Christian Froeschlin wrote:
> Hmm, if you use only radiosity with high ambient objects instead
> of light sources, wouldn't they reflect in the sphere without need
> for special highlights?

  No, because radiosity is calculated purely diffusely: The direction of
the incoming light has no effect whatsoever on the distribution of the
outgoing light, which will always be the same (a hemispherical
distribution weighted by the angle between the outgoing light and the
surface normal). No specular hightlights are possible with this method.

  Specular highlights are caused by the specular reflection properties
of many surfaces: The distribution of the reflected light depends on the
direction of the incoming light (usually in the way that more light is
reflected to the mirrored direction than in a purely diffuse case).

  The phong lighting model used by POV-Ray (not to be confused with
phong highlights; the phong lighting model is a larger concept) might
not be the most physically accurate, but it often produces results which
are visually better than a purely diffuse lighting model would. In
POV-Ray you need light sources in order to fully utilize the phong
lighting model (iow. to get the pretty highlights).

  In theory you could get highlights without light sources by having
voluminous bright objects (serving as "light sources") or a HDRI
environment (which contains luminous areas), and specifying a blurred
reflection and proper reflection exponent, but this will usually make
rendering very slow with little benefit. (The only benefit would be that
the shape of the luminous areas is correctly taken into account in the
highlights. Of course at the cost of enormous rendering times.)

Post a reply to this message