Point Deform
The Point Deform works like the Houdini Point Deform SOP: every point of the mesh is captured by the lattice points within Radius (up to Maximum Points, closer ones weigh more), then follows the blended local transforms of those lattice points between the Rest Lattice and the Deformed Lattice. Rigid Projection keeps only their rotation, and N, quaternions, 3x3 and 4x4 transforms and the other Attributes to Transform turn with the deformation, or Recompute Affected Normals computes the normals again. The capture is made once and kept while the mesh and the rest lattice stay the same, so a simulated lattice only costs the deform each frame.
Group limits the deformation to part of the mesh, Mask scales it (optionally by a point attribute), and Mode can split the work into a Capture step and a Deform step. Piece Attribute keeps separate parts from pulling on each other, and Match by ID pairs the two lattices by an ID attribute when the deformed lattice is not in the same point order. A typical use is Remesh to a low-poly lattice, simulate the lattice with cloth, soft body or Nexus, and let Point Deform move the full-resolution model with it. Display Lattice draws the lattice in the viewport.
POP - Mesh to DeformPOP - Rest LatticePOP - Deformed LatticePOP - Deformed Mesh
