Graphics Reference
In-Depth Information
Fig. 4.60 The numbers for gain and lift. Check the minor difference between these and the numbers in
the previous illustration. Now it's time to make sure that the midtones are also balanced. With the high-
lights and shadows balanced already, the midtones should be pretty close.
Notice, though, that the similar shapes that sit just under the 60IRE
line in the RGB parade are not quite at the same level in all three chan-
nels, even though the white chips and black chips are even. I'm going to
match the red and blue midtone shapes to the green midtone. The mid-
tone corrections will have a fairly strong effect on both the highlights and
shadows, so I'll have to do some back and forth corrections between the
three tonal ranges until I have the shadows, midtones, and highlights all
even across the three color channels. Figure 4.61 shows my final result
with the numbers and RGB waveform.
With the amount of clipping in the red channel, this correction is
problematic, but you can see from the vectorscope and the 3D Color
Space that most of what is supposed to be neutral is actually fairly neu-
tral. There is a slight red cast in some of the chips still, due to the clip-
ping of the red channel. In a real-world image, you could determine
whether this slight warming was appropriate and what you were will-
ing to sacrifice in other colors or tonal ranges to try to get these into
“spec.”
Let's apply this to grading a real-world image with these same tools.
Call up “piano_cool” from the DVD into your chosen color correction
application or plug-in. I'll do this one in Color Finesse.
Figure 4.62 b is the starting RGB Parade waveform display for the piano
cool image. It shows that the black levels are slightly elevated across all
 
Search WWH ::




Custom Search