On 5/27/2010 6:49 AM, Dream wrote:
> clipka<ano### [at] anonymousorg>  wrote:
>> Am 27.05.2010 13:05, schrieb Dream:
>>
>>> Therefore, here I ask for your help for how to render this scene for a
>>> realistical result? How to set the papemeters for drops meterial?
>>
>> For highest realism, you'd want to use
>>
>>     finish {
>>       reflection { 0.0, 1.0 fresnel on }
>>       conserve_energy // not really necessary in this case
>>       ambient 0 diffuse 0
>>       specular ... // try it out
>>     }
>>
>> At least that's the theory.
>>
>> You need to make sure, however, that you do /not/ get reflections at the
>> interface between the drop and the leaf, so you need to make the objects
>> overlap a little bit. I guess that's your main problem ATM.
>
>
>
> Thank you all for your answers.
>
> According to what you all suggested above, I think the most possible problem is
> caused by interface between raindrop nether surface and leaf upper surface. The
> result image attached above is generated in the situation that the raindrop
> nether surface is lower than the leaf upper surface, that is to say, the leaf
> surface traverses the raindrop body.
>
> What I want to ask now is that when the two objects (raindrop and leaf) have the
> exact coincident interface (between raindrop nether surface and leaf upper
> surface), which surface should be upper? The raindrop nether surface is upper or
> lower than leaf upper surface?
>
> Considering the fact, the raindrop is on the leaf surface, therefore I believe
> the raindrop nether surface is upper than the leaf upper surface, am I
> right？
>
> Thank you!
>
Its random. Imho, this is, in some respects, a flaw in how the algorithm 
works. There is no "sorting" of the surfaces, to determine which one 
"should be" on top, or anything, you just have, "In this one case the 
math says surface A is on top, due to limits of precision, but the next 
pixel, it says its surface B." I am not even sure if you could do 
something like a bsp tree sort, to resolve those sorts of issues, but I 
wish someone would come up with something, because there are a lot of 
cases where an overlap, or a semi-exact surface to surface touch is 
closest to reality, but raytracing just won't allow it to happen. 
Anything in a glass, being a good example. lol

-- 
void main () {

     if version = "Vista" {
       call slow_by_half();
       call DRM_everything();
     }
     call functional_code();
   }
   else
     call crash_windows();
}

<A HREF='http://www.daz3d.com/index.php?refid=16130551'>Get 3D Models, 
3D Content, and 3D Software at DAZ3D!</A>

Post a reply to this message