scott <sco### [at] scottcom> wrote:
> That's exactly the sort of thing I was thinking of.  Maybe Warp will 
> argue that the default should be "perceptual" though, and we need a 
> "linear" keyword to revert to the existing behaviour.

  The thing is that most people are used to specifying colors as perceived
by the human eye rather than in watts. For example, doubling the values of
the color components is expected to double the brightness of the color
(iow. eg. 'rgb 0.8' is assumed to be twice as bright as 'rgb 0.4').

  To elaborate, if I have understood correctly, POV-Ray 3.7 has been
changed so that regular color specifications express the power of
luminous radiation (which is called radiant flux, and in physics is
measured in watts) rather than the perceived luminosity as seen by
the human eye. The relation between these two is not linear (but closer
to logarithmic). This means that eg. doubling the radiant flux (ie.
doubling the "wattage") does not correspond to doubling the perceived
luminosity of the color, as seen by the human eye.

  This might correspond more closely to reality when calculating
illumination. For example surfaces reflect a portion of the light they
receive, and this portion is relative to the radiant flux, not to the
perceived brightness. (In other words, if the surface properties and
angle with respect to incoming light is so that it reflects exactly
half of the light it receives, this half is measures in watts, not in
what the human eye perceives as "half bright".) I suppose that at least
in theory this ought to give a more realistic end result for surface
illumination, ie. a result which corresponds more to real life.

  As said, the only problem is that people are accustomed to specifying
colors and color gradients in perceived luminosity, not in watts. This
can and will cause confusion.

  In POV-Ray 3.6 color specifications correspond directly to pixel
component values, and this happens to be close to linear with respect
to the perceived luminosity, and hence 'rgb 0.8' looks about twice as
bright as 'rgb 0.4', and this is what people are accustomed to. Likewise
linear color gradients in POV-Ray 3.6 (which, as said, simply map
directly to pixel values) happen to be close to to perceived linear
brightness, which is also what people are accustomed to.

  "You don't specify perceived brightness anymore, but absolute brightness"
is a rather radical change, and many people will get confused by it,
especially since in most systems (at least those with a gamma of 2.2)
raw pixel values map almost linearly to perceived brightness.

-- 
                                                          - Warp

Post a reply to this message