FiE-Game/Assets/PostProcessing/Resources/Shaders/AmbientOcclusion.cginc

501 lines
15 KiB
HLSL
Raw Normal View History

2023-07-24 21:52:50 +02:00
// Upgrade NOTE: commented out 'float4x4 _WorldToCamera', a built-in variable
// Upgrade NOTE: replaced '_WorldToCamera' with 'unity_WorldToCamera'
#ifndef __AMBIENT_OCCLUSION__
#define __AMBIENT_OCCLUSION__
#include "UnityCG.cginc"
#include "Common.cginc"
// --------
// Options for further customization
// --------
// By default, a 5-tap Gaussian with the linear sampling technique is used
// in the bilateral noise filter. It can be replaced with a 7-tap Gaussian
// with adaptive sampling by enabling the macro below. Although the
// differences are not noticeable in most cases, it may provide preferable
// results with some special usage (e.g. NPR without textureing).
// #define BLUR_HIGH_QUALITY
// By default, a fixed sampling pattern is used in the AO estimator. Although
// this gives preferable results in most cases, a completely random sampling
// pattern could give aesthetically better results. Disable the macro below
// to use such a random pattern instead of the fixed one.
#define FIX_SAMPLING_PATTERN
// The SampleNormal function normalizes samples from G-buffer because
// they're possibly unnormalized. We can eliminate this if it can be said
// that there is no wrong shader that outputs unnormalized normals.
// #define VALIDATE_NORMALS
// The constant below determines the contrast of occlusion. This allows
// users to control over/under occlusion. At the moment, this is not exposed
// to the editor because it<69>s rarely useful.
static const float kContrast = 0.6;
// The constant below controls the geometry-awareness of the bilateral
// filter. The higher value, the more sensitive it is.
static const float kGeometryCoeff = 0.8;
// The constants below are used in the AO estimator. Beta is mainly used
// for suppressing self-shadowing noise, and Epsilon is used to prevent
// calculation underflow. See the paper (Morgan 2011 http://goo.gl/2iz3P)
// for further details of these constants.
static const float kBeta = 0.002;
// --------
// System built-in variables
sampler2D _CameraGBufferTexture2;
sampler2D_float _CameraDepthTexture;
sampler2D _CameraDepthNormalsTexture;
float4 _CameraDepthTexture_ST;
// Sample count
#if !defined(SHADER_API_GLES)
int _SampleCount;
#else
// GLES2: In many cases, dynamic looping is not supported.
static const int _SampleCount = 3;
#endif
// Source texture properties
sampler2D _OcclusionTexture;
float4 _OcclusionTexture_TexelSize;
// Other parameters
half _Intensity;
float _Radius;
float _Downsample;
float3 _FogParams; // x: density, y: start, z: end
// Accessors for packed AO/normal buffer
fixed4 PackAONormal(fixed ao, fixed3 n)
{
return fixed4(ao, n * 0.5 + 0.5);
}
fixed GetPackedAO(fixed4 p)
{
return p.r;
}
fixed3 GetPackedNormal(fixed4 p)
{
return p.gba * 2.0 - 1.0;
}
// Boundary check for depth sampler
// (returns a very large value if it lies out of bounds)
float CheckBounds(float2 uv, float d)
{
float ob = any(uv < 0) + any(uv > 1);
#if defined(UNITY_REVERSED_Z)
ob += (d <= 0.00001);
#else
ob += (d >= 0.99999);
#endif
return ob * 1e8;
}
// Depth/normal sampling functions
float SampleDepth(float2 uv)
{
#if defined(SOURCE_GBUFFER) || defined(SOURCE_DEPTH)
float d = LinearizeDepth(SAMPLE_DEPTH_TEXTURE(_CameraDepthTexture, uv));
#else
float4 cdn = tex2D(_CameraDepthNormalsTexture, uv);
float d = DecodeFloatRG(cdn.zw);
#endif
return d * _ProjectionParams.z + CheckBounds(uv, d);
}
float3 SampleNormal(float2 uv)
{
#if defined(SOURCE_GBUFFER)
float3 norm = tex2D(_CameraGBufferTexture2, uv).xyz;
norm = norm * 2 - any(norm); // gets (0,0,0) when norm == 0
norm = mul((float3x3)unity_WorldToCamera, norm);
#if defined(VALIDATE_NORMALS)
norm = normalize(norm);
#endif
return norm;
#else
