Timeline for Is it possible to animate a transition between two XYZ Math Surfaces?
Current License: CC BY-SA 4.0
22 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Jun 13 at 19:15 | vote | accept | Lawton | ||
| Jun 13 at 6:13 | answer | added | Harry McKenzie♦ | timeline score: 1 | |
| Jun 11 at 15:40 | history | edited | Harry McKenzie♦ | CC BY-SA 4.0 |
added 16 characters in body; edited tags
|
| Jun 11 at 13:08 | answer | added | Harry McKenzie♦ | timeline score: 5 | |
| Jun 9 at 22:13 | comment | added | Harry McKenzie♦ | @Lawton if you are still new to Geometry Nodes I highly recommend to go through some tutorials first to get a good understanding of it. Here's a good one to get you started youtu.be/ZerJnivvBn4?si=DNSIamTUrbkzzJIj | |
| Jun 9 at 17:28 | comment | added | Lawton | @HarryMcKenzie As for eliminating the overlaps at the generation stage instead of manually deleting vertices, nothing I tried worked within the limitations of XYZ Math Surfaces. That might not be the case when using Geometry Nodes, but I have no experience with that feature yet. | |
| Jun 9 at 17:26 | comment | added | Lawton | @HarryMcKenzie Can you explain how to "build the node tree for each equation" for someone who has never used Geometry Nodes before (or provide a link to such an explanation)? The screenshots in the linked answers seem to imply that a separate node is required for every single mathematical operation in each equation; is that correct? I'm also not sure from the screenshots alone how the node trees in them have been constructed. | |
| Jun 9 at 10:07 | comment | added | Harry McKenzie♦ |
I think the task and node tree is way too big for 1 question. It needs to be simplified and more focused. Basically you just have to build the node tree for each equation using this method where g will be an input parameter just like a in that example. and then you can easily remove overlapping geometry using this method. And once all geometry is joined you can keyframe the g input node to animate morph it.
|
|
| Jun 9 at 5:31 | comment | added | Robin Betts♦ | I've had no luck, sorting, so far, with a mesh of this density.. analogous meshes at lower counts are working for me. | |
| Jun 8 at 19:10 | comment | added | Lawton | @HarryMcKenzie The "these questions" link doesn't work. (Are lists public, or can they only be viewed by their creator?). In answer to your question, I believe you missed the line "To get the complete mesh, the surfaces whose names end in "Fore" or "Aft" must be duplicated twice, with each duplicate rotated 120° about the z-axis." | |
| Jun 8 at 15:37 | comment | added | Harry McKenzie♦ |
@Lawton Aside from the overlaps mentioned, I'm not able to obtain a complete sphere when I set G=0 for all your 6 xyz math surfaces as seen in this link. Am I missing something? If you were able to complete the object surface you could easily plot each mesh using Geometry Nodes and merge and animate the G parameter. See these questions on examples to plot different types of equations. Then this one in particular shows how you can animate the graph.
|
|
| Jun 8 at 13:45 | history | edited | Lawton | CC BY-SA 4.0 |
Added Blend-Exchange link.
|
| Jun 7 at 16:07 | comment | added | Robin Betts♦ | @Lawton.. OK..I'm guessing there will be a Geometry Nodes way to sort you out, by 'Proximity' or 'Sort Elements' for example. If you'd like us to have a go, maybe you can share your assembled meshes to save us some trouble, on blend-exchange.com ? | |
| Jun 7 at 14:40 | history | edited | Lawton | CC BY-SA 4.0 |
Added GIFs of the meshes.
|
| Jun 7 at 14:10 | comment | added | Lawton | @RobinBetts I agree that it seems like a linear interpolation should work fine. The only issue with using Shape Keys seems to be that Blender isn't correctly matching up the vertices of the starting and ending meshes. | |
| Jun 7 at 14:09 | comment | added | Lawton | @StefLAncien The overlap between surfaces doesn't change as G changes, so the topology should be identical across the full range of values. I agree that it seems like a linear interpolation should work fine. The only issue with using Shape Keys seems to be that Blender isn't correctly matching up the vertices of the starting and ending meshes. | |
| Jun 6 at 20:10 | comment | added | Robin Betts♦ | @StefLAncien is exactly on it.. in all these expressions as given, g is a linear, in fact, a scaling component, so a Shape-Key should work,here. g is mix-factor? | |
| Jun 6 at 19:41 | comment | added | StefLAncien | Could you confirm that the expression for X, Y, and Z looks like (e.g. for X) X(F,G,U,V) = F x ( G x X1(U,V) + x2(U,V) ) ? It is not so obvious from the provided pseudo-code... If it is correct, it means that for the same topology, the transformation is linear, and as Robin mentioned, a linear interpolation should to the job. UNLESS the process written as the portions of the "Face" and "Vertex" surfaces which overlap with the "Edge" surfaces should be deleted is changing the topology as G is increasing from 0 to 1... | |
| Jun 6 at 17:23 | history | edited | Lawton | CC BY-SA 4.0 |
Added the setups for the XYZ Math Surfaces.
|
| Jun 6 at 16:41 | comment | added | Robin Betts♦ | Shape-keys will give you a linear interpolation between the surfaces at your sampled values of G. so, no, the transition will generally not be the same as varying G. The good? news is that you could certainly do this with Geometry Nodes. If you gave us an example function, we could demonstrate. | |
| Jun 6 at 16:32 | comment | added | Gordon Brinkmann | Are you sure the structure between both objects is exactly the same or is it an assumption? Because it does not sound like they were. But maybe it would help if we knew what you are creating? | |
| Jun 6 at 16:22 | history | asked | Lawton | CC BY-SA 4.0 |