First Commit

This commit is contained in:
2025-11-18 14:18:26 -07:00
parent 33eb6e3707
commit 27277ec342
6106 changed files with 3571167 additions and 0 deletions

View File

@@ -0,0 +1,71 @@
#version 460 core
#extension GL_EXT_samplerless_texture_functions : require
// Based on CAS_Shader.glsl
// Copyright(c) 2019 Advanced Micro Devices, Inc.All rights reserved.
// 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.
layout(push_constant) uniform const_buffer
{
uvec4 const0;
uvec4 const1;
ivec2 srcOffset;
};
layout(set=0, binding=0) uniform texture2D imgSrc;
layout(set=0, binding=1, rgba8) uniform writeonly image2D imgDst;
layout(constant_id=0) const int sharpenOnly = 0;
#define A_GPU 1
#define A_GLSL 1
#include "ffx_a.h"
AF3 CasLoad(ASU2 p)
{
return texelFetch(imgSrc, srcOffset + ivec2(p), 0).rgb;
}
// Lets you transform input from the load into a linear color space between 0 and 1. See ffx_cas.h
// In this case, our input is already linear and between 0 and 1
void CasInput(inout AF1 r, inout AF1 g, inout AF1 b) {}
#include "ffx_cas.h"
layout(local_size_x=64) in;
void main()
{
// Do remapping of local xy in workgroup for a more PS-like swizzle pattern.
AU2 gxy = ARmp8x8(gl_LocalInvocationID.x)+AU2(gl_WorkGroupID.x<<4u,gl_WorkGroupID.y<<4u);
// Filter.
AF4 c = vec4(0.0f);
CasFilter(c.r, c.g, c.b, gxy, const0, const1, sharpenOnly != 0);
imageStore(imgDst, ASU2(gxy), c);
gxy.x += 8u;
CasFilter(c.r, c.g, c.b, gxy, const0, const1, sharpenOnly != 0);
imageStore(imgDst, ASU2(gxy), c);
gxy.y += 8u;
CasFilter(c.r, c.g, c.b, gxy, const0, const1, sharpenOnly != 0);
imageStore(imgDst, ASU2(gxy), c);
gxy.x -= 8u;
CasFilter(c.r, c.g, c.b, gxy, const0, const1, sharpenOnly != 0);
imageStore(imgDst, ASU2(gxy), c);
}

