So you simply select a white spot or a black spot in an image and click the appropriate eye dropper in it and color correction happens?
So you simply select a white spot or a black spot in an image and click the appropriate eye dropper in it and color correction happens?
If set to RGB not individual channel just a tonal adjustment not colour.
I generally don't use the eye dropper tool in curves for that purpose, but as Paul has suggested, the black eyedropper will set the black point, the white eyedropper will set the white point and the grey eyedropper sets the part of the image one clicks on to gray,
However, if the points are not black, white or gray, it will set the individual colour channels as if they were and that can get you into trouble if the select point is not neutral. The levels adjustment (not just curves) works in a very similar manner.
I tend to use the individual colour channels in curves to do tricky white balance work when some of the simpler technique don't give me the results I want.
If you could expand upon your color Curves correction technique, GrumpyDriver, in your typically easy to read and understand manner, the entire world will be a better place.
I'd like to but it is one of those techniques that are very long and hard to write down, but much easier to demonstrate. It is only required for extreme issues like restoring a damaged / faded picture. Each colour channel is corrected for black and white points individually.
If you have the raw data, you don't need anything this sophisticated.
Last edited by Manfred M; 12th January 2017 at 01:11 AM.
Yes, I was very happy you were asked to explain...![]()
If I ever use curves to correct colour (which I did yesterday after being requested to replace a babies face with that of her father) I just tweek the individual channel curves without doing any sampling. It's right if it looks right.
If you can find a tone that should be white (i.e. a light tone that is close to white that would have equal R, G and B values) and a tone that should be black (i.e.a dark tone that is close to being blackthat would have equal R, G and B values) , it takes away some of the guesswork. Juggling 6 different data points and getting them to balance out is the tricky part).
Regardless, it is a very powerful approach, but as you have shown with your example, not really a mainstream application.