For some reason, this is way better than the last one. VERY realistic...

Norbert Kern wrote:

> hi all,
> 
> I got many encouraging statements regarding my last posting, thanks a lot.
> Furthermore some of you mentioned too uniform textures, too bright or too
> dark areas and so on.
> Those critics were all right.
> Radiosity would be a solution, but is too slow (probably too many objects).
> All shadowed points with the same texture have the same brightness with

> look like real nature.
> The main problem is the sky, which illuminates shadowed parts in nature, but
> not in povray (same is true for other raytracers, I think).
> There is no such light definition in megapov or povray. But many lights
> distributed on the sky should do the same.
> The best uniform distribution of points on a sphere is like the 60 points on
> buckminsterfullerene (like a football (soccer in U.S.)).
> I took those coordinates, deleted the points below the ground (negative y
> values.) Then I reduced the 30 lights to the most relevant 11, because with
> 30 lights test renders were too slow.
> Render time is 6 times as high as with fill lights. but the result is much
> better.
> To speed the test render up, some plants were excluded.
> 
> Comments?
> 
> 
> Norbert
> 
> 
> 
> 
> IRTC-WIP41.jpg
> 
> Content-Type:
> 
> image/jpeg
> Content-Encoding:
> 
> x-uuencode
> 
>

Post a reply to this message