reading some good shit right now: "Monte Carlo Geometry Processing: A Grid-Free Approach to PDE-Based Methods on Volumetric Domains" https://www.cs.cmu.edu/~kmcrane/Projects/MonteCarloGeometryProcessing/index.html">https://www.cs.cmu.edu/~kmcrane/... (thanks to @Alleycatsphinx)
duh, it never occurred to me that we could take random points on the sphere to make spatial queries. i always projected the local gradient, like a nerd.
coming down from reading this paper. i need to rethink my life.
of course @iquilezles has already played with it. here& #39;s two 2D shadertoys trying it out https://www.shadertoy.com/view/WsXBzl
https://www.shadertoy.com/view/WsXB... href=" https://www.shadertoy.com/view/WdXfzl ">https://www.shadertoy.com/view/WdXf...
https://www.shadertoy.com/view/WsXB... href=" https://www.shadertoy.com/view/WdXfzl ">https://www.shadertoy.com/view/WdXf...
and one shadertoy in 3D https://www.shadertoy.com/view/wdffWj ">https://www.shadertoy.com/view/wdff...