Well, everyone kept calling it a rubber duck since Rosetta
sent the first images of the comet shape, so I decided to go
with that theme. BTW, the density of the comet is only about half
that of water so may it could float - if you could wrap it into
a 6 km long yellow insulation bag and prevent it from further
compacting under gravity :)

I downloaded the model from ESA yesterday and spent today playing
with it. Well actually I mostly spent time looking at rubber duck
images and trying to figure out SSLT (this was the first time I
attempted to recreate a specific material).

The duck bill was "borrowed" (or surgically removed) from Gilles
Tran's rubber ducky at

   http://www.oyonale.com/modeles.php?lang=en&page=53

I wasn't really certain whether to go with eyes and bill since
it all seemed to be sliding down uncanny valley, but I found it
more or less tolerable with a subdued color for the bill.

I'm not really happy with the look of the material, it does not
seem very rubber-like to me. Although it did look better on a test
model of an actual rubber duck, so maybe part of it is that the
roughness and edges of the surface defy the expectation.

However, I think mostly my settings are simply incompetent.
I had a hard time getting rid of greenish tints in the darker
areas despite having no dominant green in the scene:

texture
{
   pigment {color rgb <1,0.7,0.01>}
   normal  {granite 0.05 scale 0.1}
   finish
   {
     specular 0.2 roughness 0.08 metallic 0.5
     subsurface
     {
       translucency rgb 20*<2,0.2,0.02>
     }
   }
}

interior
{
   ior 1.52
}

Looking at samples of rubber duck images it seemed they were
distinctly more reddish in their dark areas, so I tried to use
a yellowish pigment with a strong red translucency effect.

I added partial metallic highlights since they tended to
look more yellowish than whitish in sample images. The normals
were added as faint background structure because the model was
a bit plain in some areas, but they got mostly lost when
rendering with better aa and focal blur.

I found ior 1.52 listed as ior for natural rubber, although I
belatedly realized that most "rubber" ducks are probably no longer
made from rubber. However, Polyvinylchloride that seems to be used
a lot has ior of 1.54 so not much harm done.

Apart from SSLT this uses 2 bounce radiosity. It rendered
at 640x480 with +a0.1 in 2h 9m on an i7-970 with 12 threads).

While not a great render technically I still found it fun
and I hope you have a laugh too :)

Post a reply to this message

Attachments:
Download '67p.png' (219 KB)

Preview of image '67p.png'