Houdini 21.0 Nodes Geometry nodes

Point Capture geometry node

Generates connectivity aware capture weights for use with Point Deform.

On this page
Since 19.5

Overview

This node computes connectivity-aware capture weights and creates smooth deformations using small capture radii. A small capture radius allows for a reproduction of animation while preventing excessive bleeding between neighboring features. For example, the fingers on a character’s hand.

This node is designed to deform a detailed surface such as a render mesh using an animated coarse surface (ex. an FEM or Vellum simulated mesh). The coarse surface can be either a polygonal mesh or the (implied) surface of a tetrahedral mesh. This node generates the best results when the detailed surface is in close proximity to the coarse surface, and when both surfaces have roughly the same shape.

When bleeding or local chunking occurs with a Point Deform SOP node’s geometry capture, you can use this node instead of the Point Deform SOP's internally generated capture weights. This node’s capture method generally yields smoother results than the Point Deform SOP.

Note

If your capture geometry is high resolution, then this node’s capture method may perform slower than the Point Deform SOP method.

Tips and notes

  • The topology of this node’s capture geometry and animated geometry must match. Point positions can change, but primitive types and their connectivity must remain constant. It’s recommended to not use this node for capture geometry that has a very different shape or topology from the mesh it should deform. For example, deform a skin surface with a collection of muscles.

  • This node generates the best results when the target detailed surface is in close proximity to its coarse surface.

  • Make sure to set the Point Deform SOP node’s mode to Deform, otherwise it will not use the capture weights generated by this node.

  • You can turn off smoothing for this node by setting the Smoothing Method to No Smoothing.

How-to

To...Do this

Deform a high-res mesh using a low-res mesh

  1. Make sure that you have the following in your current scene:

    • The high-resolution geometry you want to capture and deform with a Point Deform SOP node. This can be a polygonal mesh or the (implied) surface of a tetrahedral mesh.

      Example: High-resolution surface to deform

    • The animated low-resolution version of the high-resolution geometry. This can be a polygonal mesh or the (implied) surface of a tetrahedral mesh.

      Example: Animated low-resolution surface to drive the deformation

  2. Make sure the high-resolution geometry and the low-resolution geometry are positioned close by each other in your scene.

  3. Create a Point Capture SOP node.

  4. Do the following:

    • Supply the static rest geometry for the capture by connecting the output of your high-resolution geometry’s node to the first input (input 1) of the Point Capture SOP node.

    • Supply the static capture pose for the capture by removing any time dependencies from your animated low-resolution geometry and then connecting its output to the second input (input 2) of the Point Capture SOP node.

    • Supply the animated pose for the capture by connecting the output of your animated low-resolution geometry’s node to the third input (input 3) of the Point Capture SOP node.

    Example: Point Capture SOP in the capture and deformation network

  5. Adjust the Point Capture SOP node’s parameter settings as desired. This will be the initial pass for adjusting its settings.

  6. Create a Point Deform SOP node.

  7. Connect all the Point Capture SOP node’s outputs to all the inputs on the Point Deform SOP node you just created.

    Example: Point Deform SOP in the capture and deformation network

  8. Select the Point Deform SOP and then go to the Parameter Editor.

  9. In the Point Deform SOP node’s parameter settings, set the Mode to Deform.

    Example: Point Deform SOP’s Mode parameter

    The Point Deform SOP node now deforms the high-resolution mesh using the capture weights generated by the Point Capture SOP node.

  10. Adjust the Point Capture SOP node’s parameter settings to obtain the desired capture result. This is your refinement pass for adjusting its settings.

    Tips

    • If you are seeing geometry bleed through or chunking in your capture result, decrease the Capture Radius.

      Example: Effect of adjusting the Capture Radius

    • If the surface of your capture result appears too rough (bumpiness, rippling, or other artifacts), increase the Smoothing Level.

Example: Deformation with Point Capture SOP generated capture weights
Example: Deformation with Point Deform SOP generated capture weights

Parameters

Group

Specifies which points on the model (input 1) to generate capture weights for. You can use the arrow select button to the right of this field to interactively select the points in the viewport. If you want to generate capture weights for all points on this node’s first input, then leave this field empty.

Capture

Capture Radius

Sets the maximum distance (in Houdini world units) away from each projected location on the capture geometry to look for connected primitives. Choose a smaller capture radius to get a more accurate reproduction of local animation.

Kernel Type

Specifies which function to convert capture distances to capture weights. The kernel function is integrated over the capture radius on the surface of the capture geometry. The resulted integral is distributed to primitives intersecting the capture radius. The shape of the kernel function determines the relative influence and which points are close or far away.

Exponential Bump

Smoothest option. Works well for most cases.

Truncated Gaussian

Less smooth than Exponential Bump, but faster to compute.

Quadratic

The falloff is the normalized squared distance from the capture sphere boundary. Uses a piecewise quadratic function.

Linear

The falloff is the normalized distance from the capture sphere boundary. Uses a piecewise linear function.

Smoothing

Tip

You can turn off smoothing by setting the Smoothing Method to No Smoothing.

Smoothing Method

Specifies which method to create smooth deformations when a small capture radius is selected.

Approximating

Produces the smoothest deformations, but follows the animation less accurately.

Interpolating

Follows the animation as accurately as possible, while still providing extra smoothness for a low capture radius.

No Smoothing

Captures without smoothing. However, when working with a small capture radius, this may result in visible chunking artifacts.

Smoothing Level

Sets the amount of smoothing for the capture. Higher values give a smoother capture surface which yields smoother results from the Point Deform SOP.

Capture Geometry Treatment

Tet Mesh Treatment

Only considers the surface of the capture tet mesh

Capturing is only done using the triangular faces of the tet mesh that lie on the surface, ignoring all points on the interior of the tet mesh

Treat the capture tet mesh as a solid volume

Capture in terms the entire tet mesh, treated as a volume.

Ignore all tets in the capture geometry

Completely ignore tetrahedra in the capture geometry for the purpose of capturing.

Tip

If your rest geometry is a higher resolution than your capture geometry, then it’s recommended to increase this parameter’s value.

Inputs

Input 1

Rest geometry. The model you want to generate capture weights. This is the higher resolution geometry to deform using Point Deform.

Input 2

Capture geometry. A copy of the animated lower resolution geometry with its points in their original starting positions.

Input 3

Animated geometry. The animated version of the capture geometry. This input is not modified and is passed directly to Point Deform.

Outputs

Output 1

Geometry capture weights.

See also

Geometry nodes