gamma function (4 terms). + // The "rolled up" continued fraction looks like this. The denominator + // is truncated, and it's calculated "from the bottom up:" + // denom = vec4('inf'); + // vec4 one = vec4(1.0); + // for(int i = 4; i > 0; --i) + // { + // denom = ((i * 2.0) - one) + z - s + (i * (s - i))/denom; + // } + // Unrolled and constant-unfolded for madds and parallelism: + const vec4 numerator = pow(z, s) * exp(-z); + vec4 denom = vec4(7.0) + z - s; + denom = vec4(5.0) + z - s + (3.0*s - vec4(9.0))/denom; + denom = vec4(3.0) + z - s + (2.0*s - vec4(4.0))/denom; + denom = vec4(1.0) + z - s + (s - vec4(1.0))/denom; + return numerator / denom; +} + +vec3 uigamma_large_z_impl(const vec3 s, const vec3 z) +{ + // vec3 version: + const vec3 numerator = pow(z, s) * exp(-z); + vec3 denom = vec3(7.0) + z - s; + denom = vec3(5.0) + z - s + (3.0*s - vec3(9.0))/denom; + denom = vec3(3.0) + z - s + (2.0*s - vec3(4.0))/denom; + denom = vec3(1.0) + z - s + (s - vec3(1.0))/denom; + return numerator / denom; +} + +vec2 uigamma_large_z_impl(const vec2 s, const vec2 z) +{ + // vec2 version: + const vec2 numerator = pow(z, s) * exp(-z); + vec2 denom = vec2(7.0) + z - s; + denom = vec2(5.0) + z - s + (3.0*s - vec2(9.0))/denom; + denom = vec2(3.0) + z - s + (2.0*s - vec2(4.0))/denom; + denom = vec2(1.0) + z - s + (s - vec2(1.0))/denom; + return numerator / denom; +} + +float uigamma_large_z_impl(const float s, const float z) +{ + // Float version: + const float numerator = pow(z, s) * exp(-z); + float denom = 7.0 + z - s; + denom = 5.0 + z - s + (3.0*s - 9.0)/denom; + denom = 3.0 + z - s + (2.0*s - 4.0)/denom; + denom = 1.0 + z - s + (s - 1.0)/denom; + return numerator / denom; +} + +// Normalized lower incomplete gamma function for small s (implementation): +vec4 normalized_ligamma_impl(const vec4 s, const vec4 z, + const vec4 s_inv, const vec4 gamma_s_inv) +{ + // Requires: 1.) s < ~0.5 + // 2.) s_inv = 1/s (precomputed for outside reuse) + // 3.) gamma_s_inv = 1/gamma(s) (precomputed for outside reuse) + // Returns: Approximate the normalized lower incomplete gamma function + // for s < 0.5. Since we only care about s < 0.5, we only need + // to evaluate two branches (not four) based on z. Each branch + // uses four terms, with a max relative error of ~0.00182. The + // branch threshold and specifics were adapted for fewer terms + // from Gil/Segura/Temme's paper here: + // http://oai.cwi.nl/oai/asset/20433/20433B.pdf + // Evaluate both branches: Real branches test slower even when available. + const vec4 thresh = vec4(0.775075); + const bool4 z_is_large = z > thresh; + const vec4 large_z = vec4(1.0) - uigamma_large_z_impl(s, z) * gamma_s_inv; + const vec4 small_z = ligamma_small_z_impl(s, z, s_inv) * gamma_s_inv; + // Combine the results from both branches: + return large_z * vec4(z_is_large) + small_z * vec4(!z_is_large); +} + +vec3 normalized_ligamma_impl(const vec3 s, const vec3 z, + const vec3 s_inv, const vec3 gamma_s_inv) +{ + // vec3 version: + const vec3 thresh = vec3(0.775075); + const bool3 z_is_large = z > thresh; + const vec3 large_z = vec3(1.0) - uigamma_large_z_impl(s, z) * gamma_s_inv; + const vec3 small_z = ligamma_small_z_impl(s, z, s_inv) * gamma_s_inv; + return large_z * vec3(z_is_large) + small_z * vec3(!z_is_large); +} + +vec2 normalized_ligamma_impl(const vec2 s, const vec2 z, + const vec2 s_inv, const vec2 gamma_s_inv) +{ + // vec2 version: + const vec2 thresh = vec2(0.775075); + const bool2 z_is_large = z > thresh; + const vec2 large_z = vec2(1.0) - uigamma_large_z_impl(s, z) * gamma_s_inv; + const vec2 small_z = ligamma_small_z_impl(s, z, s_inv) * gamma_s_inv; + return large_z * vec2(z_is_large) + small_z * vec2(!z_is_large); +} + +float normalized_ligamma_impl(const float s, const float z, + const float s_inv, const float gamma_s_inv) +{ + // Float version: + const float thresh = 0.775075; + const bool z_is_large = z > thresh; + const float large_z = 1.0 - uigamma_large_z_impl(s, z) * gamma_s_inv; + const float small_z = ligamma_small_z_impl(s, z, s_inv) * gamma_s_inv; + return large_z * float(z_is_large) + small_z * float(!z_is_large); +} + +// Normalized lower incomplete gamma function for small s: +vec4 normalized_ligamma(const vec4 s, const vec4 z) +{ + // Requires: s < ~0.5 + // Returns: Approximate the normalized lower incomplete gamma function + // for s < 0.5. See normalized_ligamma_impl() for details. + const vec4 s_inv = vec4(1.0)/s; + const vec4 gamma_s_inv = vec4(1.0)/gamma_impl(s, s_inv); + return normalized_ligamma_impl(s, z, s_inv, gamma_s_inv); +} + +vec3 normalized_ligamma(const vec3 s, const vec3 z) +{ + // vec3 version: + const vec3 s_inv = vec3(1.0)/s; + const vec3 gamma_s_inv = vec3(1.0)/gamma_impl(s, s_inv); + return normalized_ligamma_impl(s, z, s_inv, gamma_s_inv); +} + +vec2 normalized_ligamma(const vec2 s, const vec2 z) +{ + // vec2 version: + const vec2 s_inv = vec2(1.0)/s; + const vec2 gamma_s_inv = vec2(1.0)/gamma_impl(s, s_inv); + return normalized_ligamma_impl(s, z, s_inv, gamma_s_inv); +} + +float normalized_ligamma(const float s, const float z) +{ + // Float version: + const float s_inv = 1.0/s; + const float gamma_s_inv = 1.0/gamma_impl(s, s_inv); + return normalized_ligamma_impl(s, z, s_inv, gamma_s_inv); +} + + +#endif // SPECIAL_FUNCTIONS_H + + diff --git a/include/special-functions.h b/include/special-functions.h new file mode 100644 index 0000000..1f6d7a4 --- /dev/null +++ b/include/special-functions.h @@ -0,0 +1,492 @@ +///////////////////////////////// MIT LICENSE //////////////////////////////// + +// Copyright (C) 2014 TroggleMonkey +// +// 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. + + +///////////////////////////////// DESCRIPTION //////////////////////////////// + +// This file implements the following mathematical special functions: +// 1.) erf() = 2/sqrt(pi) * indefinite_integral(e**(-x**2)) +// 2.) gamma(s), a real-numbered extension of the integer factorial function +// It also implements normalized_ligamma(s, z), a normalized lower incomplete +// gamma function for s < 0.5 only. Both gamma() and normalized_ligamma() can +// be called with an _impl suffix to use an implementation version with a few +// extra precomputed parameters (which may be useful for the caller to reuse). +// See below for details. +// +// Design Rationale: +// Pretty much every line of code in this file is duplicated four times for +// different input types (vec4/vec3/vec2/float). This is unfortunate, +// but Cg doesn't allow function templates. Macros would be far less verbose, +// but they would make the code harder to document and read. I don't expect +// these functions will require a whole lot of maintenance changes unless +// someone ever has need for more robust incomplete gamma functions, so code +// duplication seems to be the lesser evil in this case. + + +/////////////////////////// GAUSSIAN ERROR FUNCTION ////////////////////////// + +vec4 erf6(vec4 x) +{ + // Requires: x is the standard parameter to erf(). + // Returns: Return an Abramowitz/Stegun approximation of erf(), where: + // erf(x) = 2/sqrt(pi) * integral(e**(-x**2)) + // This approximation has a max absolute error of 2.5*10**-5 + // with solid numerical robustness and efficiency. See: + // https://en.wikipedia.org/wiki/Error_function#Approximation_with_elementary_functions + const vec4 one = vec4(1.0); + const vec4 sign_x = sign(x); + const vec4 t = one/(one + 0.47047*abs(x)); + const vec4 result = one - t*(0.3480242 + t*(-0.0958798 + t*0.7478556))* + exp(-(x*x)); + return result * sign_x; +} + +vec3 erf6(const vec3 x) +{ + // vec3 version: + const vec3 one = vec3(1.0); + const vec3 sign_x = sign(x); + const vec3 t = one/(one + 0.47047*abs(x)); + const vec3 result = one - t*(0.3480242 + t*(-0.0958798 + t*0.7478556))* + exp(-(x*x)); + return result * sign_x; +} + +vec2 erf6(const vec2 x) +{ + // vec2 version: + const vec2 one = vec2(1.0); + const vec2 sign_x = sign(x); + const vec2 t = one/(one + 0.47047*abs(x)); + const vec2 result = one - t*(0.3480242 + t*(-0.0958798 + t*0.7478556))* + exp(-(x*x)); + return result * sign_x; +} + +float erf6(const float x) +{ + // Float version: + const float sign_x = sign(x); + const float t = 1.0/(1.0 + 0.47047*abs(x)); + const float result = 1.0 - t*(0.3480242 + t*(-0.0958798 + t*0.7478556))* + exp(-(x*x)); + return result * sign_x; +} + +vec4 erft(const vec4 x) +{ + // Requires: x is the standard parameter to erf(). + // Returns: Approximate erf() with the hyperbolic tangent. The error is + // visually noticeable, but it's blazing fast and perceptually + // close...at least on ATI hardware. See: + // http://www.maplesoft.com/applications/view.aspx?SID=5525&view=html + // Warning: Only use this if your hardware drivers correctly implement + // tanh(): My nVidia 8800GTS returns garbage output. + return tanh(1.202760580 * x); +} + +vec3 erft(const vec3 x) +{ + // vec3 version: + return tanh(1.202760580 * x); +} + +vec2 erft(const vec2 x) +{ + // vec2 version: + return tanh(1.202760580 * x); +} + +float erft(const float x) +{ + // Float version: + return tanh(1.202760580 * x); +} + +vec4 erf(const vec4 x) +{ + // Requires: x is the standard parameter to erf(). + // Returns: Some approximation of erf(x), depending on user settings. + #ifdef ERF_FAST_APPROXIMATION + return erft(x); + #else + return erf6(x); + #endif +} + +vec3 erf(const vec3 x) +{ + // vec3 version: + #ifdef ERF_FAST_APPROXIMATION + return erft(x); + #else + return erf6(x); + #endif +} + +vec2 erf(const vec2 x) +{ + // vec2 version: + #ifdef ERF_FAST_APPROXIMATION + return erft(x); + #else + return erf6(x); + #endif +} + +float erf(const float x) +{ + // Float version: + #ifdef ERF_FAST_APPROXIMATION + return erft(x); + #else + return erf6(x); + #endif +} + +/////////////////////////// COMPLETE GAMMA FUNCTION ////////////////////////// + +vec4 gamma_impl(const vec4 s, const vec4 s_inv) +{ + // Requires: 1.) s is the standard parameter to the gamma function, and + // it should lie in the [0, 36] range. + // 2.) s_inv = 1.0/s. This implementation function requires + // the caller to precompute this value, giving users the + // opportunity to reuse it. + // Returns: Return approximate gamma function (real-numbered factorial) + // output using the Lanczos approximation with two coefficients + // calculated using Paul Godfrey's method here: + // http://my.fit.edu/~gabdo/gamma.txt + // An optimal g value for s in [0, 36] is ~1.12906830989, with + // a maximum relative error of 0.000463 for 2**16 equally + // evals. We could use three coeffs (0.0000346 error) without + // hurting latency, but this allows more parallelism with + // outside instructions. + const vec4 g = vec4(1.12906830989); + const vec4 c0 = vec4(0.8109119309638332633713423362694399653724431); + const vec4 c1 = vec4(0.4808354605142681877121661197951496120000040); + const