float4 cdn = tex2D(_CameraDepthNormalsTexture, uv);
return DecodeViewNormalStereo(cdn) * float3(1.0, 1.0, -1.0);
#endif
}
float SampleDepthNormal(float2 uv, out float3 normal)
{
#if defined(SOURCE_GBUFFER) || defined(SOURCE_DEPTH)
normal = SampleNormal(uv);
return SampleDepth(uv);
#else
float4 cdn = tex2D(_CameraDepthNormalsTexture, uv);
normal = DecodeViewNormalStereo(cdn) * float3(1.0, 1.0, -1.0);
float d = DecodeFloatRG(cdn.zw);
return d * _ProjectionParams.z + CheckBounds(uv, d);
#endif
}
// Normal vector comparer (for geometry-aware weighting)
half CompareNormal(half3 d1, half3 d2)
{
return smoothstep(kGeometryCoeff, 1.0, dot(d1, d2));
}
// Common vertex shader
struct VaryingsMultitex
{
float4 pos : SV_POSITION;
half2 uv : TEXCOORD0; // Original UV
half2 uv01 : TEXCOORD1; // Alternative UV (supports v-flip case)
half2 uvSPR : TEXCOORD2; // Single pass stereo rendering UV
};
VaryingsMultitex VertMultitex(AttributesDefault v)
{
half2 uvAlt = v.texcoord.xy;
#if UNITY_UV_STARTS_AT_TOP
if (_MainTex_TexelSize.y < 0.0) uvAlt.y = 1.0 - uvAlt.y;
#endif
VaryingsMultitex o;
o.pos = UnityObjectToClipPos(v.vertex);
o.uv = v.texcoord.xy;
o.uv01 = uvAlt;
o.uvSPR = UnityStereoTransformScreenSpaceTex(uvAlt);
return o;
}
// Trigonometric function utility
float2 CosSin(float theta)
{
float sn, cs;
sincos(theta, sn, cs);
return float2(cs, sn);
}
// Pseudo random number generator with 2D coordinates
float UVRandom(float u, float v)
{
float f = dot(float2(12.9898, 78.233), float2(u, v));
return frac(43758.5453 * sin(f));
}
// Check if the camera is perspective.
// (returns 1.0 when orthographic)
float CheckPerspective(float x)
{
return lerp(x, 1.0, unity_OrthoParams.w);
}
// Reconstruct view-space position from UV and depth.
// p11_22 = (unity_CameraProjection._11, unity_CameraProjection._22)
// p13_31 = (unity_CameraProjection._13, unity_CameraProjection._23)
float3 ReconstructViewPos(float2 uv, float depth, float2 p11_22, float2 p13_31)
{
return float3((uv * 2.0 - 1.0 - p13_31) / p11_22 * CheckPerspective(depth), depth);
}
// Sample point picker
float3 PickSamplePoint(float2 uv, float index)
{
// Uniformaly distributed points on a unit sphere http://goo.gl/X2F1Ho
#if defined(FIX_SAMPLING_PATTERN)
float gn = GradientNoise(uv * _Downsample);
// FIXME: This was added to avoid a NVIDIA driver issue.
// vvvvvvvvvvvv
float u = frac(UVRandom(0.0, index + uv.x * 1e-10) + gn) * 2.0 - 1.0;
float theta = (UVRandom(1.0, index + uv.x * 1e-10) + gn) * UNITY_PI_2;
#else
float u = UVRandom(uv.x + _Time.x, uv.y + index) * 2.0 - 1.0;
float theta = UVRandom(-uv.x - _Time.x, uv.y + index) * UNITY_PI_2;
#endif
float3 v = float3(CosSin(theta) * sqrt(1.0 - u * u), u);
// Make them distributed between [0, _Radius]
float l = sqrt((index + 1.0) / _SampleCount) * _Radius;
return v * l;
}
// Fog handling in forward
half ComputeFog(float z)
{
half fog = 0.0;
#if FOG_LINEAR
fog = (_FogParams.z - z) / (_FogParams.z - _FogParams.y);
#elif FOG_EXP
fog = exp2(-_FogParams.x * z);
#else // FOG_EXP2
fog = _FogParams.x * z;
fog = exp2(-fog * fog);
#endif
return saturate(fog);
}
float ComputeDistance(float depth)
{
float dist = depth * _ProjectionParams.z;
dist -= _ProjectionParams.y;
return dist;
}
//
// Distance-based AO estimator based on Morgan 2011 http://goo.gl/2iz3P
//
half4 FragAO(VaryingsMultitex i) : SV_Target
{
float2 uv = i.uv;
// Parameters used in coordinate conversion
float3x3 proj = (float3x3)unity_CameraProjection;
float2 p11_22 = float2(unity_CameraProjection._11, unity_CameraProjection._22);
float2 p13_31 = float2(unity_CameraProjection._13, unity_CameraProjection._23);
// View space normal and depth
float3 norm_o;
float depth_o = SampleDepthNormal(UnityStereoScreenSpaceUVAdjust(uv, _CameraDepthTexture_ST), norm_o);
#if defined(SOURCE_DEPTHNORMALS)
// Offset the depth value to avoid precision error.
