*Wham!* *Wham!* *Wham!* (*bangs head on table*)

Are you /deliberately/ not following me?

So before I continue trying to answer your questions (I think I have 
answered them at least half a dozen times over), let's try to get this 
step by step:

(Chapter 1)

At the heart of POV-Ray, there is a rendering core designed to operate 
on /linear/ brightness values in three distinct spectral bands (R,G,B). 
Can we agree on this as a given fact?

If that is so (please interrupt me in case of any disagreement), then 
that core(!) rendering algorithm can be expressed as a function 
R(X,P)=Y, where X is the set of linear input brightness values, Y is the 
set of linear output brightness values, and P is the set of other 
core(!) parameters (such as scene geometry).

Let's also give a name to gamma-related transformations: Let's denote 
gamma pre-correction as g(Y)=Y', with Y in this case being a set of 
linear brightness values, and Y' being the corresponding gamma 
pre-corrected color values; let's likewise denote the inverse 
transformation as G(Y')=Y.

To finish up this chapter, can we agree that the rendering algorithm 
doesn't work properly for gamma pre-corrected values, i.e. that 
g(R(X,P)) != R(g(X),P)?

Post a reply to this message