// (C) 2015 Jan-Willem Krans (janwillem32 <at> hotmail.com)
// This file is part of Video pixel shader pack.
// This program is free software; you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation, version 2.
// This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.
// You should have received a copy of the GNU General Public License along with this program; if not, write to the Free Software Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
// 4×4 block-based error diffusion dithering
// This shader can be run as a screen space pixel shader.
// This shader requires compiling with ps_2_a, but higher is better, see http://en.wikipedia.org/wiki/Pixel_shader to look up what PS version your video card supports.
// This shader will dither down an image using block-based error diffusion.
#define sa(A, B, C) float3 A = tex2D(s0, tex+float2(B, C)*c1).rgb;
#define Quantization 255.
sampler s0 : register(s0);
float2 c0 : register(c0);
float2 c1 : register(c1);
float4 main(float2 tex : TEXCOORD0) : COLOR
{
float2 n = frac(tex*c0*.25);
tex -= (n*4.-.5)*c1;
// original pixels
sa(s1, 0., 0.) sa(s2, 1., 0.) sa(s3, 2., 0.) sa(s4, 3., 0.)
sa(s12, 0., 1.) sa(s13, 1., 1.) sa(s14, 2., 1.) sa(s5, 3., 1.)
sa(s11, 0., 2.) sa(s16, 1., 2.) sa(s15, 2., 2.) sa(s6, 3., 2.)
sa(s10, 0., 3.) sa(s9, 1., 3.) sa(s8, 2., 3.) sa(s7, 3., 3.)
float3 f1 = frac(s1*Quantization);
float3 f2 = frac(s2*Quantization);
float3 f3 = frac(s3*Quantization);
float3 f4 = frac(s4*Quantization);
float3 f5 = frac(s5*Quantization);
float3 f6 = frac(s6*Quantization);
float3 f7 = frac(s7*Quantization);
float3 f8 = frac(s8*Quantization);
float3 f9 = frac(s9*Quantization);
float3 f10 = frac(s10*Quantization);
float3 f11 = frac(s11*Quantization);
float3 f12 = frac(s12*Quantization);
float3 f13 = frac(s13*Quantization);
float3 f14 = frac(s14*Quantization);
float3 f15 = frac(s15*Quantization);
float3 f16 = frac(s16*Quantization);
float3 current;
float3 mf1 = -f1;
float3 uf1 = 1.-f1;
[flatten] if(f1.r > .5) f1.r = uf1.r;
else f1.r = mf1.r;
[flatten] if(f1.g > .5) f1.g = uf1.g;
else f1.g = mf1.g;
[flatten] if(f1.b > .5) f1.b = uf1.b;
else f1.b = mf1.b;
s1 += f1/Quantization;
current = -f1;
float3 cu2 = current+f2;
float3 mf2 = -f2;
float3 uf2 = 1.-f2;
[flatten] if(cu2.r > .5) f2.r = uf2.r;
else f2.r = mf2.r;
[flatten] if(cu2.g > .5) f2.g = uf2.g;
else f2.g = mf2.g;
[flatten] if(cu2.b > .5) f2.b = uf2.b;
else f2.b = mf2.b;
s2 += f2/Quantization;
current -= f2;
float3 cu3 = current+f3;
float3 mf3 = -f3;
float3 uf3 = 1.-f3;
[flatten] if(cu3.r > .5) f3.r = uf3.r;
else f3.r = mf3.r;
[flatten] if(cu3.g > .5) f3.g = uf3.g;
else f3.g = mf3.g;
[flatten] if(cu3.b > .5) f3.b = uf3.b;
else f3.b = mf3.b;
s3 += f3/Quantization;
current -= f3;
float3 cu4 = current+f4;
float3 mf4 = -f4;
float3 uf4 = 1.-f4;
