|
![](/i/fill.gif) |
scott wrote:
>> Normal trickery is very good for creating surface roughness. The
>> problems start when people try to use it to fake large structures such
>> as waves. (Yeah, now go look obliquely across the surface. Look wavey?
>> No, I thought not.)
>
> This is one of the best examples of parallax mapping I have seen:
>
> http://www.elderscrolls.com/images/art/ob_pc/obliv06B.jpg
>
> I imagine it looks even better when seen animated.
>
> Hard to believe that the walls are just flat polygons (well until you
> see the silhouette edge on the right under the chain!).
If you look at a "wavey" surface really obliquely, it will still not
look right. (Unless you use parallex mapping with occlusion, and nothing
intersects the polygon surface, and you can't see where the polygon
meets any other flat surfaces. I'm also not sure how refraction fits
into this picture...)
However, for "small" surface perturbations, it probably works quite well.
Post a reply to this message
|
![](/i/fill.gif) |