Scribd
Upload
10 views5 pages

simple shader algorithm

Uploaded by

Heiti Kender
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as TXT, PDF, TXT or read online on Scribd
10 views5 pages

simple shader algorithm

Uploaded by

Heiti Kender
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as TXT, PDF, TXT or read online on Scribd
You are on page 1/ 5

// Very simple algorithm. Explained great here - https://www.youtube.com/watch?

v=kbKtFN71Lfs&ab_channel=Numberphile
// You can basically do whatever transformations you want, which makes it really
fun.
// The shader is projecting points, then splatting them on the screen.
// Great demo which uses this effect - https://www.youtube.com/watch?
v=xLN3mTRlugs&ab_channel=AssemblyTV
// Example/tutorial on doing chaos game in compute.toys -
https://compute.toys/view/120

alias iv4 = vec4<i32>;alias iv3 = vec3<i32>;alias iv2 = vec2<i32>;alias uv4 =


vec4<u32>;alias uv3 = vec3<u32>;alias uv2 = vec2<u32>;alias v4 = vec4<f32>;alias v3
= vec3<f32>;alias v2 = vec2<f32>;alias m2 = mat2x2<f32>;alias m3 =
mat3x3<f32>;alias m4 = mat4x4<f32>;
#define global var<private>
var<workgroup> texCoords: array<array<float2, 16>, 16>;
#define Frame time.frame
#define T (time.elapsed*float(custom.Paused > 0.5) + 5.)
#define Tb (time.elapsed)
#define pi acos(-1.)
#define tau (acos(-1.)*2.)
global R: v2;
global U: v2;
global seed: uint;
global muv: v2;
fn rot(a: float)-> m2{ return m2(cos(a), -sin(a),sin(a), cos(a));}
fn rotX(a: float) -> m3{
let r = rot(a); return m3(1.,0.,0.,0.,r[0][0],r[0][1],0.,r[1][0],r[1][1]);
}
fn rotY(a: float) -> m3{
let r = rot(a); return m3(r[0][0],0.,r[0][1],0.,1.,0.,r[1][0],0.,r[1][1]);
}
fn rotZ(a: float) -> m3{
let r = rot(a); return m3(r[0][0],r[0][1],0.,r[1][0],r[1][1],0.,0.,0.,1.);
}
fn hash_u(_a: uint) -> uint{ var a = _a; a ^= a >> 16;a *= 0x7feb352du;a ^= a >>
15;a *= 0x846ca68bu;a ^= a >> 16;return a; }
fn hash_f() -> float{ var s = hash_u(seed); seed = s;return ( float( s ) /
float( 0xffffffffu ) ); }
fn hash_v2() -> v2{ return v2(hash_f(), hash_f()); }
fn hash_v3() -> v3{ return v3(hash_f(), hash_f(), hash_f()); }
fn hash_v4() -> v4{ return v4(hash_f(), hash_f(), hash_f(), hash_f()); }

fn sin_add(a: float)->float{return sin(a)*0.5 + 0.5;}

fn sample_disk() -> v2{


let r = hash_v2();
return v2(sin(r.x*tau),cos(r.x*tau))*sqrt(r.y);
}

#define COL_CNT 4
global kCols = array<v3, COL_CNT>(
vec3(1.,1,1), vec3(0.8,0.4,0.7),
vec3(1,1,1.)*1.5, vec3(1,1,1.)*1.5
);

fn mix_cols(_idx: float)->v3{
let idx = _idx%1.;
var cols_idx = int(idx*float(COL_CNT));
var fract_idx = fract(idx*float(COL_CNT));
fract_idx = smoothstep(0.,1.,fract_idx);
//return oklab_mix( kCols[cols_idx], kCols[(cols_idx + 1)%COL_CNT],
fract_idx );
return mix( kCols[cols_idx], kCols[(cols_idx + 1)%COL_CNT], fract_idx );
}

#storage hist_atomic array<atomic<u32>>

#storage h array<u32>

fn projParticle(_p: v3) -> v3{


var p = _p; // you still can't modify args in wgsl???
// p *= rotY(muv.x*2.);
// let dpp = dot(p,p) + 0.5 + sin(T)*0.4;

let t = T*1.;

p += v3(
sin(t+ 0.1)*0.1,
sin(t)*0.1,
sin(t)*0.2,
)*1.;
// p.z += 4.7;
// p *= rotZ(pow(sin_add(T),2.)*pi*2. * float(floor(T/pi/2.)%4 > 2.) );
// p /= p.z*0.2;
p *= 1. + sin(t*0.9 + sin(t*0.8))*0.05;
// p.x /= dpp;
// p.y /= dpp;
p.z = _p.z;
p.x /= R.x/R.y;
return p;
}

fn t_spherical(p: v3, rad: float, offs: v3) -> v3{


return p/(dot(p,p)*rad + offs);
}

#workgroup_count Splat 64 64 12
@compute @workgroup_size(256, 1,1)
fn Splat(@builtin(global_invocation_id) id: uint3) {
let Ru = uint2(textureDimensions(screen));
if (id.x >= Ru.x || id.y >= Ru.y) { return; }
R = v2(Ru); U = v2(id.xy); muv = (v2(mouse.pos) - 0.5*R)/R.y;

seed = hash_u(id.x + hash_u(Ru.x*id.y*200u)*20u + hash_u(id.x)*250u +


hash_u(id.z)*250u );
seed = hash_u(seed);

let particleIdx = id.x;

let iters = 60 - int(sin_add(T)*30.);

var t = T + sin(T*4.)*0.15*0. - hash_f()*1./30.;

let md = 5.0;
var env = ((floor(t))%md);
var env_next = ((floor(t + 1))%md);
var env_fr = fract(t);

env = mix(
env,
env_next,
smoothstep(0.,1.,
smoothstep(0.,1.,
smoothstep(0.,1.,
smoothstep(0.,1.,
env_fr
)
)
)
)
);

var p = hash_v3()*2. - 1.;


p *= 0.;

var focusDist = (custom.DOF_Focal_Dist*2. - 1.)*5.;


if(mouse.click > 0){
focusDist = muv.y;

