1// Bloom works by creating an intermediate texture with a bunch of mip levels, each half the size of the previous.2// You then downsample each mip (starting with the original texture) to the lower resolution mip under it, going in order.3// You then upsample each mip (starting from the smallest mip) and blend with the higher resolution mip above it (ending on the original texture).4//5// References:6// * [COD] - Next Generation Post Processing in Call of Duty - http://www.iryoku.com/next-generation-post-processing-in-call-of-duty-advanced-warfare7// * [PBB] - Physically Based Bloom - https://learnopengl.com/Guest-Articles/2022/Phys.-Based-Bloom9struct BloomUniforms {10 threshold_precomputations: vec4<f32>,11 viewport: vec4<f32>,12 scale: vec2<f32>,13 aspect: f32,14};16@group(0) @binding(0) var input_texture: texture_2d<f32>;17@group(0) @binding(1) var s: sampler;19@group(0) @binding(2) var<uniform> uniforms: BloomUniforms;21#ifdef FIRST_DOWNSAMPLE22// https://catlikecoding.com/unity/tutorials/advanced-rendering/bloom/#3.423fn soft_threshold(color: vec3<f32>) -> vec3<f32> {24 let brightness = max(color.r, max(color.g, color.b));25 var softness = brightness - uniforms.threshold_precomputations.y;26 softness = clamp(softness, 0.0, uniforms.threshold_precomputations.z);27 softness = softness * softness * uniforms.threshold_precomputations.w;28 var contribution = max(brightness - uniforms.threshold_precomputations.x, softness);29 contribution /= max(brightness, 0.00001); // Prevent division by 030 return color * contribution;31}32#endif34// luminance coefficients from Rec. 709.35// https://en.wikipedia.org/wiki/Rec._70936fn tonemapping_luminance(v: vec3<f32>) -> f32 {37 return dot(v, vec3<f32>(0.2126, 0.7152, 0.0722));38}40// http://graphicrants.blogspot.com/2013/12/tone-mapping.html41fn karis_average(color: vec3<f32>) -> f32 {42 // Luminance calculated based on Rec. 709 color primaries.43 // This must be done in *linear* color space.44 let luma = tonemapping_luminance(color) / 4.0;45 return 1.0 / (1.0 + luma);46}48// [COD] slide 15349fn sample_input_13_tap(uv: vec2<f32>) -> vec3<f32> {50#ifdef UNIFORM_SCALE51 // This is the fast path. When the bloom scale is uniform, the 13 tap sampling kernel can be52 // expressed with constant offsets.53 //54 // It's possible that this isn't meaningfully faster than the "slow" path. However, because it55 // is hard to test performance on all platforms, and uniform bloom is the most common case, this56 // path was retained when adding non-uniform (anamorphic) bloom. This adds a small, but nonzero,57 // cost to maintainability, but it does help me sleep at night.58 let a = textureSample(input_texture, s, uv, vec2<i32>(-2, 2)).rgb;59 let b = textureSample(input_texture, s, uv, vec2<i32>(0, 2)).rgb;60 let c = textureSample(input_texture, s, uv, vec2<i32>(2, 2)).rgb;61 let d = textureSample(input_texture, s, uv, vec2<i32>(-2, 0)).rgb;62 let e = textureSample(input_texture, s, uv).rgb;63 let f = textureSample(input_texture, s, uv, vec2<i32>(2, 0)).rgb;64 let g = textureSample(input_texture, s, uv, vec2<i32>(-2, -2)).rgb;65 let h = textureSample(input_texture, s, uv, vec2<i32>(0, -2)).rgb;66 let i = textureSample(input_texture, s, uv, vec2<i32>(2, -2)).rgb;67 let j = textureSample(input_texture, s, uv, vec2<i32>(-1, 1)).rgb;68 let k = textureSample(input_texture, s, uv, vec2<i32>(1, 1)).rgb;69 let l = textureSample(input_texture, s, uv, vec2<i32>(-1, -1)).rgb;70 let m = textureSample(input_texture, s, uv, vec2<i32>(1, -1)).rgb;71#else72 // This is the flexible, but potentially slower, path for non-uniform sampling. Because the73 // sample is not a constant, and it can fall outside of the limits imposed on constant sample74 // offsets (-8..8), we have to compute the pixel offset in uv coordinates using the size of the75 // texture.76 //77 // It isn't clear if this is meaningfully slower than using the offset syntax, the spec doesn't78 // mention it anywhere: https://www.w3.org/TR/WGSL/#texturesample, but the fact that the offset79 // syntax uses a const-expr implies that it allows some compiler optimizations - maybe more80 // impactful on mobile?81 let scale = uniforms.scale;82 let ps = scale / vec2<f32>(textureDimensions(input_texture));83 let pl = 2.0 * ps;84 let ns = -1.0 * ps;85 let nl = -2.0 * ps;86 let a = textureSample(input_texture, s, uv + vec2<f32>(nl.x, pl.y)).rgb;87 let b = textureSample(input_texture, s, uv + vec2<f32>(0.00, pl.y)).rgb;88 let c = textureSample(input_texture, s, uv + vec2<f32>(pl.x, pl.y)).rgb;89 let d = textureSample(input_texture, s, uv + vec2<f32>(nl.x, 0.00)).rgb;90 let e = textureSample(input_texture, s, uv).rgb;91 let f = textureSample(input_texture, s, uv + vec2<f32>(pl.x, 0.00)).rgb;92 let g = textureSample(input_texture, s, uv + vec2<f32>(nl.x, nl.y)).rgb;93 let h = textureSample(input_texture, s, uv + vec2<f32>(0.00, nl.y)).rgb;94 let i = textureSample(input_texture, s, uv + vec2<f32>(pl.x, nl.y)).rgb;95 let j = textureSample(input_texture, s, uv + vec2<f32>(ns.x, ps.y)).rgb;96 let k = textureSample(input_texture, s, uv + vec2<f32>(ps.x, ps.y)).rgb;97 let l = textureSample(input_texture, s, uv + vec2<f32>(ns.x, ns.y)).rgb;98 let m = textureSample(input_texture, s, uv + vec2<f32>(ps.x, ns.y)).rgb;99#endif101#ifdef FIRST_DOWNSAMPLE102 // [COD] slide 168103 //104 // The first downsample pass reads from the rendered frame which may exhibit105 // 'fireflies' (individual very bright pixels) that should not cause the bloom effect.106 //107 // The first downsample uses a firefly-reduction method proposed by Brian Karis108 // which takes a weighted-average of the samples to limit their luma range to [0, 1].109 // This implementation matches the LearnOpenGL article [PBB].110 var group0 = (a + b + d + e) * (0.125f / 4.0f);111 var group1 = (b + c + e + f) * (0.125f / 4.0f);112 var group2 = (d + e + g + h) * (0.125f / 4.0f);113 var group3 = (e + f + h + i) * (0.125f / 4.0f);114 var group4 = (j + k + l + m) * (0.5f / 4.0f);115 group0 *= karis_average(group0);116 group1 *= karis_average(group1);117 group2 *= karis_average(group2);118 group3 *= karis_average(group3);119 group4 *= karis_average(group4);120 return group0 + group1 + group2 + group3 + group4;121#else122 var sample = (a + c + g + i) * 0.03125;123 sample += (b + d + f + h) * 0.0625;124 sample += (e + j + k + l + m) * 0.125;125 return sample;126#endif127}129// [COD] slide 162130fn sample_input_3x3_tent(uv: vec2<f32>) -> vec3<f32> {131 // While this is probably technically incorrect, it makes nonuniform bloom smoother, without132 // having any impact on uniform bloom, which simply evaluates to 1.0 here.133 let frag_size = uniforms.scale / vec2<f32>(textureDimensions(input_texture));134 let x = frag_size.x;135 let y = frag_size.y;137 let a = textureSample(input_texture, s, vec2<f32>(uv.x - x, uv.y + y)).rgb;138 let b = textureSample(input_texture, s, vec2<f32>(uv.x, uv.y + y)).rgb;139 let c = textureSample(input_texture, s, vec2<f32>(uv.x + x, uv.y + y)).rgb;141 let d = textureSample(input_texture, s, vec2<f32>(uv.x - x, uv.y)).rgb;142 let e = textureSample(input_texture, s, vec2<f32>(uv.x, uv.y)).rgb;143 let f = textureSample(input_texture, s, vec2<f32>(uv.x + x, uv.y)).rgb;145 let g = textureSample(input_texture, s, vec2<f32>(uv.x - x, uv.y - y)).rgb;146 let h = textureSample(input_texture, s, vec2<f32>(uv.x, uv.y - y)).rgb;147 let i = textureSample(input_texture, s, vec2<f32>(uv.x + x, uv.y - y)).rgb;149 var sample = e * 0.25;150 sample += (b + d + f + h) * 0.125;151 sample += (a + c + g + i) * 0.0625;153 return sample;154}156#ifdef FIRST_DOWNSAMPLE157@fragment158fn downsample_first(@location(0) output_uv: vec2<f32>) -> @location(0) vec4<f32> {159 let sample_uv = uniforms.viewport.xy + output_uv * uniforms.viewport.zw;160 var sample = sample_input_13_tap(sample_uv);161 // Lower bound of 0.0001 is to avoid propagating multiplying by 0.0 through the162 // downscaling and upscaling which would result in black boxes.163 // The upper bound is to prevent NaNs.164 // with f32::MAX (E+38) Chrome fails with ":value 340282346999999984391321947108527833088.0 cannot be represented as 'f32'"165 sample = clamp(sample, vec3<f32>(0.0001), vec3<f32>(3.40282347E+37));167#ifdef USE_THRESHOLD168 sample = soft_threshold(sample);169#endif171 return vec4<f32>(sample, 1.0);172}173#endif175@fragment176fn downsample(@location(0) uv: vec2<f32>) -> @location(0) vec4<f32> {177 return vec4<f32>(sample_input_13_tap(uv), 1.0);178}180@fragment181fn upsample(@location(0) uv: vec2<f32>) -> @location(0) vec4<f32> {182 return vec4<f32>(sample_input_3x3_tent(uv), 1.0);183}
Search across packages, versions, files, declarations, and comments.
Use multiple terms to narrow results (AND), and add @vertex,
@fragment, or @compute to filter by shader stage.