# Getting something close to the convex full of SDFs using CSG operations

Inigo Quilez has an explanation of smooth operations to join sdfs.

I am trying to get something close to the convex hull of multiple sdfs. For starters let;s consider a simple case of 3 cubes int eh following configuration:

The convex hull would be a very, very similar shape, with the exception that the region inside the l shape would be a diagonal slope.

I tried using smooth joining operations with a high weight, but I get this

The area I expected to be filled is indeed filled, but everything else bloated up. I am trying to come up with a mathematical formula to test if a point would be inside the convex hull.

• Here's a counterexample: take 2 spheres pretty close to infinitely far away. Then the cylinder between those is part of the convex hull. The CSG operations on the other hand will not reproduce it. Jun 13, 2022 at 3:44
• wdym? On that specific example you can use extrusion to get the cylinder. Jun 13, 2022 at 4:29
• The point was that smooth join will not work even for very simple cases. Jun 13, 2022 at 4:49
• I understand, I am asking if there is a way to get something close to the convex hull, even if it only works when the shapes are within a certain distance or something along those lines. Jun 13, 2022 at 4:56