Re: Friday Facts #112 - Better noise
Posted: Thu Nov 26, 2015 4:34 pm
Can you clarify what you mean by "complete noise"? How is it obtained from the original hash function?
The new noise is similar to Simplex noise in this: The hash is used to select one of predefined gradient vectors in each cell vertex, then dot product of the gradient vector and position in the cell is weighted by a surflet function ((1 - min(1, position_in_cell_vector**2))**3) and the values for each vertex are added in the cell. Easy :-)Holy-Fire wrote:Can you clarify what you mean by "complete noise"? How is it obtained from the original hash function?