Equiangular Weight

Sets the probability of equiangular sampling being used instead of density sampling. Equiangular sampling improves the convergence of volumes close to light sources, while density sampling can be a better technique when dealing with volumes with dense or highly varying density. The default value of 0.5 means both techniques will be used equally and combined with multiple importance sampling. If the volumes are dense and far away from light sources then decreasing the equiangular weight may result in better convergence. This does not apply to multi-scatter volumes.

min/max Samples

This setting controls sample quality for single scatter volumes. Higher number of volume samples will improve convergence of volumes at the cost of render time. However, the trade-off in scenes with other objects is that the volume will sample more without increasing costs elsewhere in the image and may improve your results. Note convergence rates vary within volumes as well, and this means some regions can converge reasonably well with a small number of volume samples, but others might need more help.

• minSamples enforces a lower limit to sample, only valid with Single Scattering
• maxSamples are the maximum limit per ray inside the Volume

Step Size

Only valid when using the PxrUPBP integrator to render the scene, sets the stepping size for raymarching in the volume. Small values will increase quality at the cost of computation, very small volumes may add enormous overhead, adjust slowly. Large values may render more quickly but sacrifice accuracy and render incorrectly.