View File

@@ -0,0 +1,618 @@
// SPDX-FileCopyrightText: 2002-2025 PCSX2 Dev Team
// SPDX-License-Identifier: GPL-3.0+
#ifdef VERTEX_SHADER
layout(location = 0) in vec4 a_pos;
layout(location = 1) in vec2 a_tex;
layout(location = 0) out vec2 v_tex;
void main()
{
gl_Position = vec4(a_pos.x, -a_pos.y, a_pos.z, a_pos.w);
v_tex = a_tex;
}
#endif
#ifdef FRAGMENT_SHADER
layout(location = 0) in vec2 v_tex;
#if defined(ps_convert_rgba8_16bits) || defined(ps_convert_float32_32bits)
layout(location = 0) out uint o_col0;
#elif !defined(ps_datm1) && \
!defined(ps_datm0) && \
!defined(ps_datm1_rta_correction) && \
!defined(ps_datm0_rta_correction) && \
!defined(ps_convert_rgba8_float32) && \
!defined(ps_convert_rgba8_float24) && \
!defined(ps_convert_rgba8_float16) && \
!defined(ps_convert_rgb5a1_float16) && \
!defined(ps_convert_rgba8_float32_biln) && \
!defined(ps_convert_rgba8_float24_biln) && \
!defined(ps_convert_rgba8_float16_biln) && \
!defined(ps_convert_rgb5a1_float16_biln) && \
!defined(ps_depth_copy)
layout(location = 0) out vec4 o_col0;
#endif
layout(set = 0, binding = 0) uniform sampler2D samp0;
vec4 sample_c(vec2 uv)
{
return texture(samp0, uv);
}
#ifdef ps_copy
void ps_copy()
{
o_col0 = sample_c(v_tex);
}
#endif
#ifdef ps_depth_copy
void ps_depth_copy()
{
gl_FragDepth = sample_c(v_tex).r;
}
#endif
#ifdef ps_downsample_copy
layout(push_constant) uniform cb10
{
ivec2 ClampMin;
int DownsampleFactor;
int pad0;
float Weight;
vec3 pad1;
};
void ps_downsample_copy()
{
ivec2 coord = max(ivec2(gl_FragCoord.xy) * DownsampleFactor, ClampMin);
vec4 result = vec4(0);
for (int yoff = 0; yoff < DownsampleFactor; yoff++)
{
for (int xoff = 0; xoff < DownsampleFactor; xoff++)
result += texelFetch(samp0, coord + ivec2(xoff, yoff), 0);
}
o_col0 = result / Weight;
}
#endif
#ifdef ps_filter_transparency
void ps_filter_transparency()
{
vec4 c = sample_c(v_tex);
o_col0 = vec4(c.rgb, 1.0);
}
#endif
#ifdef ps_convert_rgba8_16bits
// Need to be careful with precision here, it can break games like Spider-Man 3 and Dogs Life
void ps_convert_rgba8_16bits()
{
uvec4 i = uvec4(sample_c(v_tex) * vec4(255.5f, 255.5f, 255.5f, 255.5f));
o_col0 = ((i.x & 0x00F8u) >> 3) | ((i.y & 0x00F8u) << 2) | ((i.z & 0x00f8u) << 7) | ((i.w & 0x80u) << 8);
}
#endif
#ifdef ps_datm1
void ps_datm1()
{
if(sample_c(v_tex).a < (127.5f / 255.0f)) // >= 0x80 pass
discard;
}
#endif
#ifdef ps_datm0
void ps_datm0()
{
if((127.5f / 255.0f) < sample_c(v_tex).a) // < 0x80 pass (== 0x80 should not pass)
discard;
}
#endif
#ifdef ps_datm1_rta_correction
void ps_datm1_rta_correction()
{
if(sample_c(v_tex).a < (254.5f / 255.0f)) // >= 0x80 pass
discard;
}
#endif
#ifdef ps_datm0_rta_correction
void ps_datm0_rta_correction()
{
if((254.5f / 255.0f) < sample_c(v_tex).a) // < 0x80 pass (== 0x80 should not pass)
discard;
}
#endif
#ifdef ps_rta_correction
void ps_rta_correction()
{
vec4 value = sample_c(v_tex);
o_col0 = vec4(value.rgb, value.a / (128.25f / 255.0f));
}
#endif
#ifdef ps_rta_decorrection
void ps_rta_decorrection()
{
vec4 value = sample_c(v_tex);
o_col0 = vec4(value.rgb, value.a * (128.25f / 255.0f));
}
#endif
#ifdef ps_colclip_init
void ps_colclip_init()
{
vec4 value = sample_c(v_tex);
o_col0 = vec4(roundEven(value.rgb * 255.0f) / 65535.0f, value.a);
}
#endif
#ifdef ps_colclip_resolve
void ps_colclip_resolve()
{
vec4 value = sample_c(v_tex);
o_col0 = vec4(vec3(uvec3(value.rgb * 65535.5f) & 255u) / 255.0f, value.a);
}
#endif
#ifdef ps_convert_float32_32bits
void ps_convert_float32_32bits()
{
// Convert a vec32 depth texture into a 32 bits UINT texture
o_col0 = uint(exp2(32.0f) * sample_c(v_tex).r);
}
#endif
#ifdef ps_convert_float32_rgba8
void ps_convert_float32_rgba8()
{
// Convert a vec32 depth texture into a RGBA color texture
uint d = uint(sample_c(v_tex).r * exp2(32.0f));
o_col0 = vec4(uvec4((d & 0xFFu), ((d >> 8) & 0xFFu), ((d >> 16) & 0xFFu), (d >> 24))) / vec4(255.0);
}
#endif
#ifdef ps_convert_float16_rgb5a1
void ps_convert_float16_rgb5a1()
{
// Convert a vec32 (only 16 lsb) depth into a RGB5A1 color texture
uint d = uint(sample_c(v_tex).r * exp2(32.0f));
o_col0 = vec4(uvec4(d << 3, d >> 2, d >> 7, d >> 8) & uvec4(0xf8, 0xf8, 0xf8, 0x80)) / 255.0f;
}
#endif
float rgba8_to_depth32(vec4 unorm)
{
uvec4 c = uvec4(unorm * vec4(255.5f));
return float(c.r | (c.g << 8) | (c.b << 16) | (c.a << 24)) * exp2(-32.0f);
}
float rgba8_to_depth24(vec4 unorm)
{
uvec3 c = uvec3(unorm.rgb * vec3(255.5f));
return float(c.r | (c.g << 8) | (c.b << 16)) * exp2(-32.0f);
}
float rgba8_to_depth16(vec4 unorm)
{
uvec2 c = uvec2(unorm.rg * vec2(255.5f));
return float(c.r | (c.g << 8)) * exp2(-32.0f);
}
float rgb5a1_to_depth16(vec4 unorm)
