[vtkusers] volume rendering transfer functions

[Sean McInerney]

Hi,

   This question is tangentially related to the previous discussion of 
creating transfer function editors. It is fairly straightforward to 
represent the transfer functions with a 1D plot of the color and scalar 
opacity (RGBA, HSVA, or luminance-alpha) for the range of scalars 
mapped. However, I am unsure how the gradient opacity interacts with the 
scalar opacity to produce the mapped opacity. Without necessarily 
considering integration along a ray, can anyone provide a simple 
equation explaining how an opacity value is mapped from the two opacity 
transfer functions. I am assuming that the gradient opacity transfer 
function is mapping opacity to the gradient magnitude. An editor 
interface could thereby use a 2D plot to represent the transfer 
functions with the gradient opacity plotted on a y-axis over the range 
of gradient magnitudes in a data set. Is the mapped opacity the average 
of the two lookup values ... or something spiffier? ... calling Lisa 
Avila to the rescue.

-Sean