Hiding vraylight from reflection

[percydaman]

can you show an image to illustrate your problem? What kind of AA filter are you using?

[dlparisi]

I did a simple test like you said. Its pretty obvious that the glossy/specular render is alot brighter then just glossy or just specular.

Ill tell you what though. It doesn't speak to vray's physical accurateness if someone is forced to use both the specular highlight fake, along with the 'supposed to be' correct glossiness highlight, in order to get the correct value.

I think I've figured out what's happening here (apologies to Vlado ). It appears that checking or unchecking either "Affect Specular" and "Affect Reflection" causes vray to deviate from a physically correct world and is instead forced into separating out the specular and reflection, which as dimtri said, isn't true to nature. For example, once you uncheck "affect specular" in the light, you can get the exact same image if you leave it on and take the reflective material and force a highlight of 1 (i.e., no highlight). Similarly, if you uncheck "Affect Reflection" in the light, you can get an identical render if you were to leave it on and uncheck "Trace reflections" in the material. I think the confusion comes from the nomenclature in the light controls since you can't really separate the specular and reflective components of the light itself as these are material properties.

I think we all already know that separating the highlight glossiness from the reflection glossiness in the vrayMtl is a no-no for physical accuracy (but necessary sometimes) so I think this explains why percy's test resulted in a "inaccurate" result.

[vlado]

V-Ray only guarantees an accurate result when you calculate both speculars and reflections; turning either of these off, in the lights, or the materials, is not guaranteed to produce any particularly logical result - it will remove the specified component, but the values of the remaining one can vary depending on the light and the material.

This is because, in the nature of the light calculations, V-Ray computes something that is "like" hilights, and something that is "like" reflections, which when added give you a correct result. Each of these is only a partial result though, and so is weaker than the total result.

V-Ray does this, because some cases are better computed by reflections (e.g. mirror-like materials with large light sources), and some cases are better computed with hilights (e.g. glossy materials with small light sources) - so it makes sense to compute both and combine them in a way to produce the best (least noisy) result.

Best regards,
Vlado

[percydaman]

hmmmm.... I always thought that the highlight glossiness was there for the times you choose to use a vray material and a max point light. IMO the glossiness should be able to produce the correct reflection using ONLY the glossiness when using a geometry based light. And the highlight should be the fake those that use the point light need. Its seems unnecessarily convoluted otherwise.

[vlado]

For objects with self-illuminated materials, reflections will give you the correct result (since V-Ray does not know how to directly trace them for hilights).

For point lights, there is only a hilight component and it is the correct result (obviously, there is no reflection component, as the lights are immaterial).

For VRayLight's, there is both a reflection component, and a hilight comonent, which are added together. The ratio of these components depends on the light size and position relative to the material, as well as on the material glossiness. This is not just for the sake of confusing you - it greatly reduces the noise in the speculars.

It may be interesting to observe also that diffuse illumination from VRayLight's is also split between direct lighting, and the GI calculations, for the same reason - noise reduction (only when you use GI, of course).

All in all, by default all this works correctly, so you don't really have to think about it.

Best regards,
Vlado