When using interpolation 2 (tricubic), a density_file pattern may "overshoot" if
the DF3 files has particularly high "plateaus" with steep gradients, wrapping
the pigment around to zero.

I think this requires fixing. From the density_file pattern code, I gather that
the wraparound could be suppressed by setting the "frequency" property to zero.
I also noticed that this property *is* set to zero for truly image-based
patterns, maybe for the very same reason.

Post a reply to this message