OpenSimplex2 VS FastNoise2

Wow you have been playing with noise for a while :) . It amazes me all the different noises that exist now.. Perlin, Simplex, OpenSimplex, OpenSimplex2S, OpenSimplex2F, Worley... If you are curious about OpenSimplex https://github.com/KdotJPG/OpenSimplex2

Probably true, although a neat repo for open simplex is https://github.com/KdotJPG/OpenSimplex2

This may not answer what you're actually looking for - but a few years ago I found a library for coherent noise in Java, which at least meets my needs. I use bits and pieces of the Open Simplex Noise implementation from that library. It's unlicense/CC0, so easily copy-paste-able straight in to one's own project. https://github.com/KdotJPG/OpenSimplex2/

So I was looking for a noise function, and I found this repo from the author of the OpenSimplex algorithm. And in the GLSL implementation, which is the one I'm gonna use, since I want to generate terrain with compute shaders, the function%20%7B) looks like this:

If you want to feel safer from that, you might opt to use one of these instead. Use the ImproveXY endpoint to get a Z that repeats after 500.5626833874055. Go up one directory and into the glsl directory if you need GLSL versions instead .

Fair. I definitely agree Simplex has some characteristics that don't work great in some places. Especially when you threshold near zero of a single octave. Have you tried OpenSimplex2S? It's made to be less "bubbly" than standard simplex. In own repo, In FastNoiseLite, [Shader preview (3D noise)[https://www.shadertoy.com/view/ttdGR8]. Both the 2D function and 2D slices of the 3D function might be worth looking into if you haven't yet.

Here's a link to the original implementation: github.com/KdotJPG/OpenSimplex2/blob/master/glsl/OpenSimplex2.glsl

Note the notice at the top of the readme, the author is replacing that library with FastNoise2 which is also SIMD enabled, although the library as a whole is still a work in progress, it claims to have superior performance.

I am using a very fast [noise library](https://github.com/Auburn/FastNoise2). After generating a 64x64 area of noise I turn that into a 64x64x64 array of voxel byte data. I then use my own version of this [greedy meshing algorithm](https://github.com/cgerikj/binary-greedy-meshing).

There's a nice GDC talk about noise functions here, and if you're interested in trying things out there's an Open Source noise library written by a No Man's Sky developer here.

Very cool! Have you considered using FastNoise2 as a noise source? It's SIMD optimized C++ https://github.com/Auburn/FastNoise2 But they've made C# bindings so there's less native legwork: https://github.com/Auburn/FastNoise2Bindings

Let Simplex 3D use its same gradient generator, but use 1/2 instead of 1/6 for the noise's unskew constant. This rotates the axis angles of the gradient set out of the main noise evaluation planes. I haven't implemented+run this myself, but it is implemented here in FastNoise2 (SIMD library), and I simulated the difference using OpenSimplex2 in this image (bottom with change). The result isn't perfect, but it's much nicer I think. After making this change, either set scale to 32.69428253173828125 to make a proper -1 to 1 range, or bake it into the gradient table directly.