Page 1 of 2
Glass/liquid final answer...?
Posted: Wed Jun 22, 2005 8:02 pm
by mtripoli
I know it's a pain in the rear, but can someone please tell me what the *final* answer was regarding a liquid in a glass. Do we overlap the liquid/glass poly's? "Cap" the glass to get the liquid to appear correctly (out to the "edges"?)...
Thank you!
Mike Tripoli
Posted: Wed Jun 22, 2005 11:37 pm
by tom
ROFL NOW!
Mike, tell me when you find it how....
There are several ways like ThomasAn Method, Rivoli Method, Mihai Method etc....
But we finally decided to have a ray tolerance option.
Just a wish for now...
Posted: Wed Jun 22, 2005 11:43 pm
by Maximus3D
Ray tolerance hm, that would be like what mental ray uses, it ignores normals and only uses ray tolerance to calculate the light distribution with two surfaces which are that close to eachother as a liquid in a glass is to the glass itself.
Good choice.
/ Max
Posted: Wed Jun 22, 2005 11:44 pm
by tom
But i dunno if the developers think the same or not yet...
Probably oscar would come up with a better solution when
he first make this first things first...
Posted: Thu Jun 23, 2005 12:40 am
by mtripoli
OK... I got so lost in the thread about this... (I'm not very bright!)... Hopefully it'll never come up that I need this... until I do!
MT
Posted: Thu Jun 23, 2005 5:52 am
by Thomas An.
Well, we tried all kinds of methods, but for now (and based on how Maxwell experimentally works) this seems to give accurate results ... which were also verified by a physical experiement.
http://www.maxwellrender.com/forum/viewtopic.php?t=2906
Posted: Thu Jun 23, 2005 9:06 am
by hdesbois
Ray tolerance is good, but sometimes, it's easier to deal with coplanar polygons with normals pointing opposite ways. I hope Maxwell will handle this situation (now it's just a mess : you loose the color of the liquid inside the glass, and the reflexions just go crazy).
HD
Posted: Thu Jun 23, 2005 9:13 am
by Kabe
hdesbois wrote:now it's just a mess : you loose the color of the liquid inside the glass, and the reflexions just go crazy
What do you mean by that? Do you have examples to show?
Thanks
Kabe
Posted: Thu Jun 23, 2005 9:37 am
by hdesbois
like this :
The liquid inside is supposed to be some wine.
This is an alpha render, but beta still has this kind of problem (maybe reflexions are not that bad, but color is lost anyway). I will cook another render with beta.
HD
Posted: Thu Jun 23, 2005 11:27 am
by Hervé
L'eau de la vie...Water of life... look Hdesbois is a "Fremen"... you know you only find this water on planet Dune... or also on Geidï Prime if I recall...
Posted: Thu Jun 23, 2005 11:29 am
by Mihai
hdesbois, if you use the method described by Thomas in that thread you shouldn't have any problems. It was tested several times and it works very well for stills.
Posted: Thu Jun 23, 2005 1:28 pm
by mtripoli
Sorry to sound so dense, but does the "green line" in your drawing represent the continued glass surface? In other words, is the liquid "sealed off" by capping the glass? If I saw a wireframe model it would explain all... thanks!
Mike Tripoli
Posted: Thu Jun 23, 2005 1:34 pm
by Mihai
mtripoli wrote:
Sorry to sound so dense, but does the "green line" in your drawing represent the continued glass surface? In other words, is the liquid "sealed off" by capping the glass?
That's exactly right.
Posted: Thu Jun 23, 2005 1:57 pm
by mtripoli
thank you!
MT
Posted: Thu Jun 23, 2005 5:10 pm
by hdesbois
Ok, I know it works (apart from current caustic issues) :
but this is a bit of a fake, isn't it?
HD