Can anyone explain that when a popobject is used as a source for volume fluids, why the min/max substeps on the fluid solver don't affect that part of the simulation, and the pop source is only evaluated at the dop network substep?

I know I've seen an answer to this but I cant find it now.