Originally reported on Google Code with ID 1478
Photoshop, Gimp and kColorChooser use ° for hue.
Photoshop and Gimp use % for saturation and value/brightness. RT and kColorChooser
use 0-255 for saturation and brightness.
I suggest RT use % for saturation and value, in line with the other two major programs.
Second problem, the hue value in RT differs from that reported by Gimp. I don't have
Photoshop, you test.
Tested using this photo http://rawtherapee.com/shared/test_images/k10d%20x-rite%20passport%20fluorescent.dng
Denoising values 60, 20, 50, 1.3 to smooth things out.
Zoomed to 100%.
Working space = sRGB, output space = No ICM; sRGB.
Sampled the right-most column, starting from the top:
--------RT------- -------Gimp-------
H=212, S=5, V=50 H=270, S=4%, V=20%
H=127, S=3, V=80 H=180, S=1%, V=31%
H=0, S=2, V=112 H=60, S=1%, V=44%
H=85, S=1, V=149 H=120, S=1%, V=58%
H=85, S=1, V=184 H=60, S=1%, V=72%
H=42, S=6, V=222 H=60, S=3%, V=87%
Hue was difficult to measure as it can jump 0, 60, 120 over the same sampled square.
I tried to select the value that appeared most often, though sometimes, as in the 0,60,120
example above, they all appeared more or less the same number of times.
kColorChooser's H values were similar to Gimp's.
Point is that hue in RT seems way off than in other programs. Does the counting start
rotated by 60°?
Reported by entertheyoni on 2012-07-10 16:28:28
该提问来源于开源项目：Beep6581/RawTherapee
