Quick sky study image...

moodflow · January 22, 2008, 12:01:59 PM

This is just a quick sky study. I really liked the colors, so I figured I'd post it.

Seth · January 22, 2008, 12:03:33 PM

ooooh really nice colours !!!

nvseal · January 22, 2008, 12:05:21 PM

Fantastic!

Need to see it bigger.

dhavalmistry · January 22, 2008, 12:21:42 PM

very nice moodflow...

rcallicotte · January 22, 2008, 01:59:53 PM

Colors are nice. Is it TG2 or a picture?

Tangled-Universe · January 22, 2008, 03:42:02 PM

Quote from: nvseal on January 22, 2008, 12:05:21 PM
Fantastic! Need to see it bigger.

I second that

moodflow · January 22, 2008, 05:52:57 PM

Quote from: calico on January 22, 2008, 01:59:53 PM
Colors are nice. Is it TG2 or a picture?

Its 100% TG2

Will · January 22, 2008, 06:05:30 PM

nice, the clouds seems a little flat for some reason to me but it could just be my monitor. Great colors.

moodflow · January 22, 2008, 06:22:31 PM

Quote from: Will on January 22, 2008, 06:05:30 PM
nice, the clouds seems a little flat for some reason to me but it could just be my monitor. Great colors.

Yea, I noticed that too. I am not sure of a way around it. Maybe I could try some fill lights.

Mr_Lamppost · January 22, 2008, 08:30:27 PM

Fab colours

Agree: Clouds a little flat; no lit edges.

Matt · January 23, 2008, 07:34:49 AM

The reason the cumulus clouds look flat and don't have any edge highlights is because there is little direct sunlight reaching the cumulus clouds. In this situation in the real world the cumulus would then be lit by the cirrus layer and would have some edge highlighting from the cirrus's light, but TG's volumetric GI (multiple scattering) isn't quite sophisticated enough to capture that level of detail and simply assumes that the light from the cirrus illuminates the cumulus equally in all directions. (I haven't seen any other commercial volumetric renderer for CG do this properly, either.) I have some ideas to improve this in future.

Matt

P.S. Very nice render

Cyber-Angel · January 23, 2008, 08:01:24 AM

The sky its self would also be a source of illumination providing a source of indirect illumination, in order for the illumination of the clouds to start to be realistic here Terragen really needs Mie and Raleigh scattering in an ideal world (As in nature) multiple scattering in both the forward and backward directions, not as has been proposed in some of the literature a model based on single forward scattering only.

Regards to you,

Cyber-Angel

Matt · January 23, 2008, 08:13:59 AM

Mie and Rayleigh scattering are not what's lacking here, it's lacking anisotropic multiple scattering. Multiple scattering is supported by Terragen's GI, but it's isotropic, not anisotropic.

Matt

moodflow · January 23, 2008, 09:40:51 AM

Its great to hear there will be ongoing development.

Thanks for the compliments all.

rcallicotte · January 23, 2008, 01:00:37 PM

Uhhhh. Okay.

Quote from: Matt on January 23, 2008, 07:34:49 AM
I have some ideas to improve this in future.

Matt

P.S. Very nice render