{
uvec4 c = uvec4(unorm * vec4(255.5f));
return float(((c.r & 0xF8u) >> 3) | ((c.g & 0xF8u) << 2) | ((c.b & 0xF8u) << 7) | ((c.a & 0x80u) << 8)) * exp2(-32.0f);
}
#ifdef ps_convert_float32_float24
void ps_convert_float32_float24()
{
// Truncates depth value to 24bits
uint d = uint(sample_c(v_tex).r * exp2(32.0f)) & 0xFFFFFFu;
gl_FragDepth = float(d) * exp2(-32.0f);
}
#endif
#ifdef ps_convert_rgba8_float32
void ps_convert_rgba8_float32()
{
// Convert an RGBA texture into a float depth texture
gl_FragDepth = rgba8_to_depth32(sample_c(v_tex));
}
#endif
#ifdef ps_convert_rgba8_float24
void ps_convert_rgba8_float24()
{
// Same as above but without the alpha channel (24 bits Z)
// Convert an RGBA texture into a float depth texture
gl_FragDepth = rgba8_to_depth24(sample_c(v_tex));
}
#endif
#ifdef ps_convert_rgba8_float16
void ps_convert_rgba8_float16()
{
// Same as above but without the A/B channels (16 bits Z)
// Convert an RGBA texture into a float depth texture
gl_FragDepth = rgba8_to_depth16(sample_c(v_tex));
}
#endif
#ifdef ps_convert_rgb5a1_float16
void ps_convert_rgb5a1_float16()
{
// Convert an RGB5A1 (saved as RGBA8) color to a 16 bit Z
gl_FragDepth = rgb5a1_to_depth16(sample_c(v_tex));
}
#endif
#define SAMPLE_RGBA_DEPTH_BILN(CONVERT_FN) \
ivec2 dims = textureSize(samp0, 0); \
vec2 top_left_f = v_tex * vec2(dims) - 0.5f; \
ivec2 top_left = ivec2(floor(top_left_f)); \
ivec4 coords = clamp(ivec4(top_left, top_left + 1), ivec4(0), dims.xyxy - 1); \
vec2 mix_vals = fract(top_left_f); \
float depthTL = CONVERT_FN(texelFetch(samp0, coords.xy, 0)); \
float depthTR = CONVERT_FN(texelFetch(samp0, coords.zy, 0)); \
float depthBL = CONVERT_FN(texelFetch(samp0, coords.xw, 0)); \
float depthBR = CONVERT_FN(texelFetch(samp0, coords.zw, 0)); \
gl_FragDepth = mix(mix(depthTL, depthTR, mix_vals.x), mix(depthBL, depthBR, mix_vals.x), mix_vals.y);
#ifdef ps_convert_rgba8_float32_biln
void ps_convert_rgba8_float32_biln()
{
// Convert an RGBA texture into a float depth texture
SAMPLE_RGBA_DEPTH_BILN(rgba8_to_depth32);
}
#endif
#ifdef ps_convert_rgba8_float24_biln
void ps_convert_rgba8_float24_biln()
{
// Same as above but without the alpha channel (24 bits Z)
// Convert an RGBA texture into a float depth texture
SAMPLE_RGBA_DEPTH_BILN(rgba8_to_depth24);
}
#endif
#ifdef ps_convert_rgba8_float16_biln
void ps_convert_rgba8_float16_biln()
{
// Same as above but without the A/B channels (16 bits Z)
// Convert an RGBA texture into a float depth texture
SAMPLE_RGBA_DEPTH_BILN(rgba8_to_depth16);
}
#endif
#ifdef ps_convert_rgb5a1_float16_biln
void ps_convert_rgb5a1_float16_biln()
{
// Convert an RGB5A1 (saved as RGBA8) color to a 16 bit Z
SAMPLE_RGBA_DEPTH_BILN(rgb5a1_to_depth16);
}
#endif
#ifdef ps_convert_rgb5a1_8i
layout(push_constant) uniform cb10
{
uint SBW;
uint DBW;
uint PSM;
float cb_pad1;
float ScaleFactor;
vec3 cb_pad2;
};
void ps_convert_rgb5a1_8i()
{
// Convert a RGB5A1 texture into a 8 bits packed texture
// Input column: 16x2 RGB5A1 pixels
// 0: 16 RGBA
// 1: 16 RGBA
// Output column: 16x4 Index pixels
// 0: 16 R5G2
// 1: 16 R5G2
// 2: 16 G2B5A1
// 3: 16 G2B5A1
uvec2 pos = uvec2(gl_FragCoord.xy);
// Collapse separate R G B A areas into their base pixel
uvec2 column = (pos & ~uvec2(0u, 3u)) / uvec2(1,2);
uvec2 subcolumn = (pos & uvec2(0u, 1u));
column.x -= (column.x / 128) * 64;
column.y += (column.y / 32) * 32;
// Deal with swizzling differences
if ((PSM & 0x8) != 0) // PSMCT16S
{
if ((pos.x & 32) != 0)
{
column.y += 32; // 4 columns high times 4 to get bottom 4 blocks
column.x &= ~32;
}
if ((pos.x & 64) != 0)
{
column.x -= 32;
}
if (((pos.x & 16) != 0) != ((pos.y & 16) != 0))
{
column.x ^= 16;
column.y ^= 8;
}
if ((PSM & 0x30) != 0) // PSMZ16S - Untested but hopefully ok if anything uses it.
{
column.x ^= 32;
column.y ^= 16;
}
}
else // PSMCT16
{
if ((pos.y & 32) != 0)
{
column.y -= 16;
column.x += 32;
}
if ((pos.x & 96) != 0)
{
uint multi = (pos.x & 96) / 32;
column.y += 16 * multi; // 4 columns high times 4 to get bottom 4 blocks
column.x -= (pos.x & 96);
}
if (((pos.x & 16) != 0) != ((pos.y & 16) != 0))
{
column.x ^= 16;
column.y ^= 8;
}
if ((PSM & 0x30) != 0) // PSMZ16 - Untested but hopefully ok if anything uses it.
{
column.x ^= 32;
column.y ^= 32;
}
}
uvec2 coord = column | subcolumn;
// Compensate for potentially differing page pitch.
uvec2 block_xy = coord / uvec2(64u, 64u);
uint block_num = (block_xy.y * (DBW / 128u)) + block_xy.x;
uvec2 block_offset = uvec2((block_num % (SBW / 64u)) * 64u, (block_num / (SBW / 64u)) * 64u);
coord = (coord % uvec2(64u, 64u)) + block_offset;
// Apply offset to cols 1 and 2
uint is_col23 = pos.y & 4u;
uint is_col13 = pos.y & 2u;
