PDA

View Full Version : Shadow doesn't "respect" surface imperfections



Johnny
03-20-2005, 11:40 AM
I have a light pointed generally toward the camera from behind that aqua glass wall..with RT shadows turned on, that light casts a perfectly smooth shadow onto the purple surface, while that same surface shows a squiggly reflection of the shadow-casting object.

Can anyone give me a hint as to why the shadow isn't also distorted somewhat by the surface distortion, which is a procedural bump?

thanks!

Johnny

http://www.bluemitten.com/shadow.jpg

Doran
03-20-2005, 02:56 PM
You're using a procedural to simulate surface bump.. it's not real bump and the shadow is following the true geometry. Try a deformation of the surface geometry.

Captain Obvious
03-20-2005, 06:41 PM
You're using a procedural to simulate surface bump.. it's not real bump and the shadow is following the true geometry. Try a deformation of the surface geometry.
What he said.

Bump maps only effects the surface normals. Shadows don't really care about which way a polygon is heading, only the actual position.

Johnny
03-20-2005, 07:00 PM
Ahh...ok...that makes sense...

thank you!

J

Red_Oddity
03-21-2005, 02:49 AM
Hmmm...if shadows only keep surface normals in account, would it be save to asume that a normal mapped floor would affect shadows?

Captain Obvious
03-21-2005, 03:20 AM
Hmmm...if shadows only keep surface normals in account, would it be save to asume that a normal mapped floor would affect shadows?
No no, it's the other way around. Shadows don't care about the normals, only the position.

Johnny
03-21-2005, 08:20 AM
You're using a procedural to simulate surface bump.. it's not real bump and the shadow is following the true geometry. Try a deformation of the surface geometry.


Then again, how is it that the reflection of the hypo is distorted by that very same procedural? The bump is not in the reflection channel of the surface, but in its bump channel.

That surface's geometry is no more changed for the reflection than it is for the shadow, since it's a procedural surface element.

Thanks,

J

tischbein3
03-21-2005, 09:59 AM
Then again, how is it that the reflection of the hypo is distorted by that very same procedural? The bump is not in the reflection channel of the surface, but in its bump channel.

That surface's geometry is no more changed for the reflection than it is for the shadow, since it's a procedural surface element.

Thanks,

J

Because Reflections depends on the surface normal.