Tweet
I'll try and explain better why i use "such dark" materials...
In normal conditions, without a physcam, or HDR lightsources (so say, standard max omni light with multiplier at 1, and standard camera), i always used the whole spectrum of whites.
A half step back might be required: the final pixel color can be rougly approximated, for a perfectly diffuse surface, as the surface brightness*the light brightness (do not stone me if i trash like so the shading algorithms, it's for figure of speech... :P). So in traditional rendering, a value of 255 white for the material and a value of 1.0 for the light would return me an average 255 white.
Now, with these new tools we have, the incoming Lights are EXTREMELY bright, very often (sun is a light with 1000.0 as multiplier, or thereabout, i gather).
Of course, in a perfect world, VRay would take care of everything (and given enough time it might well do), but as far as my experience goes, lowering the material brightness to 120 for a white material leaves me ROOM for specularity, on top of helping the light decrease its bounced intensity more rapidly.
I speak of having room for specularity, and here the other "strange" thing: rarely, if ever, my specular levels exceed 120, as well
120 for the surface X a value of 3.0 in some pixels of the hdri would still give me an over-the-white reflection, and modulate correctly around the hotpoints.
Notice how 120 for diffuse and 120 for specular sums up 240, which is LOWER than 255 pure white (i'm also fairly conservative).
This material, to me, would be a good candidate for white enamel, for instance:It would obey the energy conservation laws, and through exposure and lights i can easily make it look as shaded or as lit as i want, and the same i can do to reflections...
We now have the VRay COLOR map, and i am using it a lot, as it is fairly flexible, and most of all, it expresses RGB colours as normalised to 1, rather than with the old 8 bit method.
Still, the energy preservation law would dictate the material diffuse+specular (the two light bouncing components) HAVE to be lower than 1.0.
Change it to 250 white with 250 reflections, and bang: energy preservation is broken, material will likely look burned out (it would in LDR, go figure with HDR lighting), and GI methods have to work double time to try and make any sense of outgoing light being stronger than the incoming.
I hope to have made the concept more of a messy affair
Lele
In normal conditions, without a physcam, or HDR lightsources (so say, standard max omni light with multiplier at 1, and standard camera), i always used the whole spectrum of whites.
A half step back might be required: the final pixel color can be rougly approximated, for a perfectly diffuse surface, as the surface brightness*the light brightness (do not stone me if i trash like so the shading algorithms, it's for figure of speech... :P). So in traditional rendering, a value of 255 white for the material and a value of 1.0 for the light would return me an average 255 white.
Now, with these new tools we have, the incoming Lights are EXTREMELY bright, very often (sun is a light with 1000.0 as multiplier, or thereabout, i gather).
Of course, in a perfect world, VRay would take care of everything (and given enough time it might well do), but as far as my experience goes, lowering the material brightness to 120 for a white material leaves me ROOM for specularity, on top of helping the light decrease its bounced intensity more rapidly.
I speak of having room for specularity, and here the other "strange" thing: rarely, if ever, my specular levels exceed 120, as well
120 for the surface X a value of 3.0 in some pixels of the hdri would still give me an over-the-white reflection, and modulate correctly around the hotpoints.
Notice how 120 for diffuse and 120 for specular sums up 240, which is LOWER than 255 pure white (i'm also fairly conservative).
This material, to me, would be a good candidate for white enamel, for instance:It would obey the energy conservation laws, and through exposure and lights i can easily make it look as shaded or as lit as i want, and the same i can do to reflections...
We now have the VRay COLOR map, and i am using it a lot, as it is fairly flexible, and most of all, it expresses RGB colours as normalised to 1, rather than with the old 8 bit method.
Still, the energy preservation law would dictate the material diffuse+specular (the two light bouncing components) HAVE to be lower than 1.0.
Change it to 250 white with 250 reflections, and bang: energy preservation is broken, material will likely look burned out (it would in LDR, go figure with HDR lighting), and GI methods have to work double time to try and make any sense of outgoing light being stronger than the incoming.
I hope to have made the concept more of a messy affair
Lele
Comment