> the documentation says that it is an adaptive slope algorithm called > variable gradients. if it isn't, then you should look at the paper > referenced from his site that shows the results and see if that is what he > really implemented.
Yep it is, as we said, it uses the VNG ("Variable Nunber of Gradients") algorithm. This is a relatively well established technique. Love, Light and Peace, - Peter Loveday Director of Development, eyeon Software