[daros]

only a old lightscape scene to compare light distribution between lightscape and maxwell. In this case Maxwell is better.
sorry no nice details.
Obviusly Maxwell one is better but i must say Lightscape was a nice software.
Lightscape lost 70% of his rendering quality since his integration in a famous 3d software...

Maxwell one.




Lightscape one. (1998)
[/img]

[Renato Lemus]

Beautiful renders both!!! but I prefer maxwell's.

Lightscape!! man, it takes me back...!! I loved that software.

[bader]

You right Adam,,, long time ago I thought that lightscape will be more around the 3d users then any render program . I though it would be the slang language for 3d .

daros I like both render, and The Maxwell one really awesome just by looking at the materials .I also think u should do a night render. It will be tough but at the same time fun to design a light system.

[Jun In Gi]

nice i like it also, i want to see indoor lighing.

[Maya69]

yes maxwell is better


just litle question for ground

you use plastic with specular map or diffuse with reflection mat ?

a+

[daros]

The problem of lightscape was the need of a very accurate modeling...no polygon interections, no high polygon count, no polygons with aspect ratio higher than 1:10, all vertices must be coincident... you needed about 3 times the modeling time as for other software. But then the rendering was very, very fast.
I made the last vork with Lightscape (last version from 1999 ) i think one year ago, DeBeers shop in NY. Now i would do it in Maxwell.

[daros]

and one more...

[Maximus3D]

They all look great Daros! beautiful work as everyone else says.
And to see good old Lightscape still rendering just makes me smile a little extra the old days of realistic 3d hehe

/ Max

[oz42]

beautiful images daros.

What material settings are you using for the lights on the tables? It looks like SSS but I've not seen it used that well before!

[daros]

hi thanks to all for your nice comments!
x oz42: no, it'snot a geman SSS, it's a a dielectric with roughness 0.6.

[tom]

amazing results!

[iker]

Braviiisimo daros!
...if this is the test, I'm anxious to see the final result!

[daros]

Thanks tom!
x fritz.arn, not cooperative and not single frame but bucket rendering. Each node renders a region of the image. It's not the correct way to make a distributed rendering on maxwell... as you see in the second image the caustics on the wall are not completly distributed. This is a problem caused by our approach.. but we avoid in this way the high network traffic of the mxi files. This is MW 1.33 Alpha
This renderings are distributed across 70 x 3 ghz cpu and each takes from 10 to 15 minutes.
But this are not the real Maxwell times since our system is not optimized. We need more time to reduce the noise of a image as the teorical time needed by Maxwell's cooperative Rendering (without MXI merging).

[thomas lacroix]

great, nice and warm, need a little more objects just to get it alive...