Position variation: The amount of random variation in cell positions. 0=all cells are in a regular 3D grid, 1=two adjacent cells may overlap completely.

1D: The cell noise is one dimensional. 1D noise is faster to compute than 3D cell noise.

4D Input: Selects four dimensional input for the noise. When the option is set, the cell object gets a second input channel, whose first sub channel is used for the additional 4th dimension. Typically the 4th dimension is animation time. This is useful for creating animated turbulent fog, fire and other similar phenomena.

Additive: If set, the noise is sum of adjacent cells. Otherwise, the maximum value of adjacent cells defines the noise.

Centered: If set, Amplitude varies symmetrically around the Base value. If cleared, Amplitude is added to the Base value.

Cell shape: Sphere - spherical cells, Cube - box like cells, Rounded cube - Cell shape is a box with rounded edges.

Asymmetry: Defines how symmetric the cells are. If cell shape is sphere and asymmetry is zero, all cells are spheres. Non-zero asymmetry produces elliptic cells. The value 1 makes the proportions of cells fully randomized. The value of this attribute is a vector, and therefore variation of each cell dimension width/height/depth can be defined independently.

Density: Scales the input parameter. The higher the density, the smaller (=denser) patterns the cell noise creates.