Hi Will,

> I'm certainly not a math wizz so I don't know if the algorithms
> could be reworked to work with NURBS surfaces or not -

As far as I can tell it is definitely on the "not" side, I'm afraid to say.

Maybe with additional research in the future they will be able to make it more adaptable in that direction.

Possibly someone with a whole bunch of specialized reverse-engineering type of NURBS experience might be able to generate NURBS output from that, but it would be a very difficult task.

To turn something naturally into NURBS it needs to have a very regular quad-like layout to it, NURBS surfaces are a rectangular net of control points. If you see something that is a whole mess of points connected in no particular apparent order, that then is not something that has a very natural automatic conversion to NURBS.


> it "appears" like it would be compatible since it's curves that are
> used as input to the system, similar to NURBS...

Yeah, but the curves get diced up (intersected with a triangle grid) at one point.

Just because something initially starts out as curves doesn't really mean that it produces similar structured objects out the other end...

- Michael