vec4 e = vec4(2.71828182845904523536028747135266249775724709); + const vec4 sph = s + vec4(0.5); + const vec4 lanczos_sum = c0 + c1/(s + vec4(1.0)); + const vec4 base = (sph + g)/e; // or (s + g + vec4(0.5))/e + // gamma(s + 1) = base**sph * lanczos_sum; divide by s for gamma(s). + // This has less error for small s's than (s -= 1.0) at the beginning. + return (pow(base, sph) * lanczos_sum) * s_inv; +} + +vec3 gamma_impl(const vec3 s, const vec3 s_inv) +{ + // vec3 version: + const vec3 g = vec3(1.12906830989); + const vec3 c0 = vec3(0.8109119309638332633713423362694399653724431); + const vec3 c1 = vec3(0.4808354605142681877121661197951496120000040); + const vec3 e = vec3(2.71828182845904523536028747135266249775724709); + const vec3 sph = s + vec3(0.5); + const vec3 lanczos_sum = c0 + c1/(s + vec3(1.0)); + const vec3 base = (sph + g)/e; + return (pow(base, sph) * lanczos_sum) * s_inv; +} + +vec2 gamma_impl(const vec2 s, const vec2 s_inv) +{ + // vec2 version: + const vec2 g = vec2(1.12906830989); + const vec2 c0 = vec2(0.8109119309638332633713423362694399653724431); + const vec2 c1 = vec2(0.4808354605142681877121661197951496120000040); + const vec2 e = vec2(2.71828182845904523536028747135266249775724709); + const vec2 sph = s + vec2(0.5); + const vec2 lanczos_sum = c0 + c1/(s + vec2(1.0)); + const vec2 base = (sph + g)/e; + return (pow(base, sph) * lanczos_sum) * s_inv; +} + +float gamma_impl(const float s, const float s_inv) +{ + // Float version: + const float g = 1.12906830989; + const float c0 = 0.8109119309638332633713423362694399653724431; + const float c1 = 0.4808354605142681877121661197951496120000040; + const float e = 2.71828182845904523536028747135266249775724709; + const float sph = s + 0.5; + const float lanczos_sum = c0 + c1/(s + 1.0); + const float base = (sph + g)/e; + return (pow(base, sph) * lanczos_sum) * s_inv; +} + +vec4 gamma(const vec4 s) +{ + // Requires: s is the standard parameter to the gamma function, and it + // should lie in the [0, 36] range. + // Returns: Return approximate gamma function output with a maximum + // relative error of 0.000463. See gamma_impl for details. + return gamma_impl(s, vec4(1.0)/s); +} + +vec3 gamma(const vec3 s) +{ + // vec3 version: + return gamma_impl(s, vec3(1.0)/s); +} + +vec2 gamma(const vec2 s) +{ + // vec2 version: + return gamma_impl(s, vec2(1.0)/s); +} + +float gamma(const float s) +{ + // Float version: + return gamma_impl(s, 1.0/s); +} + +//////////////// INCOMPLETE GAMMA FUNCTIONS (RESTRICTED INPUT) /////////////// + +// Lower incomplete gamma function for small s and z (implementation): +vec4 ligamma_small_z_impl(const vec4 s, const vec4 z, const vec4 s_inv) +{ + // Requires: 1.) s < ~0.5 + // 2.) z <= ~0.775075 + // 3.) s_inv = 1.0/s (precomputed for outside reuse) + // Returns: A series representation for the lower incomplete gamma + // function for small s and small z (4 terms). + // The actual "rolled up" summation looks like: + // last_sign = 1.0; last_pow = 1.0; last_factorial = 1.0; + // sum = last_sign * last_pow / ((s + k) * last_factorial) + // for(int i = 0; i < 4; ++i) + // { + // last_sign *= -1.0; last_pow *= z; last_factorial *= i; + // sum += last_sign * last_pow / ((s + k) * last_factorial); + // } + // Unrolled, constant-unfolded and arranged for madds and parallelism: + const vec4 scale = pow(z, s); + vec4 sum = s_inv; // Summation iteration 0 result + // Summation iterations 1, 2, and 3: + const vec4 z_sq = z*z; + const vec4 denom1 = s + vec4(1.0); + const vec4 denom2 = 2.0*s + vec4(4.0); + const vec4 denom3 = 6.0*s + vec4(18.0); + //vec4 denom4 = 24.0*s + vec4(96.0); + sum -= z/denom1; + sum += z_sq/denom2; + sum -= z * z_sq/denom3; + //sum += z_sq * z_sq / denom4; + // Scale and return: + return scale * sum; +} + +vec3 ligamma_small_z_impl(const vec3 s, const vec3 z, const vec3 s_inv) +{ + // vec3 version: + const vec3 scale = pow(z, s); + vec3 sum = s_inv; + const vec3 z_sq = z*z; + const vec3 denom1 = s + vec3(1.0); + const vec3 denom2 = 2.0*s + vec3(4.0); + const vec3 denom3 = 6.0*s + vec3(18.0); + sum -= z/denom1; + sum += z_sq/denom2; + sum -= z * z_sq/denom3; + return scale * sum; +} + +vec2 ligamma_small_z_impl(const vec2 s, const vec2 z, const vec2 s_inv) +{ + // vec2 version: + const vec2 scale = pow(z, s); + vec2 sum = s_inv; + const vec2 z_sq = z*z; + const vec2 denom1 = s + vec2(1.0); + const vec2 denom2 = 2.0*s + vec2(4.0); + const vec2 denom3 = 6.0*s + vec2(18.0); + sum -= z/denom1; + sum += z_sq/denom2; + sum -= z * z_sq/denom3; + return scale * sum; +} + +float ligamma_small_z_impl(const float s, const float z, const float s_inv) +{ + // Float version: + const float scale = pow(z, s); + float sum = s_inv; + const float z_sq = z*z; + const float denom1 = s + 1.0; + const float denom2 = 2.0*s + 4.0; + const float denom3 = 6.0*s + 18.0; + sum -= z/denom1; + sum += z_sq/denom2; + sum -= z * z_sq/denom3; + return scale * sum; +} + +// Upper incomplete gamma function for small s and large z (implementation): +vec4 uigamma_large_z_impl(const vec4 s, const vec4 z) +{ + // Requires: 1.) s < ~0.5 + // 2.) z > ~0.775075 + // Returns: Gauss's continued fraction representation for the upper + // incomplete gamma function (4 terms). + // The "rolled up" continued fraction looks like this. The denominator + // is truncated, and it's calculated "from the bottom up:" + // denom = vec4('inf'); + // vec4 one = vec4(1.0); + // for(int i = 4; i > 0; --i) + // { + // denom = ((i * 2.0) - one) + z - s + (i * (s - i))/denom; + // } + // Unrolled and constant-unfolded for madds and parallelism: + const vec4 numerator = pow(z, s) * exp(-z); + vec4 denom = vec4(7.0) + z - s; + denom = vec4(5.0) + z - s + (3.0*s - vec4(9.0))/denom; + denom = vec4(3.0) + z - s + (2.0*s - vec4(4.0))/denom; + denom = vec4(1.0) + z - s + (s - vec4(1.0))/denom; + return numerator / denom; +} + +vec3 uigamma_large_z_impl(const vec3 s, const vec3 z) +{ + // vec3 version: + const vec3 numerator = pow(z, s) * exp(-z); + vec3 denom = vec3(7.0) + z - s; + denom = vec3(5.0) + z - s + (3.0*s - vec3(9.0))/denom; + denom = vec3(3.0) + z - s + (2.0*s - vec3(4.0))/denom; + denom = vec3(1.0) + z - s + (s - vec3(1.0))/denom; + return numerator / denom; +} + +vec2 uigamma_large_z_impl(const vec2 s, const vec2 z) +{ + // vec2 version: + const vec2 numerator = pow(z, s) * exp(-z); + vec2 denom = vec2(7.0) + z - s; + denom = vec2(5.0) + z - s + (3.0*s - vec2(9.0))/denom; + denom = vec2(3.0) + z - s + (2.0*s - vec2(4.0))/denom; + denom = vec2(1.0) + z - s + (s - vec2(1.0))/denom; + return numerator / denom; +} + +float uigamma_large_z_impl(const float s, const float z) +{ + // Float version: + const float numerator = pow(z, s) * exp(-z); + float denom = 7.0 + z - s; + denom = 5.0 + z - s + (3.0*s - 9.0)/denom; + denom = 3.0 + z - s + (2.0*s - 4.0)/denom; + denom = 1.0 + z - s + (s - 1.0)/denom; + return numerator / denom; +} + +// Normalized lower incomplete gamma function for small s (implementation): +vec4 normalized_ligamma_impl(const vec4 s, const vec4 z, + const vec4 s_inv, const vec4 gamma_s_inv) +{ + // Requires: 1.) s < ~0.5 + // 2.) s_inv = 1/s (precomputed for outside reuse) + // 3.) gamma_s_inv = 1/gamma(s) (precomputed for outside reuse) + // Returns: Approximate the normalized lower incomplete gamma function + // for s < 0.5. Since we only care about s < 0.5, we only need + // to evaluate two branches (not four) based on z. Each branch + // uses four terms, with a max relative error of ~0.00182. The + // branch threshold and specifics were adapted for fewer terms + // from Gil/Segura/Temme's paper here: + // http://oai.cwi.nl/oai/asset/20433/20433B.pdf + // Evaluate both branches: Real branches test slower even when available. + const vec4 thresh = vec4(0.775075); + bvec4 z_is_large = greaterThan(z , thresh); + vec4 z_size_check = vec4(z_is_large.x ? 1.0 : 0.0, z_is_large.y ? 1.0 : 0.0, z_is_large.z ? 1.0 : 0.0, z_is_large.w ? 1.0 : 0.0); + const vec4 large_z = vec4(1.0) - uigamma_large_z_impl(s, z) * gamma_s_inv; + const vec4 small_z = ligamma_small_z_impl(s, z, s_inv) * gamma_s_inv; + // Combine the results from both branches: + return large_z * vec4(z_size_check) + small_z * vec4(z_size_check); +} + +vec3 normalized_ligamma_impl(const vec3 s, const vec3 z, + const vec3 s_inv, const vec3 gamma_s_inv) +{ + // vec3 version: + const vec3 thresh = vec3(0.775075); + bvec3 z_is_large = greaterThan(z , thresh); + vec3 z_size_check = vec3(z_is_large.x ? 1.0 : 0.0, z_is_large.y ? 1.0 : 0.0, z_is_large.z ? 1.0 : 0.0); + const vec3 large_z = vec3(1.0) - uigamma_large_z_impl(s, z) * gamma_s_inv; + const vec3 small_z = ligamma_small_z_impl(s, z, s_inv) * gamma_s_inv; + return large_z * vec3(z_size_check) + small_z * vec3(z_size_check); +} + +vec2 normalized_ligamma_impl(const vec2 s, const vec2 z, + const vec2 s_inv, const vec2 gamma_s_inv) +{ + // vec2 version: + const vec2 thresh = vec2(0.775075); + bvec2 z_is_large = greaterThan(z , thresh); + vec2 z_size_check = vec2(z_is_large.x ? 1.0 : 0.0, z_is_large.y ? 1.0 : 0.0); + const vec2 large_z = vec2(1.0) - uigamma_large_z_impl(s, z) * gamma_s_inv; + const vec2 small_z = ligamma_small_z_impl(s, z, s_inv) * gamma_s_inv; + return large_z * vec2(z_size_check) + small_z * vec2(z_size_check); +} + +float normalized_ligamma_impl(const float s, const float z, + const float s_inv, const float gamma_s_inv) +{ + // Float version: + const float thresh = 0.775075; + float z_size_check = 0.0; + if (z > thresh) z_size_check = 1.0; + const float large_z = 1.0 - uigamma_large_z_impl(s, z) * gamma_s_inv; + const float small_z = ligamma_small_z_impl(s, z, s_inv) * gamma_s_inv; + return large_z * float(z_size_check) + small_z * float(z_size_check); +} + +// Normalized lower incomplete gamma function for small s: +vec4 normalized_ligamma(const vec4 s, const vec4 z) +{ + // Requires: s < ~0.5 + // Returns: Approximate the normalized lower incomplete gamma function + // for s < 0.5. See normalized_ligamma_impl() for details. + const vec4 s_inv = vec4(1.0)/s; + const vec4 gamma_s_inv = vec4(1.0)/gamma_impl(s, s_inv); + return normalized_ligamma_impl(s, z, s_inv, gamma_s_inv); +} + +vec3 normalized_ligamma(const vec3 s, const vec3 z) +{ + // vec3 version: + const vec3 s_inv = vec3(1.0)/s; + const vec3 gamma_s_inv = vec3(1.0)/gamma_impl(s, s_inv); + return normalized_ligamma_impl(s, z, s_inv, gamma_s_inv); +} + +vec2 normalized_ligamma(const vec2 s, const vec2 z) +{ + // vec2 version: + const vec2 s_inv = vec2(1.0)/s; + const vec2 gamma_s_inv = vec2(1.0)/gamma_impl(s, s_inv); + return normalized_ligamma_impl(s, z, s_inv, gamma_s_inv); +} + +float normalized_ligamma(const float s, const float z) +{ + // Float version: + const float s_inv = 1.0/s; + const float gamma_s_inv = 1.0/gamma_impl(s, s_inv); + return normalized_ligamma_impl(s, z, s_inv, gamma_s_inv); +} \ No newline at end of file