🧵 Untitled Thread
Anonymous at Sat, 14 Sep 2024 10:12:18 UTC No. 995236
I have this kind of 2d/3d crossover geometry with flower stems being planes and flower heads being 3d shapes, so at some point I have plane edges joining 5 gons and this seems to break smooth shading, because 3d stuff can have blue normals on all faces, but 2d plane will always have a blue (front normal) and red (back normal) parts and flipping it will just change which one is blue and which one is red. And the thing is when red face is neighbouring blue face it creates a shading artifact. Is there a way to make smooth shading work with this geometry?
Anonymous at Sat, 14 Sep 2024 16:50:23 UTC No. 995258
>>995236
Two Data Transfer modifiers.
Duplicate your stem object and flip its normals. THis object can be deleted once you've applied all modifiers.
Split the bottom edge of your 3D shape where it is connecting to the plane.
Assign a "Front" and "Back" vertex group to every side (one side only A, one side only B). Only select the bottom vertices.
Set up two Data Transfer modifiers where one grabs the normal (Face Corner Data, Custom Normal) of stem 1 and vertex group A, the other vice versa.
Here's a blend file that explains it better:
https://www.mediafire.com/file/tbn4
Anonymous at Sat, 14 Sep 2024 17:45:42 UTC No. 995262
>>995258
Oh nice, that is tricky. Thank you, not really sure I can make it work fully like I want since I also want it to be a single mesh with subdivision surface modifier, and also want it to be kinda exportable to game engines and stuff. Nevertheless this is a cool trick to know.
Anonymous at Sat, 14 Sep 2024 17:52:58 UTC No. 995263
>>995262
No problem, glad I could help.
🗑️ Sacred G at Mon, 23 Sep 2024 23:51:42 UTC No. 996228
As far as like a 3D geometry rendering software capabilities could you 3D model the actual geometry of a DNA strand and then scale out and see how that math would make the actual stem structure look cuz then in theory you could build things in DNA sequences and then zoom out and see the organism as it is and unless you edit the DNA code sequence as above so below you would see the whatever your designing simulated
Anonymous at Wed, 16 Oct 2024 08:13:32 UTC No. 998501
Hmm, tricky issue.
Anonymous at Wed, 16 Oct 2024 08:13:37 UTC No. 998502
Hmm, tricky issue.