Backscattering?

Started by Eneen, January 08, 2020, 03:36:55 AM

Previous topic - Next topic

Eneen

I've noticed something interesting on real-world spherical hdris. Take a look at this picture from hdrihaven:
quarry_01_s.jpg
quarry_01_s_marked.jpg
As you can see part of sky that is facing towards sun is brighter then "sides".
Same thing is visible on evening, clear skies from here:
http://noemotionhdrs.net/hdrevening.html

Question: is it possible to achieve this in TG?

N-drju

I might be wrong, but it might be the "Rayleigh Scattering" effect.

Try it out on a completely flat ground (like in your HDRI) and check out. Perhaps it is visible by default?

If not, the fastest (though uninventive) way to achieve such an effect is to... add another sunlight opposite the mian light source. A 0.5 - 1.25 powered, white sunlight, lowered just below the horizon. If you opt for this solution however, keep in mind that you'll need to tweak atmosphere as well to avoid colour decays.
"This year - a factory of semiconductors. Next year - a factory of whole conductors!"

WAS

Could try a localized haze layer on the horizon.

N-drju

That would have to be done carefully and with a considerable falloff in whatever masks the haze layer. Otherwise, it just might look like a bright spot pasted onto the sky. A vertical SSS could do the trick if only it had existed. 🙂
"This year - a factory of semiconductors. Next year - a factory of whole conductors!"

WAS

Quote from: N-drju on January 08, 2020, 12:16:05 PM
A vertical SSS could do the trick if only it had existed.
🙂

Just gotta flip it.

N-drju

Oh? Are you sure this can be done as an atmospheric mask...?
"This year - a factory of semiconductors. Next year - a factory of whole conductors!"

WAS

#6
Not positive. In my experience a SSS circle in soft shape will act as a dome or circle/ball in cloud density.

Maybe I'm not positive what we're after here, but I imagine it's the glow of the horizon haze which makes it look like the sun is "growing" up from the horizon (as far as light)

A huge SSS mask masking out your terrain would give you a haze on the horizon, which would play with the light against the sun, as well as the light on the opposite horizon being lit by the sun.

WAS

#7
Try this on for size.

There are two "area" masks. One is using a camera for the "rear" backscatter only, the other is 360 and will affect both the sun and backscatter. The line mask can further be rotated with the sun position. Cloud layer density can effect intensity.

I tried playing with the lighting for a scattering effect with colour modulators but it's so faint they don't really apply apparently.

Edit: you can disable the line mask and use the 360 mask instead of the distance shader for nice horizon haze, playing with the haze altitude settings.

Edit2: if you are away from 0,0,0 you can just change the SSS locations naturally.

Eneen

Thank you guys!
Ok, so looks like it can't be simulated and is rather eyeballed? Maybe @Oshyan can chime in. I think it's quite important that atmosphere model misses this...

WAS

Would be a nice feature to complete the atmosphere model. I believe the specific effect you are after is called the Belt of Venus.

Matt

Rayleigh scattering has the following phase function:

http://glossary.ametsoc.org/wiki/Rayleigh_phase_function

Terragen 4.4 doesn't take this into account, but I'm going to add this to Terragen 4.5.
Just because milk is white doesn't mean that clouds are made of milk.

WAS

Quote from: Matt on February 07, 2020, 02:44:54 PMRayleigh scattering has the following phase function:

http://glossary.ametsoc.org/wiki/Rayleigh_phase_function

Terragen 4.4 doesn't take this into account, but I'm going to add this to Terragen 4.5.

That's awesome. Awesome thanks for that link. Bookmarked.