|
Hello,
More experiments...
more madness... try this:
Take any regular RGB color shot... make it
something that has a fair amount of detail and range in it... look at the
histogram. If it is typical, the histogram will go across a fair amount of the
graph.
Now either convert it to
grayscale or go to saturation and delete all the color information out of it...
now look at the histogram... it SHRANK!
What does this mean for a
good LRGB? It probably means that if anything, the RGB information has to be
MORE than the L component for a good result... that is, the L histogram has to
fit INSIDE the RGB histogram by about 20%. THIS is why the RGB alone that I
printed came out so nice, and why the LRGB with more L than RGB was such a
disaster.
Am I obsessed?
Probably.... sorry....
Tony
|