mirror of
https://github.com/italicsjenga/slang-shaders.git
synced 2024-11-22 15:51:30 +11:00
Merge pull request #90 from Themaister/bandlimited-pixel
Add the pseudo-bandlimited pixel upscaler.
This commit is contained in:
commit
b961d1ad1d
10
retro/bandlimit-pixel.slangp
Normal file
10
retro/bandlimit-pixel.slangp
Normal file
|
@ -0,0 +1,10 @@
|
|||
shaders = 2
|
||||
shader0 = ../linear/linearize.slang
|
||||
shader1 = shaders/bandlimit-pixel.slang
|
||||
|
||||
filter_linear0 = false
|
||||
scale_type0 = source
|
||||
scale0 = 1.0
|
||||
srgb_framebuffer0 = true
|
||||
|
||||
filter_linear1 = true
|
170
retro/shaders/bandlimit-pixel.slang
Normal file
170
retro/shaders/bandlimit-pixel.slang
Normal file
|
@ -0,0 +1,170 @@
|
|||
#version 450
|
||||
|
||||
/*
|
||||
* Bandlimited pixel footprint shader.
|
||||
* Author: Themaister
|
||||
* License: MIT
|
||||
* Adapted from: https://github.com/Themaister/Granite/blob/master/assets/shaders/inc/bandlimited_pixel_filter.h
|
||||
*/
|
||||
|
||||
precision highp float;
|
||||
precision highp int;
|
||||
|
||||
layout(std140, set = 0, binding = 0) uniform UBO
|
||||
{
|
||||
mat4 MVP;
|
||||
vec4 SourceSize;
|
||||
float SMOOTHNESS;
|
||||
} global;
|
||||
|
||||
#pragma parameter SMOOTHNESS "Smoothness" 0.5 0.0 5.0 0.1
|
||||
|
||||
#pragma stage vertex
|
||||
layout(location = 0) in vec4 Position;
|
||||
layout(location = 1) in vec2 TexCoord;
|
||||
layout(location = 0) out vec2 vTexCoord;
|
||||
|
||||
void main()
|
||||
{
|
||||
gl_Position = global.MVP * Position;
|
||||
vTexCoord = TexCoord;
|
||||
}
|
||||
|
||||
#pragma stage fragment
|
||||
layout(location = 0) in vec2 vTexCoord;
|
||||
layout(location = 0) out mediump vec4 FragColor;
|
||||
layout(set = 0, binding = 1) uniform mediump sampler2D Source;
|
||||
|
||||
// The cosine filter convolved with rect has a support of 0.5 + d pixels.
|
||||
// We can sample 4x4 regions, so we can deal with 2.0 pixel range in our filter,
|
||||
// and the maximum extent value we can have is 1.5.
|
||||
const float maximum_support_extent = 1.5;
|
||||
|
||||
struct BandlimitedPixelInfo
|
||||
{
|
||||
vec2 uv0;
|
||||
vec2 uv1;
|
||||
vec2 uv2;
|
||||
vec2 uv3;
|
||||
mediump vec4 weights;
|
||||
mediump float l;
|
||||
};
|
||||
|
||||
// Our Taylor approximation is not exact, normalize so the peak is 1.
|
||||
const float taylor_pi_half = 1.00452485553;
|
||||
const float taylor_normalization = 1.0 / taylor_pi_half;
|
||||
const float PI = 3.14159265359;
|
||||
const float PI_half = 0.5 * PI;
|
||||
|
||||
#define gen_taylor(T) \
|
||||
mediump T taylor_sin(mediump T p) \
|
||||
{ \
|
||||
mediump T p2 = p * p; \
|
||||
mediump T p3 = p * p2; \
|
||||
mediump T p5 = p2 * p3; \
|
||||
return clamp(taylor_normalization * (p - p3 * (1.0 / 6.0) + p5 * (1.0 / 120.0)), -1.0, 1.0); \
|
||||
}
|
||||
// No templates in GLSL. Stamp out macros.
|
||||
gen_taylor(float)
|
||||
gen_taylor(vec2)
|
||||
gen_taylor(vec3)
|
||||
gen_taylor(vec4)
|
||||
|
||||
// Given weights, compute a bilinear filter which implements the weight.
|
||||
// All weights are known to be non-negative, and separable.
|
||||
mediump vec3 compute_uv_phase_weight(mediump vec2 weights_u, mediump vec2 weights_v)
|
||||
{
|
||||
// The sum of a bilinear sample has combined weight of 1, we will need to adjust the resulting sample
|
||||
// to match our actual weight sum.
|
||||
mediump float w = dot(weights_u.xyxy, weights_v.xxyy);
|
||||
mediump float x = weights_u.y / max(weights_u.x + weights_u.y, 0.001);
|
||||
mediump float y = weights_v.y / max(weights_v.x + weights_v.y, 0.001);
|
||||
return vec3(x, y, w);
|
||||
}
|
||||
|
||||
BandlimitedPixelInfo compute_pixel_weights(vec2 uv, vec2 size, vec2 inv_size)
|
||||
{
|
||||
// Get derivatives in texel space.
|
||||
// Need a non-zero derivative.
|
||||
vec2 extent = max(fwidth(uv) * size * (global.SMOOTHNESS + 0.5), 1.0 / 256.0);
|
||||
|
||||
// Get base pixel and phase, range [0, 1).
