I have the same problem, over and over again… I am using Millipede and Cocoon Marching Cubes.
I tried _AlignMeshVertices and _UnifyMeshNormals, as well es Combine&Clean and Mwesh WeldVertices. MeshRepair doesnt have any effect.
Actually I managed to have some results with AlignVertices Component, but it take quite long and assume for bigger meshes the render time scales linearly or worse.
I would appreciate swift guesses over solutions - need a fix urgently.
Also any suggestions on how to create faster mesh (marching cubes or else from line geometry) that have more equally sized faces would be as well highly appreciated.
Unfortunately your suggestions dont work. Any other ideas?
Also, could you help with smoothing the mesh itself? I tried MeshMachine and ReMesh, but the results dont have much effect. Again, when the meshes get bigger it computes too long.
Exosceleton works for simple geometries. It fails often from the start, because my line netowkr typology is too complex. The biggest problem are the nodes. Sometime of them are succesful and some fail. Cant recognize any pattern.
Any luck with cull unused verts after the weld (and maybe setting weld to .01)? I know @piac says it is not necessary here but I find it to solve this issue often for me (maybe I just have some dumb luck in my cases )
Do you see the orange component (in the first picture)? What is the balloon saying? (if you do not see it, Display -> Canvas widgets -> Message Balloons in Grasshopper).
There are other types of issues that a welding component cannot fix (stacked copied faces are one of them). I would suggest running the mesh through Rhino’s _MeshRepair, and see what it says.
Thanks,
Giulio
–
Giulio Piacentino
for Robert McNeel & Associates giulio@mcneel.com
Do you mean a different Weld Component than I used
Nope, I said Cull Unused Vertices After Weld with using the mesh edits weld which is a distance tolerance, not grasshoppers which uses an angle, and set that distance tolerance (t) to .01
But also check the mesh in _MeshRepair as Giulio suggests.
One of the Mesh Cleaning Methods works. But it takes too long. If yours or Michaels ideas can fix it, I would save a substantial amount of time and lower the risk of something jacking up over night.
In the end I want to create a biomorphic “thickened” geometry around a complex line network. If you or anyone has suggestions on how to do this more elegantly, that would be great.
So far, Coccon and Millipede works for me – Millipede fails more. I need to look into MeshLab, MCE https://sourceforge.net/projects/mce-marching-cube-eld/, Autodesk Netfabb, Digital Substance ISO Surfacing, but don’t find the time at this moment and probably wont at all. Unsucessful or insufficient for me was Exosceleton and Aether.
Honestly, if I were to do that I would just make some simple mesh pipes on the curves and take it right to ZBrush. Zbrush would figure it out in one nice clean mesh in a few seconds or so. Huge Rhino fan, but some jobs are not right for it.
Thanks for the suggestion. Downloading it now. I understand so far, that MeshPipes will be an input for DynaMesh and E Voila? And ZRemesher and Decimation I’d use to clean up and optimize afterwards.
Essentially yea, it will create a kind of shell mesh around those pipes. ZRemesher will make it have really good topology. Decimation master will try and reduce polygon count without loosing detail.
)
