What's new
Photoshop Gurus Forum

Welcome to Photoshop Gurus forum. Register a free account today to become a member! It's completely free. Once signed in, you'll enjoy an ad-free experience and be able to participate on this site by adding your own topics and posts, as well as connect with other members through your own private inbox!

Duplicating colours and luminosity


Marty86

New Member
Messages
3
Likes
1
Hello everyone. Photoshop newbie here, so please excuse the possibly dumb question, but this has kept me awake at night!

Let's assume I have two identical images. One (image A) has been edited and the other (image B) is the original untouched image. Is there a way to make the colours and luminosity in image B match exactly that of image A? Now I know what you're probably thinking right now; "Why not just duplicate image A?" but for the sake of this question, just assume that can't be done. Is it possible to colour match two identical images?
 
Last edited:
Just tested this - if you want to copy the exact same adjustment layers to the alternate image, place all the adjustment layers into a group. Then highlight the group.
Or alternately just use your shift key to highlight all the adjustment layers simultaneously

*Make sure both both the edited and unedited images are open

1622296082156.png

Now right click on one of the open areas on the layer (see "X" below) and a dialog box will come up
Click on "Duplicate Layers":

1622296448612.png

Now another dialog box will come up that will ask you where you want the layers copied to.
In the dropdown, menu, choose the unedited image. The click OK.

1622296684314.png

The adjustment layers should all now be in place on the original unedited image with the changes showing on the image.

1622296814472.png

There may be other ways to get this done - this one worked for me. :cheesygrin:

- Jeff
 
Hey Jeff. I could be very wrong but I think the question is how do you duplicate the color and luminosity of an edited image, without knowing what was actually done to the edited image

Perhaps something like this.....
cinderellas-revenge-before-and-after-photoshop.jpg

If you had a copy of the original and you wanted to copy the color and luminosity of another edited image of the same original but you don't know what was actually done.

I could be totally wrong though.
 
Hey Jeff. I could be very wrong but I think the question is how do you duplicate the color and luminosity of an edited image, without knowing what was actually done to the edited image

If you had a copy of the original and you wanted to copy the color and luminosity of another edited image of the same original but you don't know what was actually done.

I could be totally wrong though.


Thanks for the detailed answer Jeff, and IamSam is correct. All I have is a flat copy, with no adjustment layers. The images themselves though are 100% identical, other than the colour/luminosity. :)
 
If the two photos are extremely different because one of them has had twenty filters and effects applied to it, then it may be difficult/impossible to undo all that and get them to match. But if they are not so extremely different, then you may have success using the Color Sampler tool. The basic technique goes like this:
  • Select one or two locations on Photo A with the color sampler tool. The Info window will automatically pop-up and display the RGB color value of the specific spots you chose.
  • Select the exact same locations on Photo B with the color sampler tool.
  • Use Curves, Levels, etc on Photo B to manipulate the RGB values of those specific spots. The aim is to get the RGB values in Photo B to be identical to Photo A. Once you've successfully gotten those individual locations on each photo to match, the rest of the Photo B should (hopefully) go along for the ride and match Photo A.
A good place to start is to set the black and white points and get both photos to match. That should take care of the overall luminosity of the two images. See video below.

If your image has people as the main subject(s), then you may want to separately use the Color Sampler tool to select a representative area of skin in Photo A—let's say the center of the forehead—and then use the Color Sampler to select the exact same spot in Photo B. Then open a curves adjustment and separately adjust the Red channel, Green channel and Blue channel to get the skin tones on those specific spots to match.

You say you are a photoshop newbie, so all of these Curves adjustments may be too much, too soon. But check out the video below because, aside from your specific question, knowing how to set the black and white points is a good place to start editing on many photos.

 
Here are two photos identical in every way except for color and luminosity. But the difference is so extreme that it would be very hard (impossible for me) to get Photo #2 to match Photo #1.
In this case, all I did was duplicate the layer and change the blend mode to Exclusion. That is only one effect—I could have done ten more. But even if you know ahead of time that Exclusion is what I did, I'm not sure how to reverse the effect and restore it back to Photo A.

But my method above will usually work if the difference is merely one of exposure, or maybe a modest color cast.

1622307115846.png
 
Hello everyone. Photoshop newbie here, so please excuse the possibly dumb question, but this has kept me awake at night!

Let's assume I have two identical images. One (image A) has been edited and the other (image B) is the original untouched image. Is there a way to make the colours and luminosity in image B match exactly that of image A? Now I know what you're probably thinking right now; "Why not just duplicate image A?" but for the sake of this question, just assume that can't be done. Is it possible to colour match two identical images?

Hi @Marty86
A lot of good answers before and I will add some it. It may not put the issue to bed yet hope it lets you sleep at night :)

Here is my assumption for which you seek. You are asking can one deduce the equation/function/transformation that is applied to all pixels of the starting image that will then create the final image. It is a bit more general statement than your request yet the answer is the same.
In the general case the answer is NO.
However, if the functions/equations/transformations meet specific criteria then the answer is YES with varying levels of difficulty.

What that means is that:
1) every unique RGB outpu triple can only be created by a single RGB input triple and (Photoshop does not guarantee this)
2) It also means each unique RGB input triple unambiguously creates a single RGB output triple (I don't think Photoshop allows anything but this second option so probably a moot point).

A perfect example of one that fails the first point is the example from @Rich54 above.
It turns out that having two identical image Layers with the Exclusion blend applied is exactly the same as having a single Layer with the following Curves Adjustment Layer Applied:
Screen Shot 2021-05-30 at 5.05.41 PM.png

It cannot pass what is called the horizontal line test.
If you put a horizontal line through a the above curve, each RGB output up to 128 in output can be created by two different input levels. So the horizontal test fails. If you try and deduce which RGB output value of a given horizontal line was created by a given RGB input, you would not know which of the two RGB cross points created that output. Its ambiguous.

Case in point, you can get the same resulting output with @Rich54 example if you used the inverted version of the starting image as shown in this image below (inverted original on left and result on the right):
Screen Shot 2021-05-30 at 10.22.04 PM.png

Reverse engineering of the transformation can be done when the transformation meets the criteria provided above.
The easiest is when the transformation is done on a color channel basis and is a linear transformation
Next most difficult is when the transformation is done on a color channel basis and is non-linear
Hardest in my mind is when the transformation involves an output channel that is dependent on 2 or more color channels.

Even though you asked your question in terms of