"Chris R" <car### [at] comcastnet> wrote:
>
> I do have radiosity enabled, although I haven't influenced it with the sky
> sphere at all.  I have probably confused things with multiple versions of the
> same scene in these discussions, so do you find even the latest chess board
> closeup, (the 1920x1080 version) too dark for a moonlit scene?  The earlier
> versions had the camera too low, so too much of the scene was in the shadow.
>
>
> -- Chris R.

Yes, all seem too dark, most more importantly too contrasted : the dark areas
are pitch black as if you hadn't applied the linear workflow which generally
rather creates slight nuances in such shadowed areas, those change everything in
the unconscious perception.

And all psychology left aside, the sky light is already very important even when
the sun is up,  providing a blue tint to every shadow, so imagine as the sun
disappears that this shadow (which is not actually darkness but blue  light from
the surrounding atmosphere) grows on everything before the moon rises. Then the
moon is more like a torch light in the middle of that, that's why factoring the
sky_sphere in radioisity would be important, and probably that's why most
artists tend to refer to the moonlight scenes as bluish. (though the moon beam
is lower kelvin colored). (scattering media might here provide some interesting
gradient if the wall hidden moon light source would be pointing almost straight
into the camera, bleeding its halo into the window sky ? but it depends on the
render time cost.)

The hard thing to balance in daylight scene is the strength of the sky and the
sun, but is even harder in moon light. because you must first find the relative
weight of both, as you did, but then blow them up in synch so as to really
compensate for eye adjustment or camera exposure (which POV doesn't have, or
does it?).

Also probably using a high dynamic range format as oopen EXR to tonemap it could
be a good flexibility tool.

Enjoy your journey to infinity and beyond! :-)

Post a reply to this message