Variable Orientation - Interpolation or NN?
DhanielCarvalho Posts: 5
edited November 2020 in Leapfrog
I was checking the variable orientation results in a project with multiple surfaces as input for the VO.
It seemed that the dip/azi values are interpolated between inputs and I confirmed with a test.
As an example below (image 1), I wanted vertical/subvertical directions but LF creates a middle region with HORIZONTAL directions dips.
Even if I create a 3rd trend to fix the problem, it still gets confused (image 2).
I understand that if this was a fold then yes, this is acceptable.
In other software, this transferring could be manually adjusted accordingly using nearest neighbor instead of an interpolation.
What do you guys think of adding an option for NN instead of interpolation?
I think you’ll find that if you flip the facing direction of the left hand mesh, you will get the result you are looking for.
The VO is calculated by decomposing the normal vector to the surface into 3 component vectors and interpolating those. The facing direction is important to determine the sense of the normal vector.