Chris3606
Add noise algorithm implementation(s)
Some noise algorithm implementation(s) would be useful for procedural generation scenarios.
few examples: Perlin noise, Simplex noise, Worley noise, Fractal noise
A decent SimplexNoise (better version of perlin noise) implementation would go a long way :)
You can consider https://github.com/Auburn/FastNoise/blob/master/CSharp/FastNoiseLite.cs for Perlin noise, the Simplex-like OpenSimplex2 noise, and the various fractal adjustments to those types. It also has cellular noise, which I think may be another term for Worley, but I'm not sure. OpenSimplex2 is mostly fine; I don't think it has any quality issues and its speed is good enough (better than Perlin, usually). This new FastNoiseLite library omits 4D noise, which is a little odd. In my Java edit of the older FastNoise, I added 5D and 6D noise in most types; I can see why that wouldn't be possible with FastNoiseLite because OpenSimplex2 doesn't have working 5D or 6D implementations, to my knowledge. FastNoiseLite.cs is 2500 lines, give or take, and my (ugly) Java extension is 6900 lines, give or take, but if you want to pull the 4D/5D/6D code into C# and port it, be my guest: https://github.com/yellowstonegames/SquidSquad/blob/main/squidgrid/src/main/java/com/github/yellowstonegames/grid/Noise.java . I used regular Simplex, which does generalize up to 4D, 5D, 6D, etc. but needs some tweaking for the output to be in range. I also have some other kinds of noise that look, well, different; Foam noise depends on Value noise and just averages several rotated calls to value noise, but uses the previous value noise call to domain-warp the next one; it also emphasizes extreme results because averaging several calls (without emphasizing extremes) tends to make the noise murky and centrally-biased. Honey noise is an odd noise type that mixes a Value noise call and a Simplex noise call; because they have such different structures (honeycombs), it can reduce some artifacts. I hope some of this can be useful!