uint is_col12 = is_col23 ^ (is_col13 << 1);
coord.x ^= is_col12; // If cols 1 or 2, flip bit 3 of x
if (floor(ScaleFactor) != ScaleFactor)
coord = uvec2(vec2(coord) * ScaleFactor);
else
coord *= uvec2(ScaleFactor);
vec4 pixel = texelFetch(samp0, ivec2(coord), 0);
uvec4 denorm_c = uvec4(pixel * 255.5f);
if ((pos.y & 2u) == 0u)
{
uint red = (denorm_c.r >> 3) & 0x1Fu;
uint green = (denorm_c.g >> 3) & 0x1Fu;
float sel0 = float(((green << 5) | red) & 0xFF) / 255.0f;
o_col0 = vec4(sel0);
}
else
{
uint green = (denorm_c.g >> 3) & 0x1Fu;
uint blue = (denorm_c.b >> 3) & 0x1Fu;
uint alpha = denorm_c.a & 0x80u;
float sel0 = float((alpha | (blue << 2) | (green >> 3)) & 0xFF) / 255.0f;
o_col0 = vec4(sel0);
}
}
#endif
#ifdef ps_convert_rgba_8i
layout(push_constant) uniform cb10
{
uint SBW;
uint DBW;
uint PSM;
float cb_pad1;
float ScaleFactor;
vec3 cb_pad2;
};
void ps_convert_rgba_8i()
{
// Convert a RGBA texture into a 8 bits packed texture
// Input column: 8x2 RGBA pixels
// 0: 8 RGBA
// 1: 8 RGBA
// Output column: 16x4 Index pixels
// 0: 8 R | 8 B
// 1: 8 R | 8 B
// 2: 8 G | 8 A
// 3: 8 G | 8 A
uvec2 pos = uvec2(gl_FragCoord.xy);
// Collapse separate R G B A areas into their base pixel
uvec2 block = (pos & ~uvec2(15u, 3u)) >> 1;
uvec2 subblock = pos & uvec2(7u, 1u);
uvec2 coord = block | subblock;
// Compensate for potentially differing page pitch.
uvec2 block_xy = coord / uvec2(64u, 32u);
uint block_num = (block_xy.y * (DBW / 128u)) + block_xy.x;
uvec2 block_offset = uvec2((block_num % (SBW / 64u)) * 64u, (block_num / (SBW / 64u)) * 32u);
coord = (coord % uvec2(64u, 32u)) + block_offset;
// Apply offset to cols 1 and 2
uint is_col23 = pos.y & 4u;
uint is_col13 = pos.y & 2u;
uint is_col12 = is_col23 ^ (is_col13 << 1);
coord.x ^= is_col12; // If cols 1 or 2, flip bit 3 of x
if (floor(ScaleFactor) != ScaleFactor)
coord = uvec2(vec2(coord) * ScaleFactor);
else
coord *= uvec2(ScaleFactor);
vec4 pixel = texelFetch(samp0, ivec2(coord), 0);
vec2 sel0 = (pos.y & 2u) == 0u ? pixel.rb : pixel.ga;
float sel1 = (pos.x & 8u) == 0u ? sel0.x : sel0.y;
o_col0 = vec4(sel1); // Divide by something here?
}
#endif
#ifdef ps_convert_clut_4
layout(push_constant) uniform cb10
{
uvec2 offset;
uint doffset;
uint cb_pad1;
float scale;
vec3 cb_pad2;
};
void ps_convert_clut_4()
{
// CLUT4 is easy, just two rows of 8x8.
uint index = uint(gl_FragCoord.x) + doffset;
uvec2 pos = uvec2(index % 8u, index / 8u);
ivec2 final = ivec2(floor(vec2(offset + pos) * vec2(scale)));
o_col0 = texelFetch(samp0, final, 0);
}
#endif
#ifdef ps_convert_clut_8
layout(push_constant) uniform cb10
{
uvec2 offset;
uint doffset;
uint cb_pad1;
float scale;
vec3 cb_pad2;
};
void ps_convert_clut_8()
{
uint index = min(uint(gl_FragCoord.x) + doffset, 255u);
// CLUT is arranged into 8 groups of 16x2, with the top-right and bottom-left quadrants swapped.
// This can probably be done better..
uint subgroup = (index / 8u) % 4u;
uvec2 pos;
pos.x = (index % 8u) + ((subgroup >= 2u) ? 8u : 0u);
pos.y = ((index / 32u) * 2u) + (subgroup % 2u);
ivec2 final = ivec2(floor(vec2(offset + pos) * vec2(scale)));
o_col0 = texelFetch(samp0, final, 0);
}
#endif
#ifdef ps_yuv
layout(push_constant) uniform cb10
{
int EMODA;
int EMODC;
};
void ps_yuv()
{
vec4 i = sample_c(v_tex);
vec4 o = vec4(0.0f);
mat3 rgb2yuv;
rgb2yuv[0] = vec3(0.587, -0.311, -0.419);
rgb2yuv[1] = vec3(0.114, 0.500, -0.081);
rgb2yuv[2] = vec3(0.299, -0.169, 0.500);
vec3 yuv = rgb2yuv * i.gbr;
float Y = float(0xDB)/255.0f * yuv.x + float(0x10)/255.0f;
float Cr = float(0xE0)/255.0f * yuv.y + float(0x80)/255.0f;
float Cb = float(0xE0)/255.0f * yuv.z + float(0x80)/255.0f;
switch(EMODA)
{
case 0:
o.a = i.a;
break;
case 1:
o.a = Y;
break;
case 2:
o.a = Y/2.0f;
break;
case 3:
o.a = 0.0f;
break;
}
switch(EMODC)
{
case 0:
o.rgb = i.rgb;
break;
case 1:
o.rgb = vec3(Y);
break;
case 2:
o.rgb = vec3(Y, Cb, Cr);
break;
case 3:
o.rgb = vec3(i.a);
break;
}
o_col0 = o;
}
#endif
#if defined(ps_stencil_image_init_0) || defined(ps_stencil_image_init_1) || defined(ps_stencil_image_init_2) || defined(ps_stencil_image_init_3)
void main()
{
o_col0 = vec4(0x7FFFFFFF);
#ifdef ps_stencil_image_init_0
if((127.5f / 255.0f) < sample_c(v_tex).a) // < 0x80 pass (== 0x80 should not pass)
o_col0 = vec4(-1);
#endif
#ifdef ps_stencil_image_init_1
if(sample_c(v_tex).a < (127.5f / 255.0f)) // >= 0x80 pass
o_col0 = vec4(-1);
#endif
#ifdef ps_stencil_image_init_2
if((254.5f / 255.0f) < sample_c(v_tex).a) // < 0x80 pass (== 0x80 should not pass)
o_col0 = vec4(-1);
#endif
#ifdef ps_stencil_image_init_3
if(sample_c(v_tex).a < (254.5f / 255.0f)) // >= 0x80 pass
o_col0 = vec4(-1);
#endif
}
#endif
#endif

