Is there a way to program the same sparse noise in the Noise top in GLSL? I like the way this sparse noise looks better than perlin or simplex GPU noises.

Can anyone educate me in on some of the differences? (I have sometimes heard of sparse noise as “sparse gabor” noise - not sure if this is true in TD?)

I have started with the book of shaders noise functions and also looked at some of the LYGIA glsl noise functions - I have a decent understanding of noises in general but I’ve never seen much on the internet described as “sparse noise” and I would like to model my own sparse noise TOP in GLSL if i could… I’m curious what it is about the sparse noise that makes it not work on GPU?

for me there are 2 critical things that I like about this sparse (CPU) noise variant.

- that it only uses roughness instead of the harmonic gain + harmonic gain (sometimes harmonic spread is referred to as lacunarity) - Technically the single parameter for roughness might be a “downgrade” of control but I actually just like the way it works better.
- just the look of it. it feels more “cloth-like” when you look at the 3D noise pattern than perlin or simplex does so im kind of wondering what makes that so.

in terms of modeling the roughness vs harmonic gain/ spread is there a specific equation used to do that… most of the examples I find on the web are fbm type where they iterate through a for loop to generate the octaves which makes sense for gain+spread… but if you only have a single roughness control how does each iteration of the noise loop add/multiply to give final result?