Loft unexpected CV and Isocurve Position. (no Bug - understand CV-Weight)

hello (mainly) Mc Neel-Team, (i thing this is a bug)

hello (maybe) Everybody

Attached find a file with two R1.5mm arcs that lay on two parallel planes.
(deviation to 100% parallel 1.4872e-08 mm)
If I _loft those two curves i get a surface with unexpected Iso-Curve(s) in the middle.
(both straight and normal option)
I get similar unexpected results.
I get same results in Rhino 5 and Rhino 6

I found out, that the unexpected result has something to do with the weight of the CVs.
if i do a loft with option straight and then set the weight to 1 for all CVs the Isocurves are as expected.

But why does the different weight of the CVs of the input Curves (2 Arcs with different Angles) influence the position of the interpolated CVs in between ?
For me this looks like a bug - but maybe it is also my miss-understanding of the Nurbs Model ?

Thanks for a explanation or a bug-tracking.

Best regards - Tom

Loft_sweep_problem_00.3dm (71.1 KB)

Try Loft with the Straight sections option. No need to change the weights to 1. That will create a surface which is degree 1 and a single span in the U direction which I assume is what you expect. Loft with other options creates a degree 3 curve in the U direction. My guess is that is how Loft is intended to work.

Dear David

Interesting. Attached is a file with a Loft with Straight sections surface I created using your two arcs. It is degree 1 single span in the U direction. It does have an isoline in the middle but remember that isolines are for display purposes and can be misleading as to the structure. In contrast the surface in your Loft layer is degree 3 single span in the U direction, and I created a surface identical to that surface using Loft with Normal option. Perhaps you might verify that when you use Loft with Straight sections it gives a different result than I had.
Loft_sweep_problem_00DC2.3dm (100.4 KB)

Hi Tom - I do not have a good answer other than to point at the different weights of the middle points.If you extrude one of the arcs and then give the middle point at the far end the same weight as that on the non-extruded arc, the isocurve does the same as in the loft.


thank you a lot both Pascal and David.

this is a simple, planar U Degree = 1, V Degree = 2 surface.
the corner-Points have weight 1
the “midpoints” have weight 0.70 and 0.86 “inspired” from the arcs next two them.
isocurve-Density set to 9

i think the following explanation works (at least for me):
a controlpoint with more weight is influencing with more “force” resulting in some kind of “acceleration” - meaning there is not a continuous “speed” inside the u-Direction of the surface.
(in the middle of the surface / at the cyan isocurve)
(and i am not a strong mathematician - otherwise i could map this “behaviour” to derivations of the underlying polynomials that “mix” the coordinates ).

for me the question is solved - and my understanding of nurbs has increased. thanks.