1
$\begingroup$

Im trying to make some dice but one of the shapes is not subtracting from the cube properly. The sun on the right side is only partially subtracting.

I started with some svg vector saved from inkscape, imported and then extruded them. My plan was to subtract.

Any help?

subtraction not working

Edit:

Edit:

Sorry I never Checked my Curves from Inkscape properly. I had some unlinked shapes. I would have thought that blender would say. "Yo, dude... your vector sux"

One of the vector files I saved out from Inkscape had some duplicate nodes that were not linked.

I went back into Inkscape and tidied up the vectors and made sure the shapes were full complete enclosed shapes with no duplicate nodes

$\endgroup$
5
  • $\begingroup$ In theory, this should work. Upload your blend file to blend-exchange.giantcowfilms.com so we can take a look. $\endgroup$
    – Leander
    Commented Oct 8, 2016 at 11:36
  • $\begingroup$ if booleans act weirdly, the problem is usually always on normals (all should point outside), but also check before that you have not a negative scale... that would invert normal visualization... $\endgroup$
    – m.ardito
    Commented Oct 8, 2016 at 17:57
  • $\begingroup$ @Leander blend-exchange.giantcowfilms.com/b/2058 $\endgroup$
    – zenril
    Commented Oct 8, 2016 at 23:28
  • $\begingroup$ @m.ardito I'm not sure what negative scaling is. I have been googling it still seeing if i can figure it out. I have Checked the normal's of the sun all the lines point outside. Although all the normal's of the cube were pointing inside. I had to Mesh > Normal's > Recalculate Inside to get them pointing out.... howver this still does not work. $\endgroup$
    – zenril
    Commented Oct 9, 2016 at 8:01
  • $\begingroup$ Sorry, I meant: in Blender you can set scale to -1. It makes no sense to me, but it is possible and can happen by mistake. When you do that, normals that visually seem to point "outside", are really pointing "inside" (due to negative scaling), thus what you see is wrong, and booleans logic is inverted... $\endgroup$
    – m.ardito
    Commented Oct 10, 2016 at 8:23

1 Answer 1

1
$\begingroup$

I think I found the problem with that boolean: duplicate vertices!

I've hidden other shapes, to better see the two objects involved by the boolean, and here is you wrong result:

enter image description here

then I selected the "sun" shape, in edit mode and removed duplicates

enter image description here

It had 78 duplicate vertices:

enter image description here

And, after that, the shape works better with booleans...

enter image description here

$\endgroup$

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .