Vray Fresnel - Physically correct?

#1

Hello everyone,

My question is regarding the fresnel option in Vray. I am all about being physically accurate, linear workflow and getting it as close to real life before breaking rules.

In regards to Fresnel, if I enable the option, is this how Fresnel should be used to be accurate. Just clicked on and that is it?
The option that says “Lock Fresnel IOR to Refraction IOR”. If I unclick that and play with the Fresnel IOR, is that unrealistic? I thought in real life, things would have different fresnel values technically.

If this is the case, can someone also provide more info, is parquet flooring like 1.3 for Fresnel IOR for example?

Thank you

#2

http://www.robinwood.com/Catalog/Technical/Gen3DTuts/Gen3DPages/RefractionIndexList.html

Vray’s default 1.6 IOR is similar to glass, at 1.52. I’m not sure why he put it at 1.6.

“Lock Fresnel IOR to Refraction IOR”
Unlocking this is ‘unrealistic’. Water, glass etc. has the same IOR whether it’s reflection or refraction. But beauty trumps realism in this business, and scenes aren’t always perfect - so they added some flexibilty.

#3

Thanks for the answer. Yep, absolutely beauty over realism.

I just wanted to understand the physically correct settings prior to actually bending the rules.

My next question, if I was creating a metal material, why would that have an IOR value? After all there is no refraction isn’t it?

So as such, my result for all my materials with fresnel if I wanted to follow default, resulted in everything being 1.6.

So actually, do you guys change the IOR even if its a not refractive surface to get the fresnel to work accurately?

#4

If I unclick that and play with the Fresnel IOR, is that unrealistic? I thought in real life, things would have different fresnel values technically.

This is true, rubber, water, metal all different - but as far the exact value goes, you will never get truly physically accurate results as a lot of the time it’s not a linear thing we’re talking about.

Usually for materials I use anywhere in between 1.3-1.6 and for metals I turn off IOR. Some say to use incredibly high IOR values for reflective metal but it’s basically the same as turning it off.

I wouldn’t get too carried away with this, all render engines cheat anyway, even sometimes ones that claim to be unbiased employ a few tricks that are not exactly physically accurate.

In my experience, it’s the reference pictures that matter most - if my material looks like metal in the reference picture…great! It’s accurate enough for me and my clients!

#5

With metals we apply something called “Complex index of refraction”. I don’t think vrays IOR is physically correct though, it just tries to emulate the look of complex IOR.

Yes. With metals I usually end up with a IOR value somewhere between 3 and 10, depending on the look I’m after.

EDIT: Of course, as stated earlier, the only thing that counts is if it looks good. This is just my own workflow.

#6

Have you looked at this blog? Really good information:

Custom Fresnel Curves in Maya:

Part 1 - http://therenderblog.com/custom-fresnel-curves-in-maya/

Part 2 - http://therenderblog.com/custom-fresnel-curves-in-maya-part-2/

#7

I’ve been trying to figure this out too. IOR is basically how much faster light goes through a vacuum, than the particular element - so light goes through a vacuum about 1.5 times faster than through glass, or something like that. So does light travel 8 times slower through metallic crystals than through a vacuum? The highest one I see for metals is less than 3. But there’s no way to get a brushed aluminum look with that value.

Since it’s the literal value of that element, it doesn’t change just because its size or shape does. Meaning the IOR of glass doesn’t change when the glass is frosted.

It seems to me that we’re missing a feature - maybe glossy reflections need a ‘roughness’ attribute just like diffuse has? I’d assume that the roughness of frosted glass would be caused by the same thing as rough diffuse material; scratches and imputities at microscopic level, diffusing light more at glancing angles than at perpendicular ones, because they’re closer together in perspective - ?

Or maybe that’s uber-nitpicking and we should just keep using whatever higher IOR looks the way we want :cool: There’s no single roughness for any material, we can set it where we want.

It would be nice to know if roughness has a similar curve or not though.