3 Replies Latest reply on Oct 29, 2015 8:12 PM by Chuck Uebele

# Help me understand two axis (four pole) white balance when RGB (& most camera sensors) are three channel.

I'm trying to wrap my head around three channel RGB curve adjustments in relation to white balance adjustments. Are they directly proportional in any (interpolated) sense?

Another way of asking it I guess would be this: Can someone accomplish a RAW white balance adjustments using only curves? I'm assuming the answer is yes, but if so, where will problems begin most?

• ###### 1. Re: Help me understand two axis (four pole) white balance when RGB (& most camera sensors) are three channel.

White balance changes the overall cast of the image, where as curves modified selected tonal ranges. Two color correction, such as with White balance is very common with color correcting. Most pro labs used to just use yellow/blue and green/magenta to correct colors along with density. Red/cyan corrections was made by adjusting the two other "channels" or colors. It's faster for correcting.

• ###### 2. Re: Help me understand two axis (four pole) white balance when RGB (& most camera sensors) are three channel.

Thanks Chuck--really helpful. I didn't think of the Red/Cyan as static...

Given that these channels are present in RGB curves, can a white balance adjustment be made essentially by modifying the midpoint of the green & blue channels? Or how will a separate curves adjustment differ from a white balance adjustment?

• ###### 3. Re: Help me understand two axis (four pole) white balance when RGB (& most camera sensors) are three channel.

Using curves is actually difficult to adjust white balance, as it's hard to adjust the highlights and shadows properly. Using L*A*B* curves is more in line with adjusting white balance. As far as red/cyan being static, it's been that way with wet processing color photos for years. The adjusting color with an enlarger, one would just adjust the amount of yellow and magenta in the enlarger's filter pack. Adjusting these would change the density, so the time would be compensated. The cyan filter pack would never be touch, unless you had to remove all of either the yellow or magenta filtration. I've only had to do that a few times in the many years that I printed color. Using all three colors would just add or subtract density, so you only worked with two colors. For a while I was working as a color corrector in a pro lab. We would get slips of paper with the prints' filtration. A perfect image would be some thing like 65 30 30 30. the first number was density, and the others were cyan/red, magenta/green, and yellow/blue. The only time we touched the cyan/red was if the values got other values got down to 0. While it's not necessary to correct images this way in PS, I still find myself just modifying the G & B curves.