[flatten] if(cu4.r > .5) f4.r = uf4.r;
else f4.r = mf4.r;
[flatten] if(cu4.g > .5) f4.g = uf4.g;
else f4.g = mf4.g;
[flatten] if(cu4.b > .5) f4.b = uf4.b;
else f4.b = mf4.b;
s4 += f4/Quantization;
current -= f4;
float3 cu5 = current+f5;
float3 mf5 = -f5;
float3 uf5 = 1.-f5;
[flatten] if(cu5.r > .5) f5.r = uf5.r;
else f5.r = mf5.r;
[flatten] if(cu5.g > .5) f5.g = uf5.g;
else f5.g = mf5.g;
[flatten] if(cu5.b > .5) f5.b = uf5.b;
else f5.b = mf5.b;
s5 += f5/Quantization;
current -= f5;
float3 cu6 = current+f6;
float3 mf6 = -f6;
float3 uf6 = 1.-f6;
[flatten] if(cu6.r > .5) f6.r = uf6.r;
else f6.r = mf6.r;
[flatten] if(cu6.g > .5) f6.g = uf6.g;
else f6.g = mf6.g;
[flatten] if(cu6.b > .5) f6.b = uf6.b;
else f6.b = mf6.b;
s6 += f6/Quantization;
current -= f6;
float3 cu7 = current+f7;
float3 mf7 = -f7;
float3 uf7 = 1.-f7;
[flatten] if(cu7.r > .5) f7.r = uf7.r;
else f7.r = mf7.r;
[flatten] if(cu7.g > .5) f7.g = uf7.g;
else f7.g = mf7.g;
[flatten] if(cu7.b > .5) f7.b = uf7.b;
else f7.b = mf7.b;
s7 += f7/Quantization;
current -= f7;
float3 cu8 = current+f8;
float3 mf8 = -f8;
float3 uf8 = 1.-f8;
[flatten] if(cu8.r > .5) f8.r = uf8.r;
else f8.r = mf8.r;
[flatten] if(cu8.g > .5) f8.g = uf8.g;
else f8.g = mf8.g;
[flatten] if(cu8.b > .5) f8.b = uf8.b;
else f8.b = mf8.b;
s8 += f8/Quantization;
current -= f8;
float3 cu9 = current+f9;
float3 mf9 = -f9;
float3 uf9 = 1.-f9;
[flatten] if(cu9.r > .5) f9.r = uf9.r;
else f9.r = mf9.r;
[flatten] if(cu9.g > .5) f9.g = uf9.g;
else f9.g = mf9.g;
[flatten] if(cu9.b > .5) f9.b = uf9.b;
else f9.b = mf9.b;
s9 += f9/Quantization;
current -= f9;
float3 cu10 = current+f10;
float3 mf10 = -f10;
float3 uf10 = 1.-f10;
[flatten] if(cu10.r > .5) f10.r = uf10.r;
else f10.r = mf10.r;
[flatten] if(cu10.g > .5) f10.g = uf10.g;
else f10.g = mf10.g;
[flatten] if(cu10.b > .5) f10.b = uf10.b;
else f10.b = mf10.b;
s10 += f10/Quantization;
current -= f10;
float3 cu11 = current+f11;
float3 mf11 = -f11;
float3 uf11 = 1.-f11;
[flatten] if(cu11.r > .5) f11.r = uf11.r;
else f11.r = mf11.r;
[flatten] if(cu11.g > .5) f11.g = uf11.g;
else f11.g = mf11.g;
[flatten] if(cu11.b > .5) f11.b = uf11.b;
else f11.b = mf11.b;
s11 += f11/Quantization;
current -= f11;
float3 cu12 = current+f12;
float3 mf12 = -f12;
float3 uf12 = 1.-f12;
[flatten] if(cu12.r > .5) f12.r = uf12.r;
else f12.r = mf12.r;
[flatten] if(cu12.g > .5) f12.g = uf12.g;
else f12.g = mf12.g;
[flatten] if(cu12.b > .5) f12.b = uf12.b;
else f12.b = mf12.b;
s12 += f12/Quantization;
current -= f12;
float3 cu13 = current+f13;
float3 mf13 = -f13;
float3 uf13 = 1.-f13;
[flatten] if(cu13.r > .5) f13.r = uf13.r;
else f13.r = mf13.r;
[flatten] if(cu13.g > .5) f13.g = uf13.g;
else f13.g = mf13.g;
[flatten] if(cu13.b > .5) f13.b = uf13.b;
else f13.b = mf13.b;
s13 += f13/Quantization;
current -= f13;
float3 cu14 = current+f14;
