sdf.measure.volume.volume

sdf.measure.volume.volume(
    sdf,
    bounds=(-3, -3, -3),
    size=(6, 6, 6),
    resolution=50,
    epsilon=0.01,
)

Estimate the volume of an SDF as a differentiable JAX scalar.

Uses a smooth sigmoid indicator σ(−d/ε) as a soft approximation of the interior indicator (1 where d < 0, 0 where d > 0). The result is differentiable with respect to any free parameters of the SDF.

Parameters

Name	Type	Description	Default
sdf	SDF	The SDF to measure.	required
bounds	tuple[float, float, float]	Lower corner (x, y, z) of the sampling box.	(-3, -3, -3)
size	tuple[float, float, float]	Extent (dx, dy, dz) of the sampling box.	(6, 6, 6)
resolution	int	Number of samples per axis (resolution³ total).	50
epsilon	float	Smoothing width. Smaller → sharper / more accurate; larger → smoother gradients.	0.01

Returns

Name	Type	Description
	Array	Differentiable scalar volume estimate.