View File

@@ -0,0 +1,42 @@
// SPDX-FileCopyrightText: 2002-2025 PCSX2 Dev Team
// SPDX-License-Identifier: GPL-3.0+
#ifdef VERTEX_SHADER
layout(location = 0) in vec2 Position;
layout(location = 1) in vec2 UV;
layout(location = 2) in vec4 Color;
layout(push_constant) uniform PushConstants
{
vec2 uScale;
vec2 uTranslate;
};
layout(location = 0) out vec2 Frag_UV;
layout(location = 1) out vec4 Frag_Color;
void vs_main()
{
Frag_UV = UV;
Frag_Color = Color;
gl_Position = vec4(Position * uScale + uTranslate, 0.0f, 1.0f);
}
#endif
#ifdef FRAGMENT_SHADER
layout(binding = 0) uniform sampler2D Texture;
layout(location = 0) in vec2 Frag_UV;
layout(location = 1) in vec4 Frag_Color;
layout(location = 0) out vec4 Out_Color;
void ps_main()
{
Out_Color = Frag_Color * texture(Texture, Frag_UV.st);
}
#endif

View File

@@ -0,0 +1,203 @@
// SPDX-FileCopyrightText: 2002-2025 PCSX2 Dev Team
// SPDX-License-Identifier: GPL-3.0+
#ifdef VERTEX_SHADER
layout(location = 0) in vec4 a_pos;
layout(location = 1) in vec2 a_tex;
layout(location = 0) out vec2 v_tex;
void main()
{
gl_Position = vec4(a_pos.x, -a_pos.y, a_pos.z, a_pos.w);
v_tex = a_tex;
}
#endif
#ifdef FRAGMENT_SHADER
layout(location = 0) in vec2 v_tex;
layout(location = 0) out vec4 o_col0;
layout(push_constant) uniform cb0
{
vec4 ZrH;
};
layout(set = 0, binding = 0) uniform sampler2D samp0;
// Weave shader
#ifdef ps_main0
void ps_main0()
{
const int idx = int(ZrH.x); // buffer index passed from CPU
const int field = idx & 1; // current field
const int vpos = int(gl_FragCoord.y); // vertical position of destination texture
if ((vpos & 1) == field)
o_col0 = textureLod(samp0, v_tex, 0);
else
discard;
}
#endif
// Bob shader
#ifdef ps_main1
void ps_main1()
{
o_col0 = textureLod(samp0, v_tex, 0);
}
#endif
// Blend shader
#ifdef ps_main2
void ps_main2()
{
vec2 vstep = vec2(0.0f, ZrH.y);
vec4 c0 = textureLod(samp0, v_tex - vstep, 0);
vec4 c1 = textureLod(samp0, v_tex, 0);
vec4 c2 = textureLod(samp0, v_tex + vstep, 0);
o_col0 = (c0 + c1 * 2.0f + c2) / 4.0f;
}
#endif
// MAD shader - buffering
#ifdef ps_main3
void ps_main3()
{
// We take half the lines from the current frame and stores them in the MAD frame buffer.
// the MAD frame buffer is split in 2 consecutive banks of 2 fields each, the fields in each bank
// are interleaved (top field at even lines and bottom field at odd lines).
// When the source texture has an odd vres, the first line of bank 1 would be an odd index
// causing the wrong lines to be discarded, so a vertical offset (lofs) is added to the vertical
// position of the destination texture to force the proper field alignment
const int idx = int(ZrH.x); // buffer index passed from CPU
const int bank = idx >> 1; // current bank
const int field = idx & 1; // current field
const int vres = int(ZrH.z) >> 1; // vertical resolution of source texture
const int lofs = ((((vres + 1) >> 1) << 1) - vres) & bank; // line alignment offset for bank 1
const int vpos = int(gl_FragCoord.y) + lofs; // vertical position of destination texture
// if the index of current destination line belongs to the current fiels we update it, otherwise
// we leave the old line in the destination buffer
if ((vpos & 1) == field)
o_col0 = textureLod(samp0, v_tex, 0);
else
discard;
}
#endif
// MAD shader - reconstruction
#ifdef ps_main4
void ps_main4()
{
// we use the contents of the MAD frame buffer to reconstruct the missing lines from the current
// field.
const int idx = int(ZrH.x); // buffer index passed from CPU
const int bank = idx >> 1; // current bank
const int field = idx & 1; // current field
const int vpos = int(gl_FragCoord.y); // vertical position of destination texture
const float sensitivity = ZrH.w; // passed from CPU, higher values mean more likely to use weave
const vec3 motion_thr = vec3(1.0, 1.0, 1.0) * sensitivity; //
const vec2 bofs = vec2(0.0f, 0.5f); // position of the bank 1 relative to source texture size
const vec2 vscale = vec2(1.0f, 0.5f); // scaling factor from source to destination texture
const vec2 lofs = vec2(0.0f, ZrH.y) * vscale; // distance between two adjacent lines relative to source texture size
const vec2 iptr = v_tex * vscale; // pointer to the current pixel in the source texture
vec2 p_t0; // pointer to current pixel (missing or not) from most recent frame
vec2 p_t1; // pointer to current pixel (missing or not) from one frame back
vec2 p_t2; // pointer to current pixel (missing or not) from two frames back
vec2 p_t3; // pointer to current pixel (missing or not) from three frames back
switch (idx)
{
case 1:
p_t0 = iptr;
p_t1 = iptr;
p_t2 = iptr + bofs;
p_t3 = iptr + bofs;
break;
case 2:
p_t0 = iptr + bofs;
p_t1 = iptr;
p_t2 = iptr;
p_t3 = iptr + bofs;
break;
case 3:
p_t0 = iptr + bofs;
p_t1 = iptr + bofs;
p_t2 = iptr;
p_t3 = iptr;
break;
default:
p_t0 = iptr;
p_t1 = iptr + bofs;
p_t2 = iptr + bofs;
p_t3 = iptr;
break;
}
// calculating motion, only relevant for missing lines where the "center line" is pointed by p_t1
vec4 hn = textureLod(samp0, p_t0 - lofs, 0); // new high pixel
vec4 cn = textureLod(samp0, p_t1, 0); // new center pixel
vec4 ln = textureLod(samp0, p_t0 + lofs, 0); // new low pixel
vec4 ho = textureLod(samp0, p_t2 - lofs, 0); // old high pixel
vec4 co = textureLod(samp0, p_t3, 0); // old center pixel
vec4 lo = textureLod(samp0, p_t2 + lofs, 0); // old low pixel
vec3 mh = hn.rgb - ho.rgb; // high pixel motion
vec3 mc = cn.rgb - co.rgb; // center pixel motion
vec3 ml = ln.rgb - lo.rgb; // low pixel motion
mh = max(mh, -mh) - motion_thr;
mc = max(mc, -mc) - motion_thr;
ml = max(ml, -ml) - motion_thr;
#if 1 // use this code to evaluate each color motion separately
float mh_max = max(max(mh.x, mh.y), mh.z);
float mc_max = max(max(mc.x, mc.y), mc.z);
float ml_max = max(max(ml.x, ml.y), ml.z);
#else // use this code to evaluate average color motion
float mh_max = mh.x + mh.y + mh.z;
float mc_max = mc.x + mc.y + mc.z;
float ml_max = ml.x + ml.y + ml.z;
#endif
// selecting deinterlacing output
if ((vpos & 1) == field) // output coordinate present on current field
{
// output coordinate present on current field
o_col0 = textureLod(samp0, p_t0, 0);
}
else if ((iptr.y > 0.5f - lofs.y) || (iptr.y < 0.0 + lofs.y))
{
// top and bottom lines are always weaved
o_col0 = cn;
}
else
{
// missing line needs to be reconstructed
if(((mh_max > 0.0f) || (ml_max > 0.0f)) || (mc_max > 0.0f))
// high motion -> interpolate pixels above and below
o_col0 = (hn + ln) / 2.0f;
else
// low motion -> weave
o_col0 = cn;
}
}
#endif
#endif

