Maxwell Render | Official public forum

So I haven't used MW in quite a while and had forgotten about the way you need to model intersections for transparent materials. I'm looking at Thomas An's method, but wondering how to deal with a more complex situation.

If I have a glass with liquid, and the surface is slightly rippled, and there are ice cubes in it - how would one approach that except as a very complicated modeling problem? What about doing a Realflow pour into a glass with ice? Is the only real option to boolean out the shapes or something like that?

b

With RF, you just have to live with the result of intersections. For animations, it's not such a big problem, but for stills, there would be a lot of modelling involved to get it "correct". For example, for the ice cubes floating in water, you would have to make a very small "shell" of the ice cube geometry, trim out this shell to the water's surface where the ice cube is completely enclosed in the water with no intersections. Talk about tedious!

Tedious, yes. I might even say pointlessly so

This poses some issues for me. If I was to use Maxwell for only one thing it would be precisely for this kind of stuff, as it does such an amazing job with this kind of subject. Might be too much hassle though.....

Thanks Brian.
b

Hey Brett, do you use Max? The ProBoolean tools are very good.

I do use Max. I will have a look at that option, but the issue with that kind of workflow is that it lacks flexibility to deal with client changes. What happens when they decide to move the ice cubes over, or make them smaller and add one? The work just gets magnified, and complex (which means too much room for error with me). Those kinds of changes are a fact of life for me, right up to and even after "final" rendering. If the workflow is prohibitive I may just not be able to go that route.

Thanks for the tip though - will check it out.
b

Also check out RF5's mesh clipping options. I haven't tested it yet, but it may hold your solution.

I don't have RF5, but thanks.