Volumetric Fog feedback thread

[SirKlu]

@DanielW

I had to shelve the implementation unfortunately due to other priorities. I am probably tied up for the next 6 months =( Sorry for the bad news. It is very easy to implement it just needs more testing, and I just don’t have the time.

Love the galaxy thing, looks amazing. Wish I had gotten around to Participating media self-shadowing. Whenever you make dense volumetric materials, it looks bad due to lack of self shadowing. It’s a major task though. Like regular shadowmaps but 3d instead of 2d, so very hard to make efficient.

Volumetric Fog is lit by Volumetric Lightmaps, actually that was one of our main motivations for creating a new volumetric lightmap representation, it needs to support GPU interpolation to affect Volumetric Fog efficiently, and previously we were relying on a method which only supported CPU per-object interpolation.

This effect is due to the tonemapper and has nothing to do with Volumetric Fog - you can reproduce the same thing with a translucent sheet. The tonemapper has a non-linear high contrast ‘toe’ which causes this effect. You can reduce the contrast, or use the old tonemapper.

The depth distribution is already exponential and you can control the exponent with r.VolumetricFog.DepthDistributionScale.

Everything here sounds @DanielW !!
I understand the that you guys are currently busy working on other stuff, which I’m also looking forward to for sure.
My biggest question right now, I’ve seen that you understand the problem with selfshadowing, and of course you would like to have it, but It wasn’t clear if this would come to the engine eventually.
I mean, for a feature like this I could wait all the time needed, but it would be nice to know if you guys are planning to add it in the future.
Amazing stuff could be done with volumes with the shape of the distance fields (even more if they are dynamic ones), plus self-shadowing.
I’m really looking forward to those two features coming!
Thanks for your time!