noise-rs
noise-rs copied to clipboard
Razaekel
Procedural noise generation library for Rust.
As mentioned in #102 the output range of OpenSimplex needs to be fixed.
A lot of the code in each multifractal NoiseModule is common across PointN. Should extract this out in order to follow DRY.
The ones that are particularly bad are worley-linear-range, worley-manhattan-range, and worley-chebyshev-range. The ones that fall within [-1,1] but don't hit it exactly (such as worley-linear-squared_4d) are probably just inadequate sampling...
worley-quadratic_2d  worley-quadratic_3d  worley-quadratic_4d  worley-quadratic-range_2d  worley-quadratic-range_3d  worley-quadratic-range_4d  Unless I'm mistaken, that looks very different from what it should. For comparison, here are the linear equivalents...
Sampling worley noise at half-kernels requires a minimum of eight generating points to be samples. Your library only wanders the control points around a disk of radius 0.5 instead of...
Your current feature point selection code constrains the generated points along 8 axes (2D) and contains needless branching. It also creates relatively visually unappealing noise. A much better point distribution...
> In the future I see that these structs could simply implement Fn(Input) -> Output, but we need to wait for these to be stabilized first. > > I also...
the codegen concerns seem to be resolved https://godbolt.org/z/GGeeETdKh
```shell thread '' panicked at C:\Users\dkolsoi\.cargo\registry\src\index.crates.io-6f17d22bba15001f\noise-0.8.2\src\core\open_simplex.rs:190:18: called `Option::unwrap()` on a `None` value ``` Which points to: https://github.com/Razaekel/noise-rs/blob/7fc78b5a2b046c9589f933e255d87539af2fefe9/src/core/open_simplex.rs#L190 and https://github.com/Razaekel/noise-rs/blob/7fc78b5a2b046c9589f933e255d87539af2fefe9/src/core/open_simplex.rs#L136C59-L136C65 Not sure how to repro, but giving it a shot
