Page 1 of 1

Alaska Blue Lake

Unread postPosted: Wed Nov 02, 2016 1:43 pm
by xilofoton
Dem_Alaska_v011_Comp_01.jpg

Dem_Alaska_v012_Comp_01.jpg


Hi,

in a nutshell the displacement is driven by raw LIDAR satellite DEM (TIFF)
http://viewer.nationalmap.gov/basic/

Everything else is procedural, ie the snow cover is based on height, slope and wind exposure.

Scales are just by eyes, the trees are too virulent at certain areas... so it needs more informations from the real site.

Re: Alaska Blue Lake

Unread postPosted: Wed Nov 02, 2016 9:26 pm
by David_A
Very interesting atmosphere. Are you using the PBR for these?

Re: Alaska Blue Lake

Unread postPosted: Thu Nov 03, 2016 10:01 am
by xilofoton
Yes, PBR, path traced depth map ;)
Cheap and noise free, with comped glows it gets some illumination from the snowy areas.
Setting proper colours is much more important than doing physically based simulation (which cannot be accurate within RGB space and without wavelenght dependent scattering, especially in landscape scale environments).

Re: Alaska Blue Lake

Unread postPosted: Thu Nov 03, 2016 11:21 am
by Germanobanano
Nice!

About the PBR volumes for environments, playing with the absorption and scattering values you can get some realistic aerial perspective. I have to look more into it, but I used a volume to simulate rayleigh scattering and another for mie scattering.

Re: Alaska Blue Lake

Unread postPosted: Thu Nov 03, 2016 11:39 pm
by daaims
Really good damn atmos' here !

Cheers ;)

Re: Alaska Blue Lake

Unread postPosted: Fri Nov 04, 2016 10:51 am
by xilofoton
Dem_Alaska_v015_Volume_03_PS_v02.jpg


Thank you guys!

Now I tried rayleigh+mie like scattering with 3 volume containers. This half resolution took more time to render than the previous HD renders. I can avoid the volume noise only with the sample clamp. Earlier with the legacy I tried similar multi volume setups, it was faster and less noisy.