HOME | DD

Description

I tried to make brocoli in geonodes, and ended up with something more like agave cactus, I have named it the Bructus!
the right most is a 2 iteration of the fractal, the left 2 are three iteration version, which is the most I could get with my setup before blender crashed. (I think I could have gotten another few iterations be making a lower poly proxy of the 2nd iteration and baking a normal map for the ultra find detail, but that seemed like a lot of work for probably invisible detail)
the problem with the method here is that in real Romanesco Broccoli the size of the branches (the tiny bumps) seems to be basically consistent, with the size of the heads being determined by the number of branches.
In this method the previous iteration is arraigned in a spiral and then scaled based on its position, which means that the number of branches is fixed and the size of branches is variable, exactly the opposite of that goal.
The difference wasn't really visible at the 2nd iteration, but resulted in an increasingly different structure at higher iterations.

Description

I tried to make brocoli in geonodes, and ended up with something more like agave cactus, I have named it the Bructus!
the right most is a 2 iteration of the fractal, the left 2 are three iteration version, which is the most I could get with my setup before blender crashed. (I think I could have gotten another few iterations be making a lower poly proxy of the 2nd iteration and baking a normal map for the ultra find detail, but that seemed like a lot of work for probably invisible detail)
the problem with the method here is that in real Romanesco Broccoli the size of the branches (the tiny bumps) seems to be basically consistent, with the size of the heads being determined by the number of branches.
In this method the previous iteration is arraigned in a spiral and then scaled based on its position, which means that the number of branches is fixed and the size of branches is variable, exactly the opposite of that goal.
The difference wasn't really visible at the 2nd iteration, but resulted in an increasingly different structure at higher iterations.