Vray Dispersion

[Guest]

I am attempting to replicate the video demo shown by Chaos Group where the new dispersion tool is used to generate a rainbow of colour from a prism.

http://www.youtube.com/watch?v=Zys-yKIt-ls

So far I have had little success, I cannot seem to get the dispersion to work and the caustic effects do not look anything like the YouTube video. I have attached the file, can some one tell me what I am doing wrong?

Thanks,

[Guest]

Oops didn't have Fresnel reflections on the glass.

[Guest]

Hmm,

My results seem to have far more beams and various things going on that what you see in the YouTube vid. I thought i would get one beam of light and then out the other end a rainbow of colour. I have posted some results below. I have taken notes from other threads on this forum.

Both atmosphere shadows and optimized atmostpheric evaluation on


Both atmosphere shadows and optimized atmostpheric evaluation off


atmosphere shadows on only


optimized atmostpheric evaluation on only

[Guest]

I have just tried rendering the scene with a background and the whole image becomes washed out by the fog. Its very desaturated. My settings for for are:

fog distance 200mm
fog height 200mm
subdivs 65
step size 100mm

The size of my prism is 50mm x 50mm x 50mm. Direct light has a hotspot beam of 0.5mm and a fall off of 2.5mm. If i turn the fog odd obviously the scene renders ok. Is there a way to control the fog so it doesn't cover the whole image?

[Guest]

Managed to fix the fog issue by adjusting the fog height. But I am having other issues with sizes because thinner smaller direct lights require higher intensities than thicker larger lights. In order for caustics to work effectively you need a much thinner smaller direct light resulting in unusual high multipliers that then affect the reflection in an over exposed way.

I could turn of reflection for that light but then its cheating. If I make the prism much larger say 1000mm x 1000mm x 1000mm then the intensity of the light is much more realistic. but the size of the prism is not.