Dear DX-users Can anybody tell me what the "Manhattan" algorithm of the Gradient module implies. I have never heard of this method. Is it just a central difference method (thus a second order scheme)? If not, what are the details (like the order of the scheme).
Thanks Eelco
