SDF Grid Mean Node

The SDF Grid Mean node applies mean (box) filter smoothing to amSigned Distance Field (SDF) grid. This operation averages voxel values within a local neighborhood to smooth out small variations and noise while preserving the overall shape of the field.

The filter works as a fast, separable averaging operation—sometimes called a box filter—and is well suited for general-purpose smoothing or softening of SDF data after boolean, voxelization, or other volumetric operations.

Inputs

Grid

The input signed distance field to be smoothed.

Width

The radius of the smoothing kernel in voxels. Larger values result in a broader smoothing effect, producing softer, less detailed surfaces. Smaller values restrict smoothing to immediate neighbors, retaining more fine detail.

Iterations

The number of times the mean filter is applied. Each iteration increases the smoothness of the surface. Multiple passes can approximate a Gaussian-like blur effect.

Outputs

Grid

The resulting SDF grid after mean filtering. The surface becomes smoother and more uniform, making it useful for refining noisy or rough SDFs before further operations such as Grid to Mesh.