4
$\begingroup$

My problem, as you can see on the image below, is basically that the subsurf interpolation of the centered hole in that mirrored plane does not transform into a circle as intended.

I know that this is completely correct, since it just takes into account the additional vertex on the mirror axis, but is there any clever way to fix or work around this except fiddling around with the vertices manually?

Three quad holes on a mirrored subsurf object

$\endgroup$
5
$\begingroup$

A quad won't give you a real circle anyway. If you look at your example, the outer holes are still bulging a little in the corners. That's due to the algorithm and can't be helped. Six or 8 sides are much more precise. So, since you have 4 (6 with mirroring) verts in the middle, turn it into a hexagon or use 8 to have a nice clean result. Depending on what you want to do with the hole, an additional facering around the holes can be helpful as well as makeing a nice and undistorted bevel later.

enter image description here

| improve this answer | |
$\endgroup$
  • 1
    $\begingroup$ Thank you for the info about the "not-a-circle" problem; i also noticed the bulges, but i thought it was something with my eyes. $\endgroup$ – cdx Oct 17 '13 at 21:57
  • $\begingroup$ One is glad to be of service. Btw, even 6 or 8 verts do not make a mathematically correct circle, but it's close enough 99 percent of the time. The difference to a real circle for 8 verts is roughly 1/1000 of the diameter of the circle. $\endgroup$ – Haunt_House Oct 17 '13 at 22:24
2
$\begingroup$

You can turn the topology around to use a triangle mirrored to get a circle.

enter image description here

In your example after the mesh is mirrored you have a whole with six vertices around it, which is causing the distortion. you need to consider the mesh as it is generated after the mirror modifier.

| improve this answer | |
$\endgroup$
  • 1
    $\begingroup$ this does indeed work, but the problem is that i cant put the holes close together since the minimal distance between them in x-direction will be twice the distance between the hole and the vertex on the corner $\endgroup$ – cdx Oct 17 '13 at 9:57

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.