Hello all: I'd also be happy to help implement a "smear" (a.k.a. nearest neighbour) abyss policy, which preserves the property of being exact on constants (as the "truncate the stencil and normalize to unit sum does as well, although I think that going through an abyss policy is more elegant and faster).
In the case of YAFR for example, it would ensure that the transition through the boundary is continuously differentiable, without being as complicated as the bilinear abyss policy and without the need for "stick outside points to the boundary" to look good far from the valid pixels. Nicolas Robidoux Laurentian University/Universite Laurentienne _______________________________________________ Gegl-developer mailing list [email protected] https://lists.XCF.Berkeley.EDU/mailman/listinfo/gegl-developer