View File

@@ -0,0 +1,46 @@
// SPDX-FileCopyrightText: 2002-2025 PCSX2 Dev Team
// SPDX-License-Identifier: GPL-3.0+
#ifdef VERTEX_SHADER
layout(location = 0) in vec4 a_pos;
layout(location = 1) in vec2 a_tex;
layout(location = 0) out vec2 v_tex;
void main()
{
gl_Position = vec4(a_pos.x, -a_pos.y, a_pos.z, a_pos.w);
v_tex = a_tex;
}
#endif
#ifdef FRAGMENT_SHADER
layout(location = 0) in vec2 v_tex;
layout(location = 0) out vec4 o_col0;
layout(push_constant) uniform cb10
{
vec4 BGColor;
};
layout(set = 0, binding = 0) uniform sampler2D samp0;
void ps_main0()
{
vec4 c = texture(samp0, v_tex);
// Note: clamping will be done by fixed unit
c.a *= 2.0f;
o_col0 = c;
}
void ps_main1()
{
vec4 c = texture(samp0, v_tex);
c.a = BGColor.a;
o_col0 = c;
}
#endif

View File

@@ -0,0 +1,446 @@
// SPDX-FileCopyrightText: 2002-2025 PCSX2 Dev Team
// SPDX-License-Identifier: GPL-3.0+
#ifdef VERTEX_SHADER
layout(location = 0) in vec4 a_pos;
layout(location = 1) in vec2 a_tex;
layout(location = 0) out vec2 v_tex;
void main()
{
gl_Position = vec4(a_pos.x, -a_pos.y, a_pos.z, a_pos.w);
v_tex = a_tex;
}
#endif
#ifdef FRAGMENT_SHADER
layout(push_constant) uniform cb10
{
vec4 u_source_rect;
vec4 u_target_rect;
vec2 u_source_size;
vec2 u_target_size;
vec2 u_target_resolution;
vec2 u_rcp_target_resolution; // 1 / u_target_resolution
vec2 u_source_resolution;
vec2 u_rcp_source_resolution; // 1 / u_source_resolution
float u_time;
};
layout(location = 0) in vec2 v_tex;
layout(location = 0) out vec4 o_col0;
layout(set = 0, binding = 0) uniform sampler2D samp0;
vec4 sample_c(vec2 uv)
{
return texture(samp0, uv);
}
vec4 ps_crt(uint i)
{
vec4 mask[4] = vec4[4](
vec4(1, 0, 0, 0),
vec4(0, 1, 0, 0),
vec4(0, 0, 1, 0),
vec4(1, 1, 1, 0));
return sample_c(v_tex) * clamp((mask[i] + 0.5f), 0.0f, 1.0f);
}
vec4 ps_scanlines(uint i)
{
vec4 mask[2] =
{
vec4(1, 1, 1, 0),
vec4(0, 0, 0, 0)};
return sample_c(v_tex) * clamp((mask[i] + 0.5f), 0.0f, 1.0f);
}
#ifdef ps_copy
void ps_copy()
{
o_col0 = sample_c(v_tex);
}
#endif
#ifdef ps_filter_scanlines
void ps_filter_scanlines() // scanlines
{
uvec4 p = uvec4(gl_FragCoord);
o_col0 = ps_scanlines(p.y % 2);
}
#endif
#ifdef ps_filter_diagonal
void ps_filter_diagonal() // diagonal
{
uvec4 p = uvec4(gl_FragCoord);
o_col0 = ps_crt((p.x + (p.y % 3)) % 3);
}
#endif
#ifdef ps_filter_triangular
void ps_filter_triangular() // triangular
{
uvec4 p = uvec4(gl_FragCoord);
// output.c = ps_crt(input, ((p.x + (p.y & 1) * 3) >> 1) % 3);
o_col0 = ps_crt(((p.x + ((p.y >> 1) & 1) * 3) >> 1) % 3);
}
#endif
#ifdef ps_filter_complex
void ps_filter_complex() // triangular
{
const float PI = 3.14159265359f;
vec2 texdim = vec2(textureSize(samp0, 0));
o_col0 = (0.9 - 0.4 * cos(2 * PI * v_tex.y * texdim.y)) * sample_c(vec2(v_tex.x, (floor(v_tex.y * texdim.y) + 0.5) / texdim.y));
}
#endif
#ifdef ps_filter_lottes
#define MaskingType 4 //[1|2|3|4] The type of CRT shadow masking used. 1: compressed TV style, 2: Aperture-grille, 3: Stretched VGA style, 4: VGA style.
#define ScanBrightness -8.00 //[-16.0 to 1.0] The overall brightness of the scanline effect. Lower for darker, higher for brighter.
#define FilterCRTAmount -3.00 //[-4.0 to 1.0] The amount of filtering used, to replicate the TV CRT look. Lower for less, higher for more.
#define HorizontalWarp 0.00 //[0.0 to 0.1] The distortion warping effect for the horizontal (x) axis of the screen. Use small increments.
#define VerticalWarp 0.00 //[0.0 to 0.1] The distortion warping effect for the verticle (y) axis of the screen. Use small increments.
#define MaskAmountDark 0.50 //[0.0 to 1.0] The value of the dark masking line effect used. Lower for darker lower end masking, higher for brighter.
#define MaskAmountLight 1.50 //[0.0 to 2.0] The value of the light masking line effect used. Lower for darker higher end masking, higher for brighter.
#define BloomPixel -1.50 //[-2.0 -0.5] Pixel bloom radius. Higher for increased softness of bloom.
#define BloomScanLine -2.0 //[-4.0 -1.0] Scanline bloom radius. Higher for increased softness of bloom.
#define BloomAmount 0.15 //[0.0 1.0] Bloom intensity. Higher for brighter.
#define Shape 2.0 //[0.0 10.0] Kernal filter shape. Lower values will darken image and introduce moire patterns if used with curvature.
#define UseShadowMask 1 //[0 or 1] Enables, or disables the use of the CRT shadow mask. 0 is disabled, 1 is enabled.
float ToLinear1(float c)
{
return c <= 0.04045 ? c / 12.92 : pow((c + 0.055) / 1.055, 2.4);
}
vec3 ToLinear(vec3 c)
{
return vec3(ToLinear1(c.r), ToLinear1(c.g), ToLinear1(c.b));
}
float ToSrgb1(float c)
{
return c < 0.0031308 ? c * 12.92 : 1.055 * pow(c, 0.41666) - 0.055;
}
vec3 ToSrgb(vec3 c)
{
return vec3(ToSrgb1(c.r), ToSrgb1(c.g), ToSrgb1(c.b));
}
vec3 Fetch(vec2 pos, vec2 off)
{
pos = (floor(pos * u_target_size + off) + vec2(0.5, 0.5)) / u_target_size;
if (max(abs(pos.x - 0.5), abs(pos.y - 0.5)) > 0.5)
{
return vec3(0.0, 0.0, 0.0);
}
else
{
return ToLinear(texture(samp0, pos.xy).rgb);
}
}
vec2 Dist(vec2 pos)
{
pos = pos * vec2(640, 480);
return -((pos - floor(pos)) - vec2(0.5, 0.5));
}
float Gaus(float pos, float scale)
{
return exp2(scale * pow(abs(pos), Shape));
}
vec3 Horz3(vec2 pos, float off)
{
vec3 b = Fetch(pos, vec2(-1.0, off));
vec3 c = Fetch(pos, vec2(0.0, off));
vec3 d = Fetch(pos, vec2(1.0, off));
