new vray blend mat (carpaint) test

[davision]

Shouldn't Fresnel Reflections be checked on the sheen material (#485)???

This adjustment would require the Blend amt be pumped up of course. But I always thought that the high sheen on cars used a fresnel falloff.

[lukx]

I will try it tomorrow and post image.

[lukx]

I've changed coat to fresnel (blend value bumped to 50), also becasue base mat is metal I changed it's BRDF to ward.

[davision]

Your previous settings definitely look better. However this new version needs more punch on the coat material to be a true comparison.

I think the blend value needs to be closer to 80-90% since the reflection at near perpendicular angles should be quite strong.

If you look at the side of a parked car as your driving towards it -- at a very glancing angle -- it is VERY reflective.

So if you have the patience for another render, please humor me. I'd do it myself but I'm assuming you don't want to release the contents of your max scene into the wild.

One other thing -- I don't know if this is physically accurate -- but maybe try increasing the IOR of the fresnel on the coat material to 2.2 or 2.5 so it doesn't drop off quite so fast, and change the BRDF to Phong since it's sharper than Blinn.

[lukx]

I bumped the blend value to 100. You're right about fresnel IOR casue I had deafult and this is rather for glass and this will be transparent paint coat so it has to be bump aslo I gave it 2 but might be also more. Anyway people asked me for settings so here they are.

[beestee]

We use fresnel in our car paint gloss coats with the VRayBlendMat

Example:



But like one of the previous poster said, you have to crank your values up dramatically.

[davision]

Thanks for making that change Lukx. I think it looks pretty good -- although your original render still looks good too.

Which one do you think looks better?

The problem I have with your original render is the lack of falloff -- it looks super shiny all across it and that doesn't ring true (but looks good anyway).

My problem with the falloff version is that the highlights still seem too dull. Increasing the IOR to 3 or 3.5 on the clearcoat would help but that doesn't seem physically accurate. I envision the clearcoat as essentially being a super thin layer of glass or plexiglass on top of the regular car paint -- so I would think the IOR would be down around 1.6 (even though it obviously looks wrong).

Instead of changing the IOR (and even though it would change the whole scene) -- I almost think the HDR should be a little brighter (25-30%) (or the exposure should be lighter if you're using a Physical Cam). It will lighten up the scene a little but the highlights will start to look right on the car. And I think the scene looks a little dark right now -- for an exterior anyway.

So, if you're in the mood to keep tweaking, give it a go. I appreciate your patience.

[lukx]

I think this is it. If others think the same I will post final settings. I changed HDRI map casue wanted to show better the blackness of the paint.

[davision]

Great job. I think it looks very nice.

For comparison's sake, you should post a render with this new HDRI and your original non-fresnel material.

Thanks for all the testing... and yes, I'd like to see whatever settings you came up with (eventually).

[Pixelab]

I made some test with an okd model I did for the peugeot car design competition.

Here are the results

You can find the settings on my (new) blog