[SM3DW] Add proper scaling of ambient occlusion

This commit is contained in:
Crementif 2020-01-04 02:31:19 +01:00
parent 68534a6588
commit bf7821ac77
No known key found for this signature in database
GPG Key ID: 11A98E7DE0412B33

View File

@ -0,0 +1,292 @@
#version 430
#extension GL_ARB_texture_gather : enable
#extension GL_ARB_separate_shader_objects : enable
// shader d388f32cb9be9a7a
// Used for: Ambient-occlusion
float resScale = float($width)/float($gameWidth);
#ifdef VULKAN
#define ATTR_LAYOUT(__vkSet, __location) layout(set = __vkSet, location = __location)
#define UNIFORM_BUFFER_LAYOUT(__glLocation, __vkSet, __vkLocation) layout(set = __vkSet, binding = __vkLocation, std140)
#define TEXTURE_LAYOUT(__glLocation, __vkSet, __vkLocation) layout(set = __vkSet, binding = __vkLocation)
#define GET_FRAGCOORD() vec4(gl_FragCoord.xy*uf_fragCoordScale.xy,gl_FragCoord.zw)
#else
#define ATTR_LAYOUT(__vkSet, __location) layout(location = __location)
#define UNIFORM_BUFFER_LAYOUT(__glLocation, __vkSet, __vkLocation) layout(binding = __glLocation, std140)
#define TEXTURE_LAYOUT(__glLocation, __vkSet, __vkLocation) layout(binding = __glLocation)
#define GET_FRAGCOORD() vec4(gl_FragCoord.xy*uf_fragCoordScale,gl_FragCoord.zw)
#endif
#ifdef VULKAN
layout(set = 1, binding = 4) uniform ufBlock
{
uniform ivec4 uf_remappedPS[8];
uniform vec4 uf_fragCoordScale;
};
#else
uniform ivec4 uf_remappedPS[8];
uniform vec2 uf_fragCoordScale;
#endif
TEXTURE_LAYOUT(0, 1, 0) uniform sampler2D textureUnitPS0;
TEXTURE_LAYOUT(1, 1, 1) uniform sampler2D textureUnitPS1;
TEXTURE_LAYOUT(2, 1, 2) uniform sampler2D textureUnitPS2;
TEXTURE_LAYOUT(3, 1, 3) uniform sampler2D textureUnitPS3;
layout(location = 0) in vec4 passParameterSem0;
layout(location = 1) in vec4 passParameterSem1;
layout(location = 0) out vec4 passPixelColor0;
int clampFI32(int v)
{
if( v == 0x7FFFFFFF )
return floatBitsToInt(1.0);
else if( v == 0xFFFFFFFF )
return floatBitsToInt(0.0);
return floatBitsToInt(clamp(intBitsToFloat(v), 0.0, 1.0));
}
float mul_nonIEEE(float a, float b){ return min(a*b,min(abs(a)*3.40282347E+38F,abs(b)*3.40282347E+38F)); }
void main()
{
vec4 R0f = vec4(0.0);
vec4 R1f = vec4(0.0);
vec4 R2f = vec4(0.0);
vec4 R3f = vec4(0.0);
vec4 R4f = vec4(0.0);
vec4 R5f = vec4(0.0);
vec4 R6f = vec4(0.0);
vec4 R7f = vec4(0.0);
vec4 R122f = vec4(0.0);
vec4 R123f = vec4(0.0);
vec4 R124f = vec4(0.0);
vec4 R125f = vec4(0.0);
vec4 R126f = vec4(0.0);
vec4 R127f = vec4(0.0);
float backupReg0f, backupReg1f, backupReg2f, backupReg3f, backupReg4f;
vec4 PV0f = vec4(0.0), PV1f = vec4(0.0);
float PS0f = 0.0, PS1f = 0.0;
vec4 tempf = vec4(0.0);
float tempResultf;
int tempResulti;
ivec4 ARi = ivec4(0);
bool predResult = true;
vec3 cubeMapSTM;
int cubeMapFaceId;
R0f = passParameterSem0;
R1f = passParameterSem1;
R7f.x = (texture(textureUnitPS0, R1f.xy).x);
R2f.xy = (texture(textureUnitPS3, R1f.zw).xy);
// 0
R5f.x = (R2f.y * 2.0 + -(1.0));
PV0f.x = R5f.x;
R4f.y = R7f.x * intBitsToFloat(uf_remappedPS[0].z);
R123f.z = (-(R7f.x) * intBitsToFloat(uf_remappedPS[1].y) + 1.0);
R123f.z = clamp(R123f.z, 0.0, 1.0);
PV0f.z = R123f.z;
R0f.w = (R2f.x * 2.0 + -(1.0));
PV0f.w = R0f.w;
// 1
R4f.x = PV0f.w * intBitsToFloat(uf_remappedPS[2].z)/resScale;
R125f.y = PV0f.z * intBitsToFloat(uf_remappedPS[1].x)/resScale;
PV1f.y = R125f.y;
R4f.z = -(PV0f.x) * intBitsToFloat(uf_remappedPS[2].z)/resScale;
// 2
R127f.x = PV1f.y * intBitsToFloat(uf_remappedPS[3].x)/resScale;
PV0f.x = R127f.x;
R127f.y = PV1f.y * intBitsToFloat(uf_remappedPS[3].y)/resScale;
R127f.z = PV1f.y * intBitsToFloat(uf_remappedPS[4].y)/resScale;
PV0f.w = PV1f.y * intBitsToFloat(uf_remappedPS[4].x)/resScale;
PS0f = 1.0 / PV1f.y;
// 3
PV1f.x = R5f.x * PV0f.w;
PV1f.y = R4f.x * PV0f.w;
R5f.z = intBitsToFloat(uf_remappedPS[5].z) * PS0f;
R5f.w = intBitsToFloat(uf_remappedPS[5].z) * PS0f;
PS1f = R4f.x * PV0f.x;
// 4
PV0f.x = R5f.x * R127f.x;
R126f.y = (R4f.z * R127f.y + PS1f);
R123f.z = (R0f.w * R127f.z + PV1f.x);
PV0f.z = R123f.z;
R123f.w = (R4f.z * R127f.z + PV1f.y);
PV0f.w = R123f.w;
R6f.x = R125f.y * intBitsToFloat(uf_remappedPS[6].x)/resScale;
PS0f = R6f.x;
// 5
R2f.xyz = vec3(R1f.x,R1f.y,R1f.x) + vec3(PV0f.w,PV0f.z,-(PV0f.w));
R2f.w = R1f.y + -(PV0f.z);
R122f.x = (R0f.w * R127f.y + PV0f.x);
PS1f = R122f.x;
// 6
R3f.xyz = vec3(R1f.x,R1f.y,R1f.x) + vec3(R126f.y,PS1f,-(R126f.y));
R3f.w = R1f.y + -(PS1f);
R4f.w = R125f.y * intBitsToFloat(uf_remappedPS[6].y)/resScale;
PS0f = R4f.w;
R2f.x = (texture(textureUnitPS1, R2f.xy).x);
R2f.y = (texture(textureUnitPS1, R2f.zw).x);
R3f.x = (texture(textureUnitPS1, R3f.xy).x);
R3f.y = (texture(textureUnitPS1, R3f.zw).x);
// 0
R127f.x = R4f.y + -(R3f.x);
PV0f.x = R127f.x;
