Rongnongno Loc: FL

At first, I was considering a filter, but this idea was quickly shut down, for good reasons.

I am now looking at creating a custom white balance using magenta.

My reasoning is as follows:

uniWB reduced all channels to the same value: 1
R=G=B=G=1. When simplified this gives us B=1, G=2, B=1.
Because of the overwhelming value of G the raw is green and any JPG created is very nearly unusable.

Since the complementary color for Green is Magenta, what happens if we create a Magenta WB?
In theory we should see:
R=1, G=0, B=1, G=0 which is not good.
If the magenta WB is set properly we get
R=1, G=.5, B=1, G=.5
In effect, we get R=1, G=.5+.5, B=1. We now have a new Unitarian WB based on RGB alone. The JPG then does not suffer.

In both cases the WB offered by the camera using the JPG would be accurate.

I am probably out of my **** mind to think that way, but it seems plausible if not doable.

rmalarz Loc: Tempe, Arizona

RGG,
My only thought would be to test it. However, keep in mind that uniwb mostly takes into account the larger number of green pixels to the R and B. That's what uniwb collects and the reason for the green color.

Yes, uniwb is useless if one is not capturing in RAW. The settings are camera-dependent. So, each camera has to have its own settings.

However, digital allows for a great deal of experimentation without a lot of cost. So, let's see what happens when your theory is applied.
--Bob

anotherview Loc: California

This explanation helps understand uniWB: https://www.malch.com/nikon/UniWB.html I didn't finish reading about it. I'll stick with the usual means of setting white balance.