}
// let focusDist = (custom.DOF_Focal_Dist*2. - 1.)*5.;
let dofFac = 1./vec2(R.x/R.y,1.)*custom.DOF_Amount;

let vert_cnt = 4.0 + env;


let verts = array<vec2f,5>(
vec2f(sin(1./vert_cnt*tau),cos(1./vert_cnt*tau)),
vec2f(sin(2./vert_cnt*tau),cos(2./vert_cnt*tau)),
vec2f(sin(3./vert_cnt*tau),cos(3./vert_cnt*tau)),
vec2f(sin(4./vert_cnt*tau),cos(4./vert_cnt*tau)),
vec2f(sin(5./vert_cnt*tau),cos(5./vert_cnt*tau))
);

for(var i = 0; i < iters; i++){


let r = hash_f();

// let next_p = verts[int(r*float(vert_cnt)*0.99)];


// let f_idx = floor(r*vert_cnt*(0.99 + custom.A));
// let f_idx = floor(r*vert_cnt*(1.2 + sin(T)*0.2));
let f_idx = floor(r*vert_cnt*(0.99));
let next_p = vec3f(
sin(f_idx/vert_cnt*tau*(1. + env*1.)+ sin(T)*0.4*float(floor(T/pi +
1)%3 > 1)),
cos(f_idx/vert_cnt*tau + sin(T)*0.4*float(floor(T/pi)%3 > 1.)),
sin(f_idx/vert_cnt*tau*4.4+ env + sin(T*1. + 4.)*1.),
);

p = sin(p*1. + next_p*2. - env*0.1);


if(hash_f() < 0. + sin_add(T + 4.)*0.4){
p += p*1.;

p *= rotZ(t*0.1);
p *= rotY(t*0.1);
}

if(hash_f() < 0.){


p += next_p*1.;
p.z -= 0.1;
// p += p*2.;
// p *= 0.;
// p.z = 2.
// 4;
// p += normalize(hash_v3()*2. - 1.)*3.;
}
// let next_p = sin()

// p = mix(p,next_p,0.4 + sin(T)*0.);

var q = projParticle(p);
var k = q.xy;

k += sample_disk()*abs(q.z - focusDist + sin(t + sin(t) +


f_idx*0.5))*0.05*dofFac;

let uv = k.xy/2. + 0.5;


let cc = uv2(uv.xy*R.xy);
let idx = cc.x + Ru.x * cc.y;
if (
// q.z > -20.
uv.x > 0. && uv.x < 1.
&& uv.y > 0. && uv.y < 1.
&& idx < uint(Ru.x*Ru.y)
){
atomicAdd(&hist_atomic[idx],1);
atomicAdd(&hist_atomic[idx + Ru.x*Ru.y],uint(r*2400.));
}
}
}

@compute @workgroup_size(16, 16)


fn main_image(@builtin(global_invocation_id) id: uint3) {
let res = uint2(textureDimensions(screen));
if (id.x >= res.x || id.y >= res.y) { return; }

R = v2(res);
U = v2(id.xy);
let U = float2(float(id.x) + .5, float(res.y - id.y) - .5);

let hist_id = id.x + uint(R.x) * id.y;

var col = float(atomicLoad(&hist_atomic[hist_id]))*vec3(3);

var col_pal = float(atomicLoad(&hist_atomic[hist_id + res.x*res.y]))*vec3(3);

kCols[1] *= rotY(1.-0.347*log(col.x*1.)*4. + col_pal.x*0.0001*(1. -


sin_add(T)*0.5));
// tonemap
let sc = 124452.7;
col = log( col*(0.3+custom.B))/ log(sc);
// col = abs(col);
// col = col*mix_cols(col.x*(1. + sin(T)*0.4));
col = pow(col,v3(1. - sin_add(T)*0.))*mix_cols(col.x*(1.));
// col += 0.01;
// col = mix_cols(log(col_pal.x)*0.5);

//col = smoothstep(v3(0.),v3(1.),col*mix_cols(col.x*1.));

col = max(col,v3(0.1));
col = pow((col), float3(1./0.25454545))*4.;
col = min(col,v3(0.6));

col = clamp(col,v3(0.002),v3(0.8));
textureStore(screen, int2(id.xy), float4(col, 1.));

// clear flame
// hist[id.x + uint(R.x) * id.y] = 0;
atomicStore(&hist_atomic[hist_id],0);
atomicStore(&hist_atomic[hist_id + res.x*res.y],0);
}

You might also like

pFad - Phonifier reborn

Pfad - The Proxy pFad of © 2024 Garber Painting. All rights reserved.

Note: This service is not intended for secure transactions such as banking, social media, email, or purchasing. Use at your own risk. We assume no liability whatsoever for broken pages.


Alternative Proxies:

Alternative Proxy

pFad Proxy

pFad v3 Proxy

pFad v4 Proxy