PV0f.y = -(R2f.y) + R4f.y;
PV0f.z = -(R2f.x) + R4f.y;
R127f.w = R4f.y + -(R3f.y);
// 1
PV1f.x = PV0f.z * R5f.w;
PV1f.x /= 2.0;
R125f.y = (-(PV0f.z) * intBitsToFloat(uf_remappedPS[1].z) + 1.0);
R125f.y = clamp(R125f.y, 0.0, 1.0);
R126f.z = (-(PV0f.y) * intBitsToFloat(uf_remappedPS[1].z) + 1.0);
R126f.z = clamp(R126f.z, 0.0, 1.0);
PV1f.w = PV0f.y * R5f.z;
PV1f.w /= 2.0;
PS1f = PV0f.x * R5f.w;
PS1f /= 2.0;
// 2
PV0f.x = R127f.w * R5f.z;
PV0f.x /= 2.0;
R126f.y = (PV1f.w * intBitsToFloat(uf_remappedPS[4].z) + 0.5);
R126f.y = clamp(R126f.y, 0.0, 1.0);
PV0f.y = R126f.y;
R127f.z = (PV1f.x * intBitsToFloat(uf_remappedPS[4].z) + 0.5);
R127f.z = clamp(R127f.z, 0.0, 1.0);
PV0f.z = R127f.z;
R125f.w = (PS1f * intBitsToFloat(uf_remappedPS[3].z) + 0.5);
R125f.w = clamp(R125f.w, 0.0, 1.0);
PS0f = R4f.x * R6f.x;
// 3
R126f.x = (PV0f.x * intBitsToFloat(uf_remappedPS[3].z) + 0.5);
R126f.x = clamp(R126f.x, 0.0, 1.0);
PV1f.x = R126f.x;
PV1f.y = 0.5 + -(PV0f.z);
PV1f.z = 0.5 + -(PV0f.y);
R124f.w = (R4f.z * R4f.w + PS0f);
R125f.z = (-(R127f.x) * intBitsToFloat(uf_remappedPS[1].z) + 1.0);
R125f.z = clamp(R125f.z, 0.0, 1.0);
PS1f = R125f.z;
// 4
R127f.x = (PV1f.z * R126f.z + 0.5);
PV0f.x = R127f.x;
R127f.y = (-(R127f.w) * intBitsToFloat(uf_remappedPS[1].z) + 1.0);
R127f.y = clamp(R127f.y, 0.0, 1.0);
PV0f.y = R127f.y;
PV0f.z = 0.5 + -(PV1f.x);
R127f.w = (PV1f.y * R125f.y + 0.5);
PV0f.w = R127f.w;
PS0f = 0.5 + -(R125f.w);
// 5
R125f.x = (PV0f.z * PV0f.y + 0.5);
PV1f.y = R126f.y + -(PV0f.w);
PV1f.z = R127f.z + -(PV0f.x);
R126f.w = (PS0f * R125f.z + 0.5);
PV1f.w = R126f.w;
PS1f = R5f.x * R6f.x;
// 6
backupReg0f = R127f.x;
backupReg1f = R126f.z;
backupReg2f = R127f.w;
R127f.x = (PV1f.z * R125f.y + backupReg0f);
PV0f.y = R126f.x + -(PV1f.w);
R126f.z = (R0f.w * R4f.w + PS1f);
PV0f.z = R126f.z;
R127f.w = (PV1f.y * backupReg1f + backupReg2f);
// 7
PV1f.x = R125f.w + -(R125f.x);
R123f.y = (PV0f.y * R127f.y + R126f.w);
PV1f.y = R123f.y;
R4f.z = R1f.x + R124f.w;
R4f.w = R1f.y + PV0f.z;
R5f.x = R1f.x + -(R124f.w);
PS1f = R5f.x;
// 8
backupReg0f = R125f.x;
R125f.x = R7f.x * intBitsToFloat(uf_remappedPS[5].y);
R5f.y = R1f.y + -(R126f.z);
R123f.z = (PV1f.x * R125f.z + backupReg0f);
PV0f.z = R123f.z;
PV0f.w = PV1f.y * intBitsToFloat(uf_remappedPS[3].w);
R6f.z = 0.0;
PS0f = R6f.z;
// 9
PV1f.x = PV0f.z * intBitsToFloat(uf_remappedPS[3].w);