|
||||
vec2 pixel = uv * size - 0.5;
|
||||
vec2 base_pixel = floor(pixel);
|
||||
vec2 phase = pixel - base_pixel;
|
||||
|
||||
BandlimitedPixelInfo info;
|
||||
|
||||
mediump vec2 inv_extent = 1.0 / extent;
|
||||
if (any(greaterThan(extent, vec2(maximum_support_extent))))
|
||||
{
|
||||
// We need to just do regular minimization filtering.
|
||||
info = BandlimitedPixelInfo(vec2(0.0), vec2(0.0), vec2(0.0), vec2(0.0),
|
||||
vec4(0.0, 0.0, 0.0, 0.0), 0.0);
|
||||
}
|
||||
else if (all(lessThanEqual(extent, vec2(0.5))))
|
||||
{
|
||||
// We can resolve the filter by just sampling a single 2x2 block.
|
||||
mediump vec2 shift = 0.5 + 0.5 * taylor_sin(PI_half * clamp(inv_extent * (phase - 0.5), -1.0, 1.0));
|
||||
info = BandlimitedPixelInfo((base_pixel + 0.5 + shift) * inv_size, vec2(0.0), vec2(0.0), vec2(0.0),
|
||||
vec4(1.0, 0.0, 0.0, 0.0), 1.0);
|
||||
}
|
||||
else
|
||||
{
|
||||
// Full 4x4 sampling.
|
||||
|
||||
// Fade between bandlimited and normal sampling.
|
||||
// Fully use bandlimited filter at LOD 0, normal filtering at approx. LOD -0.5.
|
||||
mediump float max_extent = max(extent.x, extent.y);
|
||||
mediump float l = clamp(1.0 - (max_extent - 1.0) / (maximum_support_extent - 1.0), 0.0, 1.0);
|
||||
|
||||
mediump vec4 sine_phases_x = PI_half * clamp(inv_extent.x * (phase.x + vec4(1.5, 0.5, -0.5, -1.5)), -1.0, 1.0);
|
||||
mediump vec4 sines_x = taylor_sin(sine_phases_x);
|
||||
|
||||
mediump vec4 sine_phases_y = PI_half * clamp(inv_extent.y * (phase.y + vec4(1.5, 0.5, -0.5, -1.5)), -1.0, 1.0);
|
||||
mediump vec4 sines_y = taylor_sin(sine_phases_y);
|
||||
|
||||
mediump vec2 sine_phases_end = PI_half * clamp(inv_extent * (phase - 2.5), -1.0, 1.0);
|
||||
mediump vec2 sines_end = taylor_sin(sine_phases_end);
|
||||
|
||||
mediump vec4 weights_x = 0.5 * (sines_x - vec4(sines_x.yzw, sines_end.x));
|
||||
mediump vec4 weights_y = 0.5 * (sines_y - vec4(sines_y.yzw, sines_end.y));
|
||||
|
||||
mediump vec3 w0 = compute_uv_phase_weight(weights_x.xy, weights_y.xy);
|
||||
mediump vec3 w1 = compute_uv_phase_weight(weights_x.zw, weights_y.xy);
|
||||
mediump vec3 w2 = compute_uv_phase_weight(weights_x.xy, weights_y.zw);
|
||||
mediump vec3 w3 = compute_uv_phase_weight(weights_x.zw, weights_y.zw);
|
||||
|
||||
info = BandlimitedPixelInfo((base_pixel - 0.5 + w0.xy) * inv_size,
|
||||
(base_pixel + vec2(1.5, -0.5) + w1.xy) * inv_size,
|
||||
(base_pixel + vec2(-0.5, 1.5) + w2.xy) * inv_size,
|
||||
(base_pixel + 1.5 + w3.xy) * inv_size,
|
||||
vec4(w0.z, w1.z, w2.z, w3.z), l);
|
||||
}
|
||||
|
||||
return info;
|
||||
}
|
||||
|
||||
mediump vec4 sample_bandlimited_pixel(sampler2D samp, vec2 uv, BandlimitedPixelInfo info, float lod)
|
||||
{
|
||||
mediump vec4 color = texture(samp, uv);
|
||||
if (info.l > 0.0)
|
||||
{
|
||||
mediump vec4 bandlimited = info.weights.x * textureLod(samp, info.uv0, lod);
|
||||
if (info.weights.x < 1.0)
|
||||
{
|
||||
bandlimited += info.weights.y * textureLod(samp, info.uv1, lod);
|
||||
bandlimited += info.weights.z * textureLod(samp, info.uv2, lod);
|
||||
bandlimited += info.weights.w * textureLod(samp, info.uv3, lod);
|
||||
}
|
||||
color = mix(color, bandlimited, info.l);
|
||||
}
|
||||
return color;
|
||||
}
|
||||
|
||||
void main()
|
||||
{
|
||||
BandlimitedPixelInfo info = compute_pixel_weights(vTexCoord, global.SourceSize.xy, global.SourceSize.zw);
|
||||
mediump vec3 result = sample_bandlimited_pixel(Source, vTexCoord, info, 0.0).rgb;
|
||||
FragColor = vec4(sqrt(clamp(result, 0.0, 1.0)), 1.0);
|
||||
}
|
Loading…
Reference in a new issue