In Shaded view I set to use single color for backfaces:
It works great until I join surfaces in different direction:
into closed polysurface:
Now it shows backfaces totally transparent and _dir command can’t flip individual faces.
poysurface-dir_3.3dm (169.6 KB)
P.S.
Originally, I had this issue in Rhino8 but for whatever reason I’m no longer able reproduce it there, only in Rhino9
Why are you trying to create bad geometry with normals pointing the wrong way?
I’m not “trying”, I use _sweep2 command a lot which requires specific “clock-wise” sequence selections to control which direction the surface will be created and until I discovered the setting for backface color it was very difficult to know if created surface was the correct direction, and _dir command is a mess when it comes to closed polysurfaces, nearly impossible to see what direction each face oriented (it’s fixed in latest 9.0.26251.12303, 09/08/2026)
Just saying that perhaps closed polysurface should automatically turn faces the correct direction…or at least give us ability to turn individual face with _dir command.
It does… or at least it’s supposed to. If Join, BooleanUnion or any similar command doesn’t automatically unifies all the normals of a closed polysurface, it’s most certainly a bug.
-Jakob
AFAIK it is usual for Rhino to display backfaces as transparent on polysurfaces where the faces have inconsistent normals.
_dir can’t flip individual faces of a polysurface, only the whole thing.
Intriguing that you were able to create this polysurface with mixed normals - can you remember step by step how you did it?
And I presume you know that if you explode it and then select the pieces and join them the result has correct normals throughout?
The screenshots are made with exploded and then joined surfaces, so no, it does not automatically correct normals neither in Rhino8, nor Rhino9beta
Rhino8 however shows the blackface color correctly instead of transparent. The attached file above shows differently on my Rhino8 and Rhino9beta.
The second screenshot shows you ran _Undo on the explode, not _Join…
That implies you were reverting to bad geometry, not creating a new clean polysurface. If I run _Join on your exploded surfaces they do join with normals standardised in both R8 and R9.