R1f.y = (R127f.w * intBitsToFloat(uf_remappedPS[4].w) + PV0f.w);
R6f.w = 1.0;
// 10
R1f.z = (R127f.x * intBitsToFloat(uf_remappedPS[4].w) + PV1f.x);
// 11
R123f.z = (R7f.x * intBitsToFloat(uf_remappedPS[5].z) + intBitsToFloat(uf_remappedPS[5].x)/3);
PV1f.z = R123f.z;
// 12
backupReg0f = R0f.x;
backupReg1f = R0f.z;
PV0f.x = backupReg0f * PV1f.z;
PV0f.z = backupReg1f * PV1f.z;
PV0f.w = R0f.y * PV1f.z;
// 13
backupReg0f = R125f.x;
backupReg0f = R125f.x;
backupReg0f = R125f.x;
R125f.x = (backupReg0f * intBitsToFloat(uf_remappedPS[7].y) + PV0f.w);
R127f.y = (backupReg0f * intBitsToFloat(uf_remappedPS[7].x) + PV0f.x);
R123f.w = (backupReg0f * intBitsToFloat(uf_remappedPS[7].z) + PV0f.z);
PV1f.w = R123f.w;
// 14
PS0f = 1.0 / PV1f.w;
// 15
R0f.x = R127f.y * PS0f;
R0f.y = R125f.x * PS0f;
R7f.x = (texture(textureUnitPS1, R4f.zw).x);
R5f.y = (texture(textureUnitPS1, R5f.xy).x);
R6f.y = (texture(textureUnitPS2, R0f.xy).x);
// 0
PV0f.x = R4f.y + -(R7f.x);
PV0f.w = R4f.y + -(R5f.y);
// 1
R127f.x = (-(PV0f.x) * intBitsToFloat(uf_remappedPS[1].z) + 1.0);
R127f.x = clamp(R127f.x, 0.0, 1.0);
PV1f.y = PV0f.w * R5f.z;
PV1f.y /= 2.0;
PV1f.z = PV0f.x * R5f.w;
PV1f.z /= 2.0;
R126f.w = (-(PV0f.w) * intBitsToFloat(uf_remappedPS[1].z) + 1.0);
R126f.w = clamp(R126f.w, 0.0, 1.0);
// 2
R125f.x = (PV1f.z * intBitsToFloat(uf_remappedPS[6].z) + 0.5);
R125f.x = clamp(R125f.x, 0.0, 1.0);
PV0f.x = R125f.x;
R127f.w = (PV1f.y * intBitsToFloat(uf_remappedPS[6].z) + 0.5);
R127f.w = clamp(R127f.w, 0.0, 1.0);
PV0f.w = R127f.w;
// 3
PV1f.x = 0.5 + -(PV0f.w);
PV1f.w = 0.5 + -(PV0f.x);
// 4
R127f.y = (PV1f.w * R127f.x + 0.5);
PV0f.y = R127f.y;
R126f.z = (PV1f.x * R126f.w + 0.5);
PV0f.z = R126f.z;
// 5
PV1f.x = R125f.x + -(PV0f.z);
PV1f.w = R127f.w + -(PV0f.y);
// 6
R123f.y = (PV1f.w * R126f.w + R127f.y);
PV0f.y = R123f.y;
R123f.z = (PV1f.x * R127f.x + R126f.z);
PV0f.z = R123f.z;
// 7
R123f.x = (PV0f.z * intBitsToFloat(uf_remappedPS[6].w) + R1f.z);
PV1f.x = R123f.x;
R123f.w = (PV0f.y * intBitsToFloat(uf_remappedPS[6].w) + R1f.y);
PV1f.w = R123f.w;
// 8
PV0f.z = PV1f.x + PV1f.w;
// 9
PV1f.y = PV0f.z + intBitsToFloat(uf_remappedPS[1].w);
// 10
PV0f.x = PV1f.y + -(0.5);
// 11
PV1f.w = PV0f.x * intBitsToFloat(uf_remappedPS[0].y);
PV1f.w = clamp(PV1f.w, 0.0, 1.0);
// 12
R6f.x = -(PV1f.w) + 1.0;
// export
passPixelColor0 = vec4(R6f.x, R6f.y, R6f.z, R6f.w);
}