It's explained in the README. The SDF is evaluated on whole batches of points (32x32x32 tiles by default) using numpy. (And it's multithreaded.)
It's explained in the README. The SDF is evaluated on whole batches of points (32x32x32 tiles by default) using numpy. (And it's multithreaded.)
So uniform sampling? And then I guess marching cubes or sth?
I see you still haven't read the README.
marching cubes yep.