SubD bugs in Rhino 8

Rhino SubD
,
1

Hi all.
Reporting 2 minor bugs about SubDs.

  • 1

Toroidal SubD, 1 whole loop hard crease, 1 soft crease.
Moving one face result in wrong meshing near the vertex intersection from hard crease and soft crease:
2024-02-16 23_29_57-Window
sub crease meshing bug.3dm (146.5 KB)

EDIT:
this actually result in an open brep/surface after ToNurbs:
2024-02-17 00_14_32-Window
the surface has the same control points structure as if the soft crease didn’t exist.

  • 2

Toroidal SubD, 1 whole loop hard crease.
ToNurbs() output results in a single closed, periodic surface, as expected… but the seam is in a wrong place:

2024-02-16 23_31_59-Window
2024-02-16 23_31_59-Window448×590 122 KB

subd hard crease tonurbs seam bug.3dm (243.6 KB)

Iso-trimming a part of of the surface:
2024-02-16 23_37_04-Window
sharp without edge, quite odd…

Normally with SubDs I expect any hard crease to convert to a surface/brep edge after ToNurbs…

  • 3
    I take this occasion to ask once again for a manual SubD packing for ToNurbs feature.

Thanks for the support, have nice weekend.

1 Like
2

Hi Riccardo - I see that on your example, I cannot reproduce it from scratch so far.

-Pascal

3

sub crease meshing bug 2

2 Likes
4

… @pascal were you able to replicate this?

…both?

My initial post is about 2 bugs, mainly…

Thanks.

5

@dalelear any ideas about this?

I can reproduce it here, odd thing is if you move one face the problem appears, BUT if you move a second face with the problem shown, then undo it, the problem disappears.

my soft crease is set to 90

1 Like
6

I think I’ve fixed this bug and the fix will ship in 8.6.
I’ve added https://mcneel.myjetbrains.com/youtrack/issue/RH-80676 to make sure I test it.

3 Likes
7

It turns out https://mcneel.myjetbrains.com/youtrack/issue/RH-80676 is not fixed by the recent changes. This is a bug in the code that updates render meshes. The SubD itself is fine - only the mesh used to display it is bad. I’m working on fixing this bug now.

8

Hi Dale, thanks for the support.

With this

are you referring at the bug i’ve reported here?

Because it also applies to the brep/surface result of ToNurbs command:

Those control points ^ are of the brep/surface, not the SubD’s …
My guess is it is a problem somewhere before the meshing step.
But it’s just a guess from seeing from outside…

Output ToNurbs brep seems to be correct, but it’s only a “pulled” edge, if you launch _Explode and _RebuildEdges you can see the same defect in the actual surface.

(in the proposed steps to recreate the bug, instead of moving a face, extrude one, so you can easier explode the brep, do rebuildedges, see…)

Better than many words:
subd bug

9

The root cause of the gaps in ToNurbs and the display mesh were identical. Minutes ago I submitted a fix to our version control system. If the fix passes testing, it will ship in 8.6.

I apologize for the inconvenience. Thank your for taking the time to investigate this bug and provide such a clear reports of the problem.

Feb 29 UPDATE: I’ve tested this bug RH-80676 (and 3 related bugs RH-76871, RH-80633, and a report that has a model we cannot share publicly) in our daily internal release (8.6.24060.6001, 2024-02-29).
This problem and the related problems look like they are all fixed and the fixes will ship in 8.6.

3 Likes
10

works here in internal testing…I’m calling this fixed in 8.6