Glass test

sprocket · Post by **sprocket** » Sat Jan 13, 2007 9:43 am

Hi Ono,
I've made a lava lamp, so I have wax inside water inside a glass outer casing. What do I need to set the precedence values to? (what do they actually mean as well?!)
I've got 3 specular mats; glass with uniform absorbtion (clear), oil inside it with blue absorbtion, and an SSS mat for the wax.

I assume it has to be modelled like the drink glass on the first page also?

Thanks!

CTZn · Post by **CTZn** » Sat Jan 13, 2007 3:06 pm

What do I need to set the precedence values to? (what do they actually mean as well?!)

Bump !

is that like an override for geometry normals (higher precedence for material with higher ior? Don't make much sense tho)

I assume it has to be modelled like the drink glass on the first page also?

I think so, think as of "interfaces":

1° air-glass
2° glass-water
3° water-wax

and in the special case wher wax is touching glass:

4° glass-wax (thus one surface for glass, one for wax and a third one for the interface of them both)

ryjo · Post by **ryjo** » Sat Jan 13, 2007 10:44 pm

CTZn wrote:
What do I need to set the precedence values to? (what do they actually mean as well?!)
Bump ! is that like an override for geometry normals (higher precedence for material with higher ior? Don't make much sense tho)

If several media occupy the same space, the renderer will consider the medium with the highest precedence. I don't think it is related to ior.

CTZn · Post by **CTZn** » Sun Jan 14, 2007 4:50 am

Oh ok, a simple priority flag then, thx ryjo... could that be confirmed please ?

Edit: that was confirmed !

sprocket · Post by **sprocket** » Mon Jan 15, 2007 7:14 am

ryjo wrote:
If several media occupy the same space, the renderer will consider the medium with the highest precedence. I don't think it is related to ior.

Thanks!

The new documentation is so handy thanks Ono for the update!

CTZn wrote: 1° air-glass
2° glass-water
3° water-wax

and in the special case wher wax is touching glass:

4° glass-wax (thus one surface for glass, one for wax and a third one for the interface of them both)

Damn, looks like I'll have to do some hefty remodelling to get those interfaces correct!

Post by **OnoSendai** » Tue Jan 16, 2007 3:11 am

To build the lava lamp, u need to define 3 media: oil, water, and glass.
Then define 3 specular materials, using the 3 media.

Glass should be modeled as you would expect.
The water mesh should overlap the glass mesh, as diagrammed in the first post.
The lava (oil) mesh can be modelled to sit inside the water; as long as it has a higher precedence than the water, it will 'displace' the water.

precedences should be

glass > oil > water.

sprocket · Post by **sprocket** » Tue Jan 16, 2007 7:07 am

Thanks for clarifying Ono!

One more question (sorry

) Is it possible to get an sss material with an IOR greater than 1, but no reflections on the surface?

Post by **OnoSendai** » Tue Jan 16, 2007 7:22 am

Yes it is possible, but it's not really physically realistic.
You can use <null_material/>

What effect are u trying to achieve?

sprocket · Post by **sprocket** » Tue Jan 16, 2007 7:49 am

Thx for the speedy reply!
Still the wax in the lava lamp I'm afraid..
The reflections are ok in that they're physically correct and I do get them on my lamp at home, but I needed a way to get them more subtle.
Also would be cool for diffuse stuff like if you want sss transmission through a sheet of paper, though i suppose the IOR doesnt really matter in that situation as the paper is so thin.

How would you use <null_material/> in this way?
It just seems to make objects disappear when I try...

Post by **OnoSendai** » Tue Jan 16, 2007 8:03 am

null_material is basically a bsdf that does nothing; light passes right thru it, but u will see the object if it has SSS.

You might want to try glossy_transparent for the lava.

Yeah... i need to think of some ways to handle nice materials for single polygon models with no thickness

sprocket · Post by **sprocket** » Tue Jan 16, 2007 8:32 am

Glossy transparent was awesome, I tried it a while back and its the closest to real life ive got so far

Can't get the null_material do do anything with sss still... i've tried it in all the possible places I could think of, it either makes the object disappear completely or Indigo throws up an error

CTZn · Post by **CTZn** » Tue Jan 16, 2007 4:54 pm

Yes, correct rendering of glass/liquids is possible now.
Make the model to the specifications in the first post.

oops, didn't had a look at the first post, sorry, I was assuming an interface had to be one surface with normals indicating where each medium was... now I got it, makes clear what precedence is for

Glass test

Who is online