azi,

The transition between lit and not-lit reflectance (called the terminator) is decided by the algorithm used for Diffuse reflections. In Poser, the default built-in calculation is the Lambertian reflectance model. There is another model called Oren-Nayer, implemented in the Poser Clay node. This has a roughness parameter that can somewhat sharpen the terminator.

But you've used the word "realism" and then described a behavior that is decidedly unrealistic.

I have tested the reflectance of paper and it is exactly reproduced by Poser's Diffuse node. So asking it to do anything different is asking for non-realism.

The reason the light fades as the angle changes is a simple consequence of the geometry of the situation. Imagine tracking millions of photons that are 1 micron apart, approaching a surface perpendicularly to the light path.  These photons will strike the surface with a spacing of 1 micron, so you'll have 1 million photons per linear meter.

Now change the angle so that they approach at a 45 degree angle. The photons striking the surface will land 1.414 microns apart. resulting in only about 707000 photons per linear meter and the surface will appear darker. At an approach angle of 1 degree, the photons land about 57 microns apart, or 17000 per meter, resulting in the surface being only 1.7% as bright as the 90 degree case.

---

Renderosity forum reply notifications are wonky. If I read a follow-up in a thread, but I don't myself reply, then notifications no longer happen AT ALL on that thread. So if I seem to be ignoring a question, that's why. (Updated September 23, 2019)