When modeling the volume of a liquid for caustics, is it still necessary to make the volume's geometry go beyond its container walls?
For the surface of the liquid, will the solver recognize either a bump map or displacement map for the surface irregularities?
First question: we always do this, and it looks correct, so yes, overshoot. But it's not necessarily essential.
Second question: either will work - displacement is of course more realistic, and will look correct at the edges where the water meets other surfaces, but bump is a good quick way of getting more or less the same result without worrying about the displ map screwing things up on the edges e.g. wave height becoming a challenge. Just not as realistic to look at. Depends on use case really.
The above are assuming you're using phoenixoceantex (which is great), but you don't have to.
You have a highly appropriate username for these questions!