float dst = Dist(pos).x;
// Convert distance to weight.
float scale = FilterCRTAmount;
float wb = Gaus(dst - 1.0, scale);
float wc = Gaus(dst + 0.0, scale);
float wd = Gaus(dst + 1.0, scale);
return (b * wb + c * wc + d * wd) / (wb + wc + wd);
}
vec3 Horz5(vec2 pos, float off)
{
vec3 a = Fetch(pos, vec2(-2.0, off));
vec3 b = Fetch(pos, vec2(-1.0, off));
vec3 c = Fetch(pos, vec2(0.0, off));
vec3 d = Fetch(pos, vec2(1.0, off));
vec3 e = Fetch(pos, vec2(2.0, off));
float dst = Dist(pos).x;
// Convert distance to weight.
float scale = FilterCRTAmount;
float wa = Gaus(dst - 2.0, scale);
float wb = Gaus(dst - 1.0, scale);
float wc = Gaus(dst + 0.0, scale);
float wd = Gaus(dst + 1.0, scale);
float we = Gaus(dst + 2.0, scale);
return (a * wa + b * wb + c * wc + d * wd + e * we) / (wa + wb + wc + wd + we);
}
vec3 Horz7(vec2 pos, float off)
{
vec3 a = Fetch(pos, vec2(-3.0, off));
vec3 b = Fetch(pos, vec2(-2.0, off));
vec3 c = Fetch(pos, vec2(-1.0, off));
vec3 d = Fetch(pos, vec2( 0.0, off));
vec3 e = Fetch(pos, vec2( 1.0, off));
vec3 f = Fetch(pos, vec2( 2.0, off));
vec3 g = Fetch(pos, vec2( 3.0, off));
float dst = Dist(pos).x;
// Convert distance to weight.
float scale = BloomPixel;
float wa = Gaus(dst - 3.0, scale);
float wb = Gaus(dst - 2.0, scale);
float wc = Gaus(dst - 1.0, scale);
float wd = Gaus(dst + 0.0, scale);
float we = Gaus(dst + 1.0, scale);
float wf = Gaus(dst + 2.0, scale);
float wg = Gaus(dst + 3.0, scale);
// Return filtered sample.
return (a * wa + b * wb + c * wc + d * wd + e * we + f * wf + g * wg) / (wa + wb + wc + wd + we + wf + wg);
}
// Return scanline weight.
float Scan(vec2 pos, float off)
{
float dst = Dist(pos).y;
return Gaus(dst + off, ScanBrightness);
}
float BloomScan(vec2 pos, float off)
{
float dst = Dist(pos).y;
return Gaus(dst + off, BloomScanLine);
}
vec3 Tri(vec2 pos)
{
vec3 a = Horz3(pos, -1.0);
vec3 b = Horz5(pos, 0.0);
vec3 c = Horz3(pos, 1.0);
float wa = Scan(pos, -1.0);
float wb = Scan(pos, 0.0);
float wc = Scan(pos, 1.0);
return (a * wa) + (b * wb) + (c * wc);
}
vec3 Bloom(vec2 pos)
{
vec3 a = Horz5(pos,-2.0);
vec3 b = Horz7(pos,-1.0);
vec3 c = Horz7(pos, 0.0);
vec3 d = Horz7(pos, 1.0);
vec3 e = Horz5(pos, 2.0);
float wa = BloomScan(pos,-2.0);
float wb = BloomScan(pos,-1.0);
float wc = BloomScan(pos, 0.0);
float wd = BloomScan(pos, 1.0);
float we = BloomScan(pos, 2.0);
return a * wa + b * wb + c * wc + d * wd + e * we;
}
vec2 Warp(vec2 pos)
{
pos = pos * 2.0 - 1.0;
pos *= vec2(1.0 + (pos.y * pos.y) * HorizontalWarp, 1.0 + (pos.x * pos.x) * VerticalWarp);
return pos * 0.5 + 0.5;
}
vec3 Mask(vec2 pos)
{
#if MaskingType == 1
// Very compressed TV style shadow mask.
float lines = MaskAmountLight;
float odd = 0.0;
if (fract(pos.x / 6.0) < 0.5)
{
odd = 1.0;
}
if (fract((pos.y + odd) / 2.0) < 0.5)
{
lines = MaskAmountDark;
}
pos.x = fract(pos.x / 3.0);
vec3 mask = vec3(MaskAmountDark, MaskAmountDark, MaskAmountDark);
if (pos.x < 0.333)
{
mask.r = MaskAmountLight;
}
else if (pos.x < 0.666)
{
mask.g = MaskAmountLight;
}
else
{
mask.b = MaskAmountLight;
}
mask *= lines;
return mask;
#elif MaskingType == 2
// Aperture-grille.
pos.x = fract(pos.x / 3.0);
vec3 mask = vec3(MaskAmountDark, MaskAmountDark, MaskAmountDark);
if (pos.x < 0.333)
{
mask.r = MaskAmountLight;
}
else if (pos.x < 0.666)
{
mask.g = MaskAmountLight;
}
else
{
mask.b = MaskAmountLight;
}
return mask;
#elif MaskingType == 3
// Stretched VGA style shadow mask (same as prior shaders).
pos.x += pos.y * 3.0;
vec3 mask = vec3(MaskAmountDark, MaskAmountDark, MaskAmountDark);
pos.x = fract(pos.x / 6.0);
if (pos.x < 0.333)
{
mask.r = MaskAmountLight;
}
else if (pos.x < 0.666)
{
mask.g = MaskAmountLight;
}
else
{
mask.b = MaskAmountLight;
}
return mask;
#else
// VGA style shadow mask.
pos.xy = floor(pos.xy * vec2(1.0, 0.5));
pos.x += pos.y * 3.0;
vec3 mask = vec3(MaskAmountDark, MaskAmountDark, MaskAmountDark);
pos.x = fract(pos.x / 6.0);
if (pos.x < 0.333)
{
mask.r = MaskAmountLight;
}
else if (pos.x < 0.666)
{
mask.g = MaskAmountLight;
}
else
{
mask.b = MaskAmountLight;
}
return mask;
#endif
}
vec4 LottesCRTPass()
{
vec4 color;
vec4 fragcoord = gl_FragCoord - u_target_rect;
vec2 inSize = u_target_resolution - (2 * u_target_rect.xy);
vec2 pos = Warp(fragcoord.xy / inSize);
color.rgb = Tri(pos);
color.rgb += Bloom(pos) * BloomAmount;
#if UseShadowMask
color.rgb *= Mask(fragcoord.xy);
#endif
color.rgb = ToSrgb(color.rgb);
return color;
}
void ps_filter_lottes()
{
o_col0 = LottesCRTPass();
}
#endif
#ifdef ps_4x_rgss
void ps_4x_rgss()
{
vec2 dxy = vec2(dFdx(v_tex.x), dFdy(v_tex.y));
vec3 color = vec3(0);
float s = 1.0/8.0;
float l = 3.0/8.0;
color += sample_c(v_tex + vec2( s, l) * dxy).rgb;
color += sample_c(v_tex + vec2( l,-s) * dxy).rgb;
color += sample_c(v_tex + vec2(-s,-l) * dxy).rgb;
color += sample_c(v_tex + vec2(-l, s) * dxy).rgb;
o_col0 = vec4(color * 0.25,1);
}
#endif
#ifdef ps_automagical_supersampling
void ps_automagical_supersampling()
{
vec2 ratio = (u_source_size / u_target_size) * 0.5;
vec2 steps = floor(ratio);
vec3 col = sample_c(v_tex).rgb;
float div = 1;
for (float y = 0; y < steps.y; y++)
{
for (float x = 0; x < steps.x; x++)
{
vec2 offset = vec2(x,y) - ratio * 0.5;
col += sample_c(v_tex + offset * u_rcp_source_resolution * 2.0).rgb;
div++;
}
}
o_col0 = vec4(col / div, 1);
}
#endif
#endif

View File

@@ -0,0 +1,74 @@
// SPDX-FileCopyrightText: 2002-2025 PCSX2 Dev Team
// SPDX-License-Identifier: GPL-3.0+
//#version 420 // Keep it for editor detection
#ifdef VERTEX_SHADER
layout(location = 0) in vec4 a_pos;
layout(location = 1) in vec2 a_tex;
layout(location = 0) out vec2 v_tex;
void main()
{
gl_Position = vec4(a_pos.x, -a_pos.y, a_pos.z, a_pos.w);
v_tex = a_tex;
}
#endif
/*
** Contrast, saturation, brightness
** Code of this function is from TGM's shader pack
** http://irrlicht.sourceforge.net/phpBB2/viewtopic.php?t=21057
** TGM's author comment about the license (included in the previous link)
** "do with it, what you want! its total free!
** (but would be nice, if you say that you used my shaders :wink: ) but not necessary"
*/
#ifdef FRAGMENT_SHADER
layout(push_constant) uniform cb0
{
vec4 params;
};
layout(set = 0, binding = 0) uniform sampler2D samp0;
layout(location = 0) in vec2 v_tex;
layout(location = 0) out vec4 o_col0;
// For all settings: 1.0 = 100% 0.5=50% 1.5 = 150%
vec4 ContrastSaturationBrightness(vec4 color)
{
float brt = params.x;
float con = params.y;
float sat = params.z;
// Increase or decrease these values to adjust r, g and b color channels separately
const float AvgLumR = 0.5;
const float AvgLumG = 0.5;
const float AvgLumB = 0.5;
const vec3 LumCoeff = vec3(0.2125, 0.7154, 0.0721);
vec3 AvgLumin = vec3(AvgLumR, AvgLumG, AvgLumB);
vec3 brtColor = color.rgb * brt;
float dot_intensity = dot(brtColor, LumCoeff);
vec3 intensity = vec3(dot_intensity, dot_intensity, dot_intensity);
vec3 satColor = mix(intensity, brtColor, sat);
vec3 conColor = mix(AvgLumin, satColor, con);
color.rgb = conColor;
return color;
}
void main()
{
vec4 c = texture(samp0, v_tex);
o_col0 = ContrastSaturationBrightness(c);
}
#endif

File diff suppressed because it is too large Load Diff