Cornell Box

[Meelis]

Hi

When someone would make holes in water surface using boolean then also (fix me here if i'm wrong)
but we need to extrude from hole edges to bottom and make water mesh solid, or not?

When we have very complex shape there where now the cubes are, it would be extreamly complesx to achieve.

[neo2068]

Hi

When someone would make holes in water surface using boolean then also (fix me here if i'm wrong)
but we need to extrude from hole edges to bottom and make water mesh solid, or not?

That is not nessessary. You only have to model the interfaces, i.e. water/air, water/glass, glass/air. The interfaces water/glass and glass/air are modeled with the box geometry.

When we have very complex shape there where now the cubes are, it would be extreamly complesx to achieve.

Because of that I first applied the boolean modifier to a simple plane and after that I subdivided the mesh and applied the displace modifier.

[Meelis]

That is not nessessary. You only have to model the interfaces, i.e. water/air, water/glass, glass/air. The interfaces water/glass and glass/air are modeled with the box geometry.

But how does red light ray know it's in water or has been in meanwhile?

[neo2068]

That is not nessessary. You only have to model the interfaces, i.e. water/air, water/glass, glass/air. The interfaces water/glass and glass/air are modeled with the box geometry.

But how does red light ray know it's in water or has been in meanwhile?

That's right. I only addressed the issue of the dotted red water surface inside the green object. That has to be removed. The other problem is to correctly assign the interfaces. You have to split the green object, i.e. the glass box, into 2 materials. One for the glass/air interface and another for the glass/water interface.

[Meelis]

That's right. I only addressed the issue of the dotted red water surface inside the green object. That has to be removed. The other problem is to correctly assign the interfaces. You have to split the green object, i.e. the glass box, into 2 materials. One for the glass/air interface and another for the glass/water interface.

Seems logicall.

That took me to think about situation with instanced volumes (water drops for example on surface), but they are on foggy environment, that would also render incorrect,
unless i set water drops exterior to fog (world 2). I have to look some of my older stuff, maybe something renders better if i set stuff up correct way

[slime]

I thought that was what the glass2 material was meant for, volumes and automatically working out the different between the external and internal IOR. On the glass balls that are fully submerged there is a glass - water interface, so the cubes need to be split along where the water reaches and given two materials.

Neo can the boolean thing do this? Split the cube to where it's above water and below water

[Meelis]

Neo can the boolean thing do this? Split the cube to where it's above water and below water

Yes

Correct is that way:

[neo2068]

Your sketch implies that there are two surfaces on the water and glass border. Actually, there only has to be only one surface, the glass/water interface. There has been other threads that discussed this matter:
viewtopic.php?f=36&t=3767&p=36463&hilit=water+glass+interface#p36420
or
viewtopic.php?f=16&t=1645&p=26225&hilit=water+glass+interface#p14844

[Meelis]

Your sketch implies that there are two surfaces on the water and glass border. Actually, there only has to be only one surface, the glass/water interface. There has been other threads that discussed this matter:
viewtopic.php?f=36&t=3767&p=36463&hilit=water+glass+interface#p36420
or
viewtopic.php?f=16&t=1645&p=26225&hilit=water+glass+interface#p14844

Oo yes this method looks quite good.

[hoodedmanwithsythe]

Neo can the boolean thing do this? Split the cube to where it's above water and below water

Yes

Correct is that way:

This would be true for a 100% hydrophobic material but nothing else.