I’ve been working on a model-wrapping project (topology transfer) in Houdini, and I’ve run into an issue with preserving the point count and order after separating, transforming, and re-merging parts of a mesh. I’m wondering if anyone else has faced this issue and can offer insights. 😢
The Setup
Separate Parts Using Blast: I use a
Blast
node to isolate specific parts of the mesh, focusing on the “InnerMouth” group.
Apply Transformations: I move the “InnerMouth” part to a new position.
Merge Parts Back Together: After moving, I use a
Merge
node to combine the “InnerMouth” with the rest of the model.
The Problem After merging, I’ve noticed that the point count in the merged model is higher than the original! I attempted to correct this using a
Fuse
node, but Fuse only works if the “InnerMouth” part remains in its original position. Moving the part makes Fuse ineffective, leaving the point count incorrect.
Strangely, this issue only happens with the “InnerMouth” group. Other groups work fine even after transformations, which makes me wonder if it’s a model-specific problem.
Edited by BarryAllen_WalleyWest - Oct. 27, 2024 10:35:57
My goal in simple words is: I have 2 models, one is A, one is B(target). I need to wrap A to B, which will result in a model having A's topology and B's shape. I blast the inner mouth part, so the wrapping process will leave out the inner mouth. After the wrap, I want to move the inner mouth part(using point deform), to match B's position, without manually use the edit node(because the point deform is automatic, less manual adjustment is good). But the problem is, if I merge, there will be extra points and it'll fail to create a blendshape
Without seeing the geo I can only guess but if the "inner mouth" and the rest of the body are a continuous mesh then of course blasting a section will duplicate the points along the edges where the mesh was split.
Can you utilize the rigid primitives group on topotransfer? That way you don't have to separate the mouth and it should come along for the ride without being deformed. Then you can adjust it's placement after.
You can also Point Deform the whole original A using your rest and deformed Geos without inner mouth instead of deforming just mouth
And if needed you can use different point deforms on different groups to have the outer part match exactly without smoothing or even post blend that part directly
freshbaked Without seeing the geo I can only guess but if the "inner mouth" and the rest of the body are a continuous mesh then of course blasting a section will duplicate the points along the edges where the mesh was split.
Can you utilize the rigid primitives group on topotransfer? That way you don't have to separate the mouth and it should come along for the ride without being deformed. Then you can adjust it's placement after.
Yes the "inner mouth" part is onnected to the lips, I found out they share the same points, that's the reason. Thanks for the advice! I don't really see too much of a difference between specifying the rigid primitives group and not, the result looks very familiar to me~
tamte You can also Point Deform the whole original A using your rest and deformed Geos without inner mouth instead of deforming just mouth
And if needed you can use different point deforms on different groups to have the outer part match exactly without smoothing or even post blend that part directly
Thanks! I've actually tried to deform the whole orginal A model But after I deform the whole original A, how to merge the "inner mouth" from the deformed A, to the deformed Geo(A topology, B shape)? I fear it will still create more points( when merging the A "inner mouth" and the deformed Geo without inner mouth).
And also, I don't really understand what this mean
And if needed you can use different point deforms on different groups to have the outer part match exactly without smoothing or even post blend that part directly
Does it mean that, I can use a point deoform, select the "Inner Mouth" only, and then deform? And what does "the outer part match exactly" mean?
