I’ve a really easy question for you, not sure about the easiness reply:
What’s the fastest and best method to pass from Metaballs isocurves to Meshes in 2020 with Rhino 6?
So going from this
I’m trying to create and update thread for this interesting topic.
I did my reasearch and I’ve read a lot of topics:
Here they are suggesting cocoon (too slow) and chromodoris (for rhino 5, I’ve to check if its compatible with rhino 6).
After they are suggesting this magic “Ball Pivot” component, but I’m not finding it (is it related to “The Milkbox” but honestly I’ve no clue how to download and use it).
Here there is an onld good component by @david stasiuk in VB, quite old and not that fast. it’s an “okay” solution.
Here appears for the first time the MeshFromPoints component did by @Tudor_Cosmatu, but it’s for Rhino and for Rhino 5 moreover.
There is a link redirecting to a veeery old thread by @DavidRutten (11 years ago)
But I’ve found no explanation.
someone is asking also about surface, a solution seems to be passing through Rhino, but I wanted to stay in GH only.
Last thread is linking both MeshFromPoints both Ball Pivot.
Maybe there is already an update solution to this problem, but it’s quite well hide in the GH’s Q&A world
Rather than generating the section curves of metaballs, and then getting an isosurface of these curves with Dendro, you could generate the metaball isosurface directly from the points. This should be much faster. dendroballs.gh (8.8 KB)
In the example you send me that volume offset was taking 42s don’t know the PC u have but for the sake of the project was too much… the solution of @DanielPiker with volume from points seems to be quite good for time saving.
Cool discussion the one u sent me about Connolly Boundary, during my research I didn’t find it. And cool solution u gave him!
Dendro is sensible to the size of Voxel, I put a lower value than Daniel. Negative Offset needs better surface quality so low voxel size. I am sure that for the same size of Voxel Cocoon will be longer than Dendro. But Cocoon will be better quality for Metaballs. It is difficult to have everything.
It would be nice if there was a tool for generating isosurfaces that would let users define custom scalar fields.
Dendro seems to be limited to linear falloff, so you get something just like a boolean union of spheres (and to get something smooth you need to do further smoothing operations), while Cocoon uses I think inverse square falloff, which gives you proper metaball shapes directly.
There are all sorts of other interesting scalar fields to explore though (see all the stuff people used to do with K3dSurf). I imagine something using an interface, where the user could define in a script any function that returns a scalar value for a point in space (using any combination of equations and geometry functions), and then plug that into the isosurfacer and get a contour at any value.