// (depth in the DepthNormals mode has only 16-bit precision)
depth_o -= _ProjectionParams.z / 65536;
#endif
// Reconstruct the view-space position.
float3 vpos_o = ReconstructViewPos(i.uv01, depth_o, p11_22, p13_31);
float ao = 0.0;
for (int s = 0; s < _SampleCount; s++)
{
// Sample point
#if defined(SHADER_API_D3D11)
// This 'floor(1.0001 * s)' operation is needed to avoid a NVidia
// shader issue. This issue is only observed on DX11.
float3 v_s1 = PickSamplePoint(uv, floor(1.0001 * s));
#else
float3 v_s1 = PickSamplePoint(uv, s);
#endif
v_s1 = faceforward(v_s1, -norm_o, v_s1);
float3 vpos_s1 = vpos_o + v_s1;
// Reproject the sample point
float3 spos_s1 = mul(proj, vpos_s1);
float2 uv_s1_01 = (spos_s1.xy / CheckPerspective(vpos_s1.z) + 1.0) * 0.5;
// Depth at the sample point
float depth_s1 = SampleDepth(UnityStereoScreenSpaceUVAdjust(uv_s1_01, _CameraDepthTexture_ST));
// Relative position of the sample point
float3 vpos_s2 = ReconstructViewPos(uv_s1_01, depth_s1, p11_22, p13_31);
float3 v_s2 = vpos_s2 - vpos_o;
// Estimate the obscurance value
float a1 = max(dot(v_s2, norm_o) - kBeta * depth_o, 0.0);
float a2 = dot(v_s2, v_s2) + EPSILON;
ao += a1 / a2;
}
ao *= _Radius; // intensity normalization
// Apply other parameters.
ao = pow(ao * _Intensity / _SampleCount, kContrast);
// Apply fog when enabled (forward-only)
#if !FOG_OFF
float d = Linear01Depth(SAMPLE_DEPTH_TEXTURE(_CameraDepthTexture, uv));
d = ComputeDistance(d);
ao *= ComputeFog(d);
#endif
return PackAONormal(ao, norm_o);
}
// Geometry-aware separable bilateral filter
half4 FragBlur(VaryingsMultitex i) : SV_Target
{
#if defined(BLUR_HORIZONTAL)
// Horizontal pass: Always use 2 texels interval to match to
// the dither pattern.
float2 delta = float2(_MainTex_TexelSize.x * 2.0, 0.0);
#else
// Vertical pass: Apply _Downsample to match to the dither
// pattern in the original occlusion buffer.
float2 delta = float2(0.0, _MainTex_TexelSize.y / _Downsample * 2.0);
#endif
#if defined(BLUR_HIGH_QUALITY)
// High quality 7-tap Gaussian with adaptive sampling
fixed4 p0 = tex2D(_MainTex, i.uvSPR);
fixed4 p1a = tex2D(_MainTex, i.uvSPR - delta);
fixed4 p1b = tex2D(_MainTex, i.uvSPR + delta);
fixed4 p2a = tex2D(_MainTex, i.uvSPR - delta * 2.0);
fixed4 p2b = tex2D(_MainTex, i.uvSPR + delta * 2.0);
fixed4 p3a = tex2D(_MainTex, i.uvSPR - delta * 3.2307692308);
fixed4 p3b = tex2D(_MainTex, i.uvSPR + delta * 3.2307692308);
#if defined(BLUR_SAMPLE_CENTER_NORMAL)
fixed3 n0 = SampleNormal(i.uvSPR);
#else
fixed3 n0 = GetPackedNormal(p0);
#endif
half w0 = 0.37004405286;
half w1a = CompareNormal(n0, GetPackedNormal(p1a)) * 0.31718061674;
half w1b = CompareNormal(n0, GetPackedNormal(p1b)) * 0.31718061674;
half w2a = CompareNormal(n0, GetPackedNormal(p2a)) * 0.19823788546;
half w2b = CompareNormal(n0, GetPackedNormal(p2b)) * 0.19823788546;
half w3a = CompareNormal(n0, GetPackedNormal(p3a)) * 0.11453744493;
half w3b = CompareNormal(n0, GetPackedNormal(p3b)) * 0.11453744493;
half s;
s = GetPackedAO(p0) * w0;
s += GetPackedAO(p1a) * w1a;
s += GetPackedAO(p1b) * w1b;
s += GetPackedAO(p2a) * w2a;
s += GetPackedAO(p2b) * w2b;
s += GetPackedAO(p3a) * w3a;
s += GetPackedAO(p3b) * w3b;
s /= w0 + w1a + w1b + w2a + w2b + w3a + w3b;
#else
// Fater 5-tap Gaussian with linear sampling
fixed4 p0 = tex2D(_MainTex, i.uvSPR);
fixed4 p1a = tex2D(_MainTex, i.uvSPR - delta * 1.3846153846);
fixed4 p1b = tex2D(_MainTex, i.uvSPR + delta * 1.3846153846);
fixed4 p2a = tex2D(_MainTex, i.uvSPR - delta * 3.2307692308);
fixed4 p2b = tex2D(_MainTex, i.uvSPR + delta * 3.2307692308);
#if defined(BLUR_SAMPLE_CENTER_NORMAL)
fixed3 n0 = SampleNormal(i.uvSPR);
#else
fixed3 n0 = GetPackedNormal(p0);
#endif
half w0 = 0.2270270270;
half w1a = CompareNormal(n0, GetPackedNormal(p1a)) * 0.3162162162;
half w1b = CompareNormal(n0, GetPackedNormal(p1b)) * 0.3162162162;
half w2a = CompareNormal(n0, GetPackedNormal(p2a)) * 0.0702702703;
half w2b = CompareNormal(n0, GetPackedNormal(p2b)) * 0.0702702703;
half s;
s = GetPackedAO(p0) * w0;
s += GetPackedAO(p1a) * w1a;
s += GetPackedAO(p1b) * w1b;
s += GetPackedAO(p2a) * w2a;
s += GetPackedAO(p2b) * w2b;
s /= w0 + w1a + w1b + w2a + w2b;
#endif
return PackAONormal(s, n0);
}
// Gamma encoding (only needed in gamma lighting mode)
half EncodeAO(half x)
{
half x_g = 1.0 - max(1.055 * pow(1.0 - saturate(x), 0.416666667) - 0.055, 0.0);
// ColorSpaceLuminance.w == 0 (gamma) or 1 (linear)
return lerp(x_g, x, unity_ColorSpaceLuminance.w);
}
// Geometry-aware bilateral filter (single pass/small kernel)
half BlurSmall(sampler2D tex, float2 uv, float2 delta)
{
fixed4 p0 = tex2D(tex, uv);
fixed4 p1 = tex2D(tex, uv + float2(-delta.x, -delta.y));
fixed4 p2 = tex2D(tex, uv + float2(+delta.x, -delta.y));
fixed4 p3 = tex2D(tex, uv + float2(-delta.x, +delta.y));
fixed4 p4 = tex2D(tex, uv + float2(+delta.x, +delta.y));
fixed3 n0 = GetPackedNormal(p0);
half w0 = 1.0;
half w1 = CompareNormal(n0, GetPackedNormal(p1));
half w2 = CompareNormal(n0, GetPackedNormal(p2));
half w3 = CompareNormal(n0, GetPackedNormal(p3));
half w4 = CompareNormal(n0, GetPackedNormal(p4));
half s;
s = GetPackedAO(p0) * w0;
s += GetPackedAO(p1) * w1;
s += GetPackedAO(p2) * w2;
s += GetPackedAO(p3) * w3;
s += GetPackedAO(p4) * w4;
return s / (w0 + w1 + w2 + w3 + w4);
}
// Final composition shader
half4 FragComposition(VaryingsMultitex i) : SV_Target
{
float2 delta = _MainTex_TexelSize.xy / _Downsample;
half ao = BlurSmall(_OcclusionTexture, i.uvSPR, delta);
half4 color = tex2D(_MainTex, i.uvSPR);
#if !defined(DEBUG_COMPOSITION)
color.rgb *= 1.0 - EncodeAO(ao);
#else
color.rgb = 1.0 - EncodeAO(ao);
#endif
return color;
}
// Final composition shader (ambient-only mode)
VaryingsDefault VertCompositionGBuffer(AttributesDefault v)
{
VaryingsDefault o;
o.pos = v.vertex;
#if UNITY_UV_STARTS_AT_TOP
o.uv = v.texcoord.xy * float2(1.0, -1.0) + float2(0.0, 1.0);
#else
o.uv = v.texcoord.xy;
#endif
o.uvSPR = UnityStereoTransformScreenSpaceTex(o.uv);
return o;
}
#if !SHADER_API_GLES // excluding the MRT pass under GLES2
struct CompositionOutput
{
half4 gbuffer0 : SV_Target0;
half4 gbuffer3 : SV_Target1;
};
CompositionOutput FragCompositionGBuffer(VaryingsDefault i)
{
// Workaround: _OcclusionTexture_Texelsize hasn't been set properly
// for some reasons. Use _ScreenParams instead.
float2 delta = (_ScreenParams.zw - 1.0) / _Downsample;
half ao = BlurSmall(_OcclusionTexture, i.uvSPR, delta);
CompositionOutput o;
o.gbuffer0 = half4(0.0, 0.0, 0.0, ao);
o.gbuffer3 = half4((half3)EncodeAO(ao), 0.0);
return o;
}
#else
fixed4 FragCompositionGBuffer(VaryingsDefault i) : SV_Target0
{
return 0.0;
}
#endif
#endif // __AMBIENT_OCCLUSION__