32x32x32

Spiral Zoom by OldEclipse

"Controlled Explosions" by OldEclipse!

It's been a hectic year. My brother's brain surgery, buying a house, moving, wedding and now my sister's condition. I miss coding

Graphics Programming weekly - Issue 444 - June 28th, 2026 www.jendrikillner.com/post/graphic...

I made a video about the smooth-maximum function, I hope you like it! youtu.be/6Qb6QtC6QMs

Procedural Noise Collection by lumiey

RADAR f4 p = rnd(C*.1)/.1-R.xyyy/2 O = .005/modf((atn(p.y,p.x)+T-f4(,.01,.02,))/6.3+1,O)/frc(dot(p+O,sin(p.yxyx/R.y)));

The best way to write shaders. Get the character count for free!

I Eat Pixels

My niece painted this! She thought it was terrible

The depth and thoroughness of @bgolus.bsky.social articles, when you are blessed with one relevant to what you're currently doing. (In this case, using normal maps with triplanar texture mapping.) bgolus.medium.com/normal-mappi...

Good to see

It's so important to get good at giving and receiving critical feedback. When making things, you need people who can tell you exactly what's wrong

A live session coding a procedural tower of bricks. Turned out ugly, I fixed it a bit in shadertoy after I closed the recording: youtu.be/DAMiS2PGTEE

Coming soon

This is a great thread @hillsguy.bsky.social, I'll add a bit more context here for anyone interested. Yes there's no fancy sim - it just uses a water interaction buffer as a target for the particle system rendering, so we could use the power of particle system we already have.

Disco Sun by WorkingClassHacker

Can we stop using this for water caustics?

Thank you to whoever gave me a heart

Life Force by OldEclipse

The JSON Importer is much more reliable now and works with ShaderToy built-in textures and respects privacy settings (so private shaders stay private). Try it!

Pin your favorite shaders, tutorials or code snippets on your profile. And just for fun, I made it so you can also react to profiles. Leave a reaction on mine maybe?

Aurora Paint by @noztol.bsky.social

When selling shader licenses on FragCoord you can choose your own variants or even write your own license

Listed a bunch of commercial shaders

FragCoord 1.2 -Pro Mode for publishing tutorials, commercial licenses and early access. -Compute shaders and HDR with WebGPU -Rebuilt debug modes: Tuner, Inspect, Speed -Market: for tutorials and commercial licensing

Skyline Radio by otaviogood and koncreate

Using wallscaper to generate my desktop backgrounds

Sneak peak for the new speed profiler (projects frame costs at common resolutions)

FragCoord 1.1 tomorrow!

Orange and purple

When learning shaders most tend to overcomplicate common practices like noise or color corrections. Here are 5 common noob mistakes:

Wow! f d,a;f3 p,v=R.xyy-2*C.rgb,c;v.yz*=.1*mat2(6,8,-8,6);c.yz+=3.;p=c; @(i,150)i==80?O.r++,v=c,p+=(c-p)/xp(2*frc(dot(C,sin(C+T)))):p,d=min(len(f2(len(p.xz)-2,p.y))-.3*cos(a=atn(p.x,p.z)*13+T)+.1*sin(a/.31)-.4,.6-p.z/7)/3,O*=(d*d)**.01,p-=d*nor(v);O=tanh(O+.7*tex(B,C.xy/R,2.));

Bromine Clouds by Kaso

Little shader by @altunenes.bsky.social

Quartz Dream by "0nicron" fragcoord.xyz/s/qxg7boz1

Ribbons f z,d,i @(99) { f3 p = z * nor(R.xyy-2*C.rgb) d=2 @(6) d /= .9, p = p.zxy+sin(p*d+d+T/2) / d z += d = .001+abs(2.-mix(z,p.z,.4))/9 O += (sin(z+.06*i+++f4(,1,2,)) + 1) / d } O = tanh(O / 3e4)

FragCoord is now the best place online to share: -Complex graphics demos (fluid simulation, Radiance Cascades, audio visualizers, multi-pass post-processing) -Shader functions and code snippets -HLSL, MSL, live demos -Code golfing -Advanced shader debugging

8-bit forest by @noztol.bsky.social

Hit 1000 users on FragCoord!

Light Matter f z,d @(80) { f3 p = z * nor(2*C.rgb - R.xyy); p.z++,d=4 @(3) d *= .8, p += sin(p.yzx/dot(p,p)*d+T) / len(p) / d; z += d = len(p) / 45, O += (sin(z+f4(,1,8,)) + 1) / d / d * z } O = tanh(O / 8e5)

Neon Mountains by @noztol

Switched to Perlin noise. Normally value noise is good enough, but it left a grid-like pattern

When my own website features me

On some GPUs (eg GCN/RDNA) 24bit integer muls/mads are 4x faster than 32bit ones. 24bit instructions are not exposed in HLSL/GLSL but you can encourage the compiler to use the 24bit intrinsics by zeroing the 8 most significant bits, if the range allows it, for faster integer muls/mads.

Building on @xordev.com's tutorial on glows — here's a shader I wrote using a Lorentzian function for brightness falloff. Its closed-form antiderivative makes analytic motion blur straightforward: fragcoord.xyz/s/lmojqltl. I've made it public domain, enjoy!

Do you like oklab, but dislike numbers? Introducing: klab -- it's shorter than oklab #generative #genartclub #codegolf #glsl