Page 1 of 1

a simple cocktail - problem with ice

Posted: Wed Jan 07, 2009 9:48 am
by Onfame
i want to model a cocktail, its a wip so please be patient, but my problem is that the ice cubes (dunno if they are so called in english) always are kind of black, i tried several hdri´s and many gain factors. i used the ice material from the m-database, first i thought it´s the bump map, but then i tried to change the material from a glossy transparent with bump to a specular without bump, but no effect.

furthermore, since i use cindigo, my global scale factor is 0.001, all other materials work correctly.

so my problem is how to get those ice cubes look like ice?

Also i tried MLT with and without bidir, but no effect on the cubes.

i use indigo 1.1.15 and the newest cindigo version. Also i tried aperture diffraction and without, but still no significant changes in the result.

Posted: Wed Jan 07, 2009 10:09 am
by Juniorsatan
Check normals of the ice-cubes.

Perhaps the precedence althoug affect the effect. (lol)

Ist es für das Webdesigner-Battle? :)

Posted: Wed Jan 07, 2009 10:14 am
by Onfame
if i changed the normals the cubes get totally black, which means the normals should be okay, i also tried to change precedence from 1 to 10 to 20 to 50 but no significant effect.

And yes, this is my first sss liquid test for a battle on "" :)

Posted: Wed Jan 07, 2009 10:51 am
by Kram1032
precedence doesn't need to be a cerain value. It only should be like that:

Precendence of 1<Liquid<Glass<=Ice :)

0.001 ... did you model in millimetres?

Posted: Wed Jan 07, 2009 12:22 pm
by CTZn
Are you by any chance using the hybrid mode ? It's not to recommend...

Posted: Thu Jan 08, 2009 1:15 am
by Phr0stByte
I know you said you tried several HDRIs, but from what I can see in your posted pic, is a very black environment with the light source coming from the top. I would guess that the black in the ice is just your environment being reflected on the ice. I don't remember having any issues with the ice material when I first made it, though I have not used it in recent versions of Indigo.
Try enclosing the scene in a cube and light it with a couple of area lights to test this theory.

Posted: Thu Jan 08, 2009 2:31 am
by PureSpider
Thats what I think, too!
The Ice looks totally fine to me, judged by its refractions and such!

Posted: Thu Jan 08, 2009 3:36 am
by neepneep
Also, make sure the ice cubes are not instances - instanced objects with transparent materials turn black for some reason.

Posted: Thu Jan 08, 2009 3:48 am
by Onfame
hmm i tried krams tip, now it works just fine, i think the problem was a too high precedence of liquid and glass :) Also i changed the global scale factor from 0.001 to 0.01. Here´s the result:

Posted: Thu Jan 08, 2009 5:51 am
by crojack
it looks nice, but the ice is too far out of the water. at least it looks like that to me.

Posted: Thu Jan 08, 2009 5:55 am
by Phr0stByte
What crojack said. I dont think it should even be floating at all. I would make the ice much bigger, deform them so that they don't all look the same, and arrange them pile up from the bottom of the glass with the top ones sticking out of the liquid a bit.

Posted: Thu Jan 08, 2009 6:45 am
by Onfame
as i said it´s just a WIP to test the material settings :)

Posted: Thu Jan 08, 2009 7:45 am
by Kram1032
ok, so only the liquid's precedence was too high: as the glass doesn't touch the ice cubes, it doesn't affect them.
But if the liquids precedence was even higher than the one of the glass, it affected the liquid.

Look at precedence like this:
It's a layering system.
If you map it to 2D, it would be like that:
The uppermost layer would partly hide layers underneath. Those layers underneath could hide other layers lying under them.
And so on.

Now, the higher your layer lies, the higher is the number it has.
"1" is already presaved for standard air.

So, if you make it like this:

liquid = 2
glass = 3
ice = 3

that'll work just as fine as if you say

liquid = 10
glass = 25
ice = 99

It's recommended for more complicated scenes that you don't use double precedence value. Also, as that's allready the surrounding air, never use precedence of 1!

Things start to get odd, if you have two similar precedences and the two materials intersect:

liquid = 2
glass = 2
ice = 1

Indigo couldn't decide propperly, wether to use liquid first or rather glass. What you get is a nebulous buggy something.
Even worse, ice has a precedence of 1, which lets it "mix" in the same undefinable way but with air.

Next problem in that case would be, that the liquid would also cut off the ice, instead of the other way round!

Posted: Thu Jan 08, 2009 9:17 am
by Onfame
THANKS for that wonderfull explanation, now i got it :D