float3 mf14 = -f14;
float3 uf14 = 1.-f14;
[flatten] if(cu14.r > .5) f14.r = uf14.r;
else f14.r = mf14.r;
[flatten] if(cu14.g > .5) f14.g = uf14.g;
else f14.g = mf14.g;
[flatten] if(cu14.b > .5) f14.b = uf14.b;
else f14.b = mf14.b;
s14 += f14/Quantization;
current -= f14;
float3 cu15 = current+f15;
float3 mf15 = -f15;
float3 uf15 = 1.-f15;
[flatten] if(cu15.r > .5) f15.r = uf15.r;
else f15.r = mf15.r;
[flatten] if(cu15.g > .5) f15.g = uf15.g;
else f15.g = mf15.g;
[flatten] if(cu15.b > .5) f15.b = uf15.b;
else f15.b = mf15.b;
s15 += f15/Quantization;
current -= f15;
float3 cu16 = current+f16;
float3 mf16 = -f16;
float3 uf16 = 1.-f16;
[flatten] if(cu16.r > .5) f16.r = uf16.r;
else f16.r = mf16.r;
[flatten] if(cu16.g > .5) f16.g = uf16.g;
else f16.g = mf16.g;
[flatten] if(cu16.b > .5) f16.b = uf16.b;
else f16.b = mf16.b;
s16 += f16/Quantization;
//current -= f16;
[flatten] if(n.y > .75) {
[flatten] if(n.x > .75) return s7.rgbb;
else [flatten] if(n.x > .5) return s8.rgbb;
else [flatten] if(n.x > .25) return s9.rgbb;
else return s10.rgbb;}
else [flatten] if(n.y > .5) {
[flatten] if(n.x > .75) return s6.rgbb;
else [flatten] if(n.x > .5) return s15.rgbb;
else [flatten] if(n.x > .25) return s16.rgbb;
else return s11.rgbb;}
else [flatten] if(n.y > .25) {
[flatten] if(n.x > .75) return s5.rgbb;
else [flatten] if(n.x > .5) return s14.rgbb;
else [flatten] if(n.x > .25) return s13.rgbb;
else return s12.rgbb;}
else {
[flatten] if(n.x > .75) return s4.rgbb;
else [flatten] if(n.x > .5) return s3.rgbb;
else [flatten] if(n.x > .25) return s2.rgbb;
else return s1.rgbb;}
}// (C) 2015 Jan-Willem Krans (janwillem32 <at> hotmail.com)
// This file is part of Video pixel shader pack.
// This program is free software; you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation, version 2.
// This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.
// You should have received a copy of the GNU General Public License along with this program; if not, write to the Free Software Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
// 3×3 block-based error diffusion dithering
// This shader can be run as a screen space pixel shader.
// This shader requires compiling with ps_2_a, but higher is better, see http://en.wikipedia.org/wiki/Pixel_shader to look up what PS version your video card supports.
// This shader will dither down an image using block-based error diffusion.
#define sa(A, B, C) float3 A = tex2D(s0, tex+float2(B, C)*c1).rgb;
#define Quantization 255.
sampler s0 : register(s0);
float2 c0 : register(c0);
float2 c1 : register(c1);
float4 main(float2 tex : TEXCOORD0) : COLOR
{
float2 n = frac(tex*c0/3.);
tex -= (n*3.-.5)*c1;
// original pixels
sa(s1, 0., 0.) sa(s2, 1., 0.) sa(s3, 2., 0.)
sa(s8, 0., 1.) sa(s9, 1., 1.) sa(s4, 2., 1.)
sa(s7, 0., 2.) sa(s6, 1., 2.) sa(s5, 2., 2.)
float3 f1 = frac(s1*Quantization);
float3 f2 = frac(s2*Quantization);
float3 f3 = frac(s3*Quantization);
float3 f4 = frac(s4*Quantization);
float3 f5 = frac(s5*Quantization);
float3 f6 = frac(s6*Quantization);
float3 f7 = frac(s7*Quantization);
float3 f8 = frac(s8*Quantization);
float3 f9 = frac(s9*Quantization);
float3 current;
float3 mf1 = -f1;
float3 uf1 = 1.-f1;
[flatten] if(f1.r > .5) f1.r = uf1.r;
else f1.r = mf1.r;
[flatten] if(f1.g > .5) f1.g = uf1.g;
else f1.g = mf1.g;
[flatten] if(f1.b > .5) f1.b = uf1.b;
else f1.b = mf1.b;
s1 += f1/Quantization;
current = -f1;
float3 cu2 = current+f2;
float3 mf2 = -f2;
float3 uf2 = 1.-f2;
[flatten] if(cu2.r > .5) f2.r = uf2.r;
else f2.r = mf2.r;
[flatten] if(cu2.g > .5) f2.g = uf2.g;
else f2.g = mf2.g;
[flatten] if(cu2.b > .5) f2.b = uf2.b;
else f2.b = mf2.b;
s2 += f2/Quantization;
current -= f2;
float3 cu3 = current+f3;
float3 mf3 = -f3;
float3 uf3 = 1.-f3;
[flatten] if(cu3.r > .5) f3.r = uf3.r;
else f3.r = mf3.r;
[flatten] if(cu3.g > .5) f3.g = uf3.g;
else f3.g = mf3.g;
[flatten] if(cu3.b > .5) f3.b = uf3.b;
else f3.b = mf3.b;
s3 += f3/Quantization;
current -= f3;
float3 cu4 = current+f4;
float3 mf4 = -f4;
float3 uf4 = 1.-f4;
[flatten] if(cu4.r > .5) f4.r = uf4.r;
else f4.r = mf4.r;
[flatten] if(cu4.g > .5) f4.g = uf4.g;
else f4.g = mf4.g;
[flatten] if(cu4.b > .5) f4.b = uf4.b;
else f4.b = mf4.b;
s4 += f4/Quantization;
current -= f4;
float3 cu5 = current+f5;
float3 mf5 = -f5;
float3 uf5 = 1.-f5;
[flatten] if(cu5.r > .5) f5.r = uf5.r;
else f5.r = mf5.r;
[flatten] if(cu5.g > .5) f5.g = uf5.g;
else f5.g = mf5.g;
[flatten] if(cu5.b > .5) f5.b = uf5.b;
else f5.b = mf5.b;
s5 += f5/Quantization;
current -= f5;
float3 cu6 = current+f6;
float3 mf6 = -f6;
float3 uf6 = 1.-f6;
[flatten] if(cu6.r > .5) f6.r = uf6.r;
else f6.r = mf6.r;
[flatten] if(cu6.g > .5) f6.g = uf6.g;
else f6.g = mf6.g;
[flatten] if(cu6.b > .5) f6.b = uf6.b;
else f6.b = mf6.b;
s6 += f6/Quantization;
current -= f6;
float3 cu7 = current+f7;
float3 mf7 = -f7;
float3 uf7 = 1.-f7;
[flatten] if(cu7.r > .5) f7.r = uf7.r;
else f7.r = mf7.r;
[flatten] if(cu7.g > .5) f7.g = uf7.g;
else f7.g = mf7.g;
[flatten] if(cu7.b > .5) f7.b = uf7.b;
else f7.b = mf7.b;
s7 += f7/Quantization;
current -= f7;
float3 cu8 = current+f8;
float3 mf8 = -f8;
float3 uf8 = 1.-f8;
[flatten] if(cu8.r > .5) f8.r = uf8.r;
else f8.r = mf8.r;
[flatten] if(cu8.g > .5) f8.g = uf8.g;
else f8.g = mf8.g;
[flatten] if(cu8.b > .5) f8.b = uf8.b;
else f8.b = mf8.b;
s8 += f8/Quantization;
current -= f8;
float3 cu9 = current+f9;
float3 mf9 = -f9;
float3 uf9 = 1.-f9;
[flatten] if(cu9.r > .5) f9.r = uf9.r;
else f9.r = mf9.r;
[flatten] if(cu9.g > .5) f9.g = uf9.g;
else f9.g = mf9.g;
[flatten] if(cu9.b > .5) f9.b = uf9.b;
else f9.b = mf9.b;
s9 += f9/Quantization;
//current -= f9;
[flatten] if(n.y > 2./3.) {
[flatten] if(n.x > 2./3.) return s5.rgbb;
else [flatten] if(n.x > 1./3.) return s6.rgbb;
else return s7.rgbb;}
else [flatten] if(n.y > 1./3.) {
[flatten] if(n.x > 2./3.) return s4.rgbb;
else [flatten] if(n.x > 1./3.) return s9.rgbb;
else return s8.rgbb;}
else {
[flatten] if(n.x > 2./3.) return s3.rgbb;
else [flatten] if(n.x > 1./3.) return s2.rgbb;
else return s1.rgbb;}
}