PDA

View Full Version : Problem with "Unseen by Rays"



dpartridge
07-05-2004, 05:57 AM
I am always finding a short fall with "unseen by rays", When I am seting up a scene with two image. one hdri image for illuminating the scene and a tif image for the backdrop.

I have set up an outside scene which as two images. Both images are full 360 views. This scene is a animation where the camera rotates around the scene. I have placed the hdri image in the "backdrop:texture environment" and turn on radiosity by "backdrop only", the scene works fine and is illuminated by the hdri image. The problem is the hdri image is incorrect for the backdrop of the render image, and I need to use the tif file as the backdrop and reflection. I have placed a large sphere in the scene and mapped the tiff file to it for the backdrop of the render. The sphere blocks the hdri texture environment from illuminating the scene. So I have checked "unseen by ray" for the sphere. This stops the sphere from blocking hdri texture illumination, but now the backdrop tif ( sphere ) is not seen in the reflections. so in other words, i need two types of "unseen by ray". one for radiosity and one for reflections.

does any one know of a way around this?

Panikos
07-05-2004, 06:20 AM
Yes, this is a well-known limit and several users have requested an "Unseen by Radiosity" switch.

Personally I prefer to have two giant-spheres with different maps and different roles so I can exlude the one from the camera and the other from the rays. It allows you some more control since its impossible to exclude the backdrop from any calculation.

Hopefully we will see some "Unseen by Radiosity" in the future.

:(

Exper
07-06-2004, 10:31 AM
Yep...
LW needs some less "tricky" approach when speaking about the renderer! ;)

toonafish
07-06-2004, 11:26 AM
Couldn't you use a spherical reflection map instead of a dome ?

Panikos
07-06-2004, 02:21 PM
Spherical Reflection map is additive to the surface properties and this forces the object to emmit a lot of irradiance, which is wrong.

The nature of the problem here is different, and adding spherical reflection map will make the problem more complex than giving a solution.

toby
07-06-2004, 09:49 PM
This looks like a job for Multiple passes!