[jomaga]

Do you remember the reflectance test made with Beta?
http://www.maxwellrender.com/forum/viewtopic.php?t=4585


I've updated the test with a scene with two areas, one hard lit interior with all the walls and the ceiling with a white material with 90% roughness and 90º reflectance variable, and another exterior with the same materials in the walls and ceiling.


In Beta version, white materials with a reflectance over 85% produce very flat images with loss of contrast and very high render times. In v1.1 this behaviour seem to be different, and high reflectances don't increase significantly render times, and the contrast keep almost constant.
So... good news
I want to make some more tests with longer render times and different scenes to confirm this behaviour, stay tuned.

[Frances]

Thanks Jose!

[Thomas An.]

Thank you for this test Jose !
(v1.1 is lovely indeed)

[yadikrisnadi]

Thanks Jose.

[sandykoufax]

Very useful test, Jose. Thanks.

[Hybaj]

Nice!

[devista]

gracias jose
saludos
luis

[samsam]

Jose - thanks for these new tests - your first set of tests were probably the most helpful tests ever in terms of getting to grips with the idiosyncracies of maxwell beta.

I would be interested in what material you are using for the wall - are you using a modified version of one of the diffuse plastic materials? Perhaps you might be kind enough to post the mxm?

This revised test reassures me about using full range white materials in V1.1, as well as using the full RGB (0-255) colour range for texturemapped channels.

[jomaga]

Thanks!
Samsam thanks for your kind words
The white material is very simple: one layer with Reflectance 0º variable from 200 to 250, Reflectance 90º full white, Transmittance full black, Nd 2, and roughness 79



Btw, here is the MXED dialog seen from Vista 64bits and Aero

[Mihai]

Nice! Could you please test the scenes also with a white lambert?

[Eric Lagman]

Were the new tests done with a co-op or on one computer. I am doing an interior now that has white walls like your test, on one pc. After 25 minutes or so at 640 x 480 it the noise is really bad. It looks like it would proably take days to clear up if at all on my 3ghz office pc.

[goncalo]

Thanks for the TEST

[abgrafx3d]

Thanks Jose for this info and testing!

[zoppo]

, Nd 2,

why did you use nd 2?

[Micha]

I think about the new test and wonder me. The first new scene looks like a semi exterior scene with much direct light and only a few effective GI bounce. So, we can not compare it to the beta scene.

The second new scene let me wonder, I can change the brightness of the wall, but get the same brightness ratio between the white wall and the interior materials at every image. I'm not sure - will we see fine color variations anymore? We can change the wall color brightness 20% and don't see any effect?
Also, the front part of the scene is very dark, so I feel, the scene dosn't use much bounced lights.

Why not the same scene again like for the beta? Are colors above 200 cutted? Most, textures are between 0 ... 255. Will this textures compressed in the bright part?

At Vray I know the option to set the secondary GI engine at 80%. Than I would get a GI light solution without dull contrast too. Do Maxwell the same now? Or is color 255 equal to 80% and Maxwell correct internal all textures and colors?