triple_r nous apporta ses lumieres en ce 23-04-2007 10:05:
> It's explaned in the following link somewhere in the middle, but it
> basically amounts to the vorticity minus the shear, leaving the rotation.
> Not my paper of course, but I just tried (unsuccessfully) to replicate the
> results.  I couldn't get mine even close to turbulent.

> http://www.bgu.ac.il/~yakhot/homepage/publications/JFM_2006.pdf

> Say, does anyone know how to get decent render times out of good-looking
> media?  Method 3 tends to give me 'slices' and Method 2 needs too many
> samples.  Here are my current settings for a smokey-looking wake, but it's
> taking quite a while.  Maybe that's to be expected.

> media {
>   method 3
>   intervals 30
>   ratio 0.9
>   samples 4,4
>   jitter 1.0
>   aa_level 3
>   aa_threshold 0.05
>   emission <1,1,1> * 8
>   absorption <1,1,1> * 2
>   scattering {
>     1, <1,1,1>*10
>     extinction 2.5
>   }
>   confidence 0.99
>   variance 1/100
>   density {
>     density_file df3 "output.df3"
>     interpolate 1
>     color_map {
>       [0.0 rgb 0]
>       [1.0 rgb 1]
>     }
>   }
> }

>  - Ricky




When using method 3, intervals >1 is a NO-NO.
Leave it at 1 and increace samples, MUCH faster!

-- 
Alain
-------------------------------------------------
My toes were playing the piano under the blankets.
They made so much noise that I woke up singing.

Post a reply to this message