I think a physically plausible way would be challenging since Vray (to my knowledge) doesn’t deal with energy outside the visible spectrum (If I’m wrong, please let me know!). Since florescence takes non-visible light and makes it visible, it’s essentially the same thing as the ‘emit’ value (emit based on exposure to UV light)
Vray scans is super cool - but overkill for what I am looking for.
I wonder if I could use a surface luminance node or something similar to sample the light on the object and setup the self illumination (in a blend node with add?) to be a percentage to push the shader past physically plausible?
ie the amount of self illumination would be an extra percentage of the physical light that hits the object.
Is there a Vray node that anyone knows of to sample luminance in this way?
The other thing that makes ‘safety orange’ and other high-visibility materials so challenging is they often are coated with a special reflective surface similar to a bike reflector
“They reflect most of the incoming light back at the direction it came from, mostly independent of the incident angle of the light. That means, that the object appears brightest when the light source and the camera are coming from the same direction. For example, when in a car, you (the camera) and the headlights (light source) are pretty much in the same spot and thus the reflector lights up nicely.” source
Interesting! You could try two identical orange diffuse materials and add them together using a VRayBlendMtl set to additive mode. It should look fluorescent if the diffuse colours add up to >1, so for example RGB .6 .4 .2 + RGB .6 .4 .2 = RGB 1.2 0.8 0.4.
For a retroreflector try a somewhat glossy reflective material with the reflection amount set to / multiplied with the facing ratio of a samplerinfo. It will reflect more towards the camera.
Rens - could you elaborate on how your approach for the retroreflective material and samplerinfo would solve the issue of retroreflectance? I think I’m missing something.
The amount of reflection doesn’t matter as much as the direction of the reflection (surface normal). It seems like an appropriate approach would be to have a shader always point the surface normal toward the the camera. Unfortunately I don’t know enough about shaders to know if that is possible without writing a custom shader.
Yes you’re absolutely correct, the approach I mentioned above only multiplies the reflections of surfaces facing the camera, however in reality it should reflect objects near the camera back to the camera independent on the surface angle. So yes the normals should be bent towards the camera.
I’d like to dig into this with OSL at some point however you can try the following setup right now, seems to work pretty decently:
So this is obviously max not maya however it should work the same.
- Create a VRaySamplerInfo set to Refraction Vector, Camera, Unsigned.
- Create a VRayColor, RGB 0.5 0.5 1.0 (= default surface normal direction).
- Create a Blend Colors node. Attach the two previous maps. Use this to vary the strength.
- Creata a VrayNormalMap node (is that in maya?). Enable Flip Green for this to work. Plug in the Blend Colors node.
- Create a VRayMtl, have it reflect and give it some roughness/glossiness. Plug the VRayNormalMap into the bump.
Oh and by the way here’s the setup for fluorescence:
It works best with highly saturated, high value colours. Keep in mind that this should be two regularly shaded materials added together, not a self-illuminating material added on top, as even fluorescent materials won’t light up in shadowed areas (it only relfects a different wavelength of the same light).
