0
$\begingroup$

So, I have a fairly simple control cage for what's supposed to be a smooth surface. IOW, it's getting a subdivide modifier slapped on it. However, it also needs some sharp corners... and thickness. So it's getting a solidify modifier also.

I can get the sharp corners on the "main" (non-rim) faces... but I can't get the rim right. The trouble is that I need to solidify first (due to varying thickness that results in my model looking like garbage if I subdivide first; basically, I need the subdivision to smooth out the transitions between thicknesses), but when I do that, the rim wants to bulge in at the pointy spots, because there is no edge crease. But just turning on edge crease makes the rim (and in fact, the entire edge of the shape) pointy everywhere, which is also wrong.

It seems like what Blender should be doing is directly copying the input vertex crease to the rim edge crease, which would solve the problem.

Is there some way to do that that I'm missing? Is there some other way to fix this mess? (Again, solidify must precede subdivision.)

Here's a quick sample of the sort of problem I'm trying to solve: example


This may be a duplicate of Subdivision and solidify sharp corners, but the only answer there isn't working for this case: applying a bevel modifier by angle won't work because some the corners (at least on my real model) that should be smooth are greater angle than some of the corners that should be sharp, and using 'complex' solidify (for some reason?!) removes the ability to set creases on the edges. I'd also really prefer to avoid beveling the sharp corners... while that "mostly" works, this model is ultimately going to participate in cloth simulation, and I'd really prefer to avoid the tiny faces that get generated by such method. Really, what I want is a way to just set the rim edge crease.

$\endgroup$
2
  • $\begingroup$ Can you spell out/illustrate the 'varying thickness -> garbage' part of the problem? $\endgroup$
    – Robin Betts
    Commented Oct 9, 2022 at 19:29
  • 1
    $\begingroup$ @RobinBetts, basically, I get linear interpolation of the heights rather than CC-smoothed. In the attached model, add every other vertex (checkerboard pattern) to a group and smooth based on that. The results with subdivision first vs. subdivision second are completely different. In particular, subdivision first gives a very jaggy/faceted "top", which isn't what I want. $\endgroup$
    – Matthew
    Commented Oct 9, 2022 at 19:59

0

You must log in to answer this question.

Browse other questions tagged .