[ishook]

Has anybody experimented with ice in liquid? I've seen the beer stuff here, but that was liquid around ice, not penetrating. I haven't had much visual luck with rendering ice cubes intersecting a liquid. I know we're all patiently waiting for a double sided material (an in ND and an out ND) but until then, we need a half-decent workaround.

anybody?

[JTB]

I know we're all patiently waiting for a double sided material (an in ND and an out ND)

This is the first time I hear this

[tom]

Yes, we have some examples here:
http://www.maxwellrender.com/forum/view ... p?p=225526

[Fernando Tella]

Maybe a way to specify whether a surface is adherent or not when is very close to another surface so there's no air gap in between would be enough. It could be an object property with a max distance to become adherent. This way you would model all liquids and objects inside as they really are and Maxwell would calculate the correct Nd for all the transitions.

Just thinking aloud.

[MS]

Maybe a way to specify whether a surface is adherent or not when is very close to another surface so there's no air gap in between would be enough. It could be an object property with a max distance to become adherent. This way you would model all liquids and objects inside as they really are and Maxwell would calculate the correct Nd for all the transitions.

It could help a lot!

[Fernando Tella]

You also could use realflow meshes exactly as they are generated...

[ishook]

Maybe a way to specify whether a surface is adherent or not when is very close to another surface so there's no air gap in between would be enough. It could be an object property with a max distance to become adherent. This way you would model all liquids and objects inside as they really are and Maxwell would calculate the correct Nd for all the transitions.

Just thinking aloud.

The problem is that light rays leaving the ice cube don't know that its entering into a liquid again, since there is no new surface. Your idea is good, and technically would solve this problem, but its difficult to set up like that without a liquid simulation, as booleaning often fails.

[Fernando Tella]

Starting from ice cube's mesh it wouldn't be so hard to make the corresponding water mesh by cutting, detaching, inverting mesh, attach to water mesh and weld vertexes.

[simmsimaging]

I am not sure exactly what you are after here because ice "penetrating" liquid, and not having liquid "around" ice doesn't actually make sense to me. If what you want is to have ice cubes floating in a liquid medium then you are actually after the latter, and fluid sim is probably the best way to go.

I do mine almost exclusively with Realflow now, but I have done a few with Glu3D as well. I'm sure Boolean operations can work, but for us it was pretty buggy. It gets very slow and apt to crash Max with complex geometries.

b

[ishook]

When ice in a glass isn't the focus of the render, it can be a waste to do a realflow simulation. That could be more than half a day just trying to get ice in a glass. Booleaning is a waste, so the logical thing is for maxwell to develop a material that knows when its encased in another medium. I'm sure NL wants us to do a realflow simulation for every liquid+solid object setup, but its not gonna happen.

[simmsimaging]

whatever works man.
b