A procedural noise generation library for Rust.
use noise::{Brownian3, Seed};
let seed = Seed::new(12);
let noise = Brownian3::new(noise::perlin3, 4).wavelength(32.0);
let val = noise.apply(&seed, &[42.0, 37.0, 2.0]);Gradient noise produces a smooth, continuous value over space. It's achieved by dividing space into regions, and placing a random gradient at each vertex, then blending between those gradients.
A very fast and reasonable quality gradient noise:
perlin2perlin3perlin4
A slower but higher quality form of gradient noise:
open_simplex2open_simplex3
A way of combining multiple octaves of a noise function to create a richer and more varied output:
Brownian2Brownian3Brownian4
Cell noise, also called worley noise or voronoi noise, is based on dividing space into cells based on proximity to a random set of seed points. In this API, this is accomplished in three categories.
These are the actual noise functions, which just take a coordinate and return a value. They use the functions from the later categories to generate the most common and useful types of cell noise. Most of the time, these are the only ones you'll be interested in.
These functions, when given a point, will return a value generated from the cell coordinates of the nearest seed point, so all points within the same cell will return the same value:
cell2_valuecell3_valuecell4_value
These functions, when given a point, will return the range to the nearest seed point:
cell2_rangecell3_rangecell4_range
These functions, when given a point, will return the range to the nearest cell border. This is accomplished by subtracting the range to the nearest point from the range to the second nearest point:
cell2_range_invcell3_range_invcell4_range_inv
These functions, when given a point, will return a value generated from the cell coordinates of the nearest seed point, but using manhattan distance. This results in more squared-off cells:
cell2_value_manhattancell3_value_manhattancell4_value_manhattan
These functions, when given a point, will return range to the nearest seed point, but using manhattan distance:
cell2_manhattancell3_manhattancell4_manhattan
These functions, when given a point, will return the range to the nearest cell border, but in manhattan distance:
cell2_manhattan_invcell3_manhattan_invcell4_manhattan_inv
These are the set of functions for determining range between two points, which are used to determine which seed point is actually closest to the test point.
Calculate the square of the euclidian range between two points. This is what one normally thinks of as range. It's squared because that's cheaper to calculate, and we only actually care about which distance is greater in this context:
range_sqr_euclidian2range_sqr_euclidian3range_sqr_euclidian4
Calculate the manhattan range between two points. This is the sum of the coordinates in the offset vector, equivalent to the distance one would walk on the Manhattan streets between two points:
range_manhattan2range_manhattan3range_manhattan4
These are the set of functions for returning the nearest point, nearest 2 points, or the cell of the nearest point. If you want to do something a little unusual with the cell noise, you can call these directly to operate on this intermediate data.
These functions, when given a point, will return the nearest seed point, and the range to that point:
cell2_seed_pointcell3_seed_pointcell4_seed_point
These functions, when given a point, will return the nearest 2 seed points, and the range to those points. The first point is the nearest one, and the second point the second nearest:
cell2_seed_2_pointscell3_seed_2_pointscell4_seed_2_points
These functions, when given a point, will return the cell of the nearest seed point:
cell2_seed_cellcell3_seed_cellcell4_seed_cell
Everything below this line is planned, but not yet implemented:
open_simplex4