PDA

View Full Version : Same light on different MR shaders


KungKrille
03-25-2006, 05:04 PM
I have modeled a perfume bottle where the glass is colored and the liquid is clear (I think). I have applied three different dielectric materials as the scheme below:
http://web.telia.com/~u13309256/perfume_materials_01.jpg

The background is a dgs material with white color and no glossy or specular. The bottle neck and the spray are supposed to be made of metal and has a dgs material. All materials (both dgs and dielectric) are linked to the same light that is a spot light. I use the spot as an area light with a physical light shader applied to it and here is my problem. In http://web.telia.com/~u13309256/perfume_render_04.jpg the light is set to 4000. In http://web.telia.com/~u13309256/perfume_render_05.jpg it is set to 30000. You see that the glass is very dark in the first render and the metal is overexposed in the second render. There is no caustic effect in any of the pictures. The trace depth is set to 20. Whatīs wrong with this scene?

KungKrille
03-27-2006, 07:23 AM
I guess my last post wasnīt clear so I try again. As you see on the renders I get very different exposures on different materials. The metal gets very bright while the glass is dark. They are linked to the same physical light. Should I use seperate lights for each material or is there some way to adjust how much light they catch? How would you set up the light for a scene like this?

Lorecanth
03-27-2006, 08:16 AM
use lobbers mental ray glass shader.

dagon1978
03-27-2006, 08:49 AM
what's the scale of your scene?
in my experience the default dielectric settings works well with very tiny scenes, try to set your V color attribute to 0,99 or 0,999, it means the % of your light absorption and it's strictly scale-dependant

KungKrille
03-27-2006, 11:53 AM
My scene is set to 1 unit=1 cm. The bottle is about 8 units wide, 7 units high and the distance to the light 37 units. I have been experimenting with the scale and found that too small scale will make it difficult to adjust the transparency of dielectric materials (on or off actually). And I still get huge brightness differences between the materials whatever scale I use. I must be doing something wrong here. I will take a closer look tonight after work again (have spent 1 1/2 week on this problem already).

dagon1978
03-27-2006, 12:06 PM
mmm maybe it's just a dgs setup problem, what's the d+g+s sum of your bottle neck shader?
are u sure the vcolor sum is <1? i can't see any problem in the background diffuse-only shader ;)

KungKrille
03-27-2006, 06:07 PM
Yes, mr dagon! Thatīs it! Thank you :) I didnīt think about the v-sum. Iīm new at this and have read a lot of tutorials and the v-sum was mentioned in some of them but I forgot it. The render looks like below. Can you answer me another thing - why is the bottle blue at the bottom? The light is white, the bottle greenblue and the metal as well, the liquid is clear and the backdrop is white. The reflection map has no blue color so where does it come from?

http://web.telia.com/~u13309256/perfume_render_06.jpg

dagon1978
03-27-2006, 06:31 PM
:scream:
are u using an hdri image or something like...?
it's an absorption problem i think... try to increase your raytrace settings or the FG settings, i dont know exactly what's your scene setup...

cgtriguy
03-27-2006, 08:39 PM
In looking at you diagram the bottom of the bottle is not filled with fluid but is rather behaving as solid glass. This is causing much more light absorption thus making the glass appear darker blue. The rest of the glass is lighter because it is thinner.

KungKrille
03-28-2006, 09:58 AM
Yes cgtriguy, the bottom is made of solid glass. If I compare the real perfume bottle with my cg I donīt get that blue effect in "real life". But you might be right anyway because probably the mr dielectric material works that way as you describe.

CGTalk Moderation
03-28-2006, 09:58 AM
This thread has been automatically closed as it remained inactive for 12 months. If you wish to continue the discussion, please create a new thread in the appropriate forum.