# Whorled leaf curling with phyllotaxis

I'm trying to curl leaves/branches along a stem in the same direction whilst also having phyllo-taxis rotation. Rotating the instances or setting curve tilt messes up the curl rotation and I cannot figure out why.

• Is that what you want? i.stack.imgur.com/bwR1i.jpg Dec 19, 2022 at 19:37
• Yes however with curving/noise in the main stem. Last night I found your solution to curve tilting here blender.stackexchange.com/questions/271761/… and was able to utilize it in this case. Though this feels like a hack since I only partially know why this is working. I'll upload the new file and possibly mark this as my answer Dec 19, 2022 at 23:12

Even if my node group you used in your own answer could help you: It's not a good idea to use it in this case, since this was actually developed for a completely different use case.

I took a closer look at the issue, and you can actually solve this much easier.

You'd just have to structure things a little differently:

The biggest differences are as follows:

1. Instead of instantiating individual arcs along the curve and using their points to instantiate the branches, I just duplicate the points and use those as the base.
1. I apply the rotations of the branches directly during instantiation.
1. The distance to the trunk is obtained by a local offset.
2. The curling of the branches I also create, very similar to the answer before, with the relative position of the individual points to the anchor point of the branches.

However, the result in this case is: fewer nodes, better performance, but still the same idea.

(Blender 3.4)

• I knew you would have a much better way. You are really filling in the gaps in everyone's knowledge. Since there is no real consolidated source for geometry nodes training, and the nodes are evolving so rapidly I'm finding it very hard to learn. I find the Blender manual is not very helpful to us noobs. I think a comprehensive course on vector math and vector nodes would be the most helpful. The Vector Math->Subtract before the Cross Product is what I was missing in my trials. Dec 20, 2022 at 18:26
• @LGood I understand exactly what you mean. It's really not very clear in parts. But as motivation, I started learning Geometry Nodes in ~February 2022. My only source was the multitude of interesting questions here on BSE, which I just started answering. ...Since then I could not stop doing it. So there is a quite useful rule: Just keep going, it gets easier and easier! Dec 20, 2022 at 23:18

I think this is the solution..

It seems the issue relates to curve tilting as addressed in this post by quellenform

• OK, glad that worked out for you! I have never used this node group in such a use case. Still, it would work much easier in your case, and the answer has already been given once to one of your previous questions (just not quite applied correctly by you so far). But anyway, I'm happy for you that you were able to solve the problem. Dec 19, 2022 at 23:27
• Yes the previous question was very helpful but did not work when any amount of curve or noise was applied to the main stem - which is where your Align Curve to Vector node group saved the day. Dec 19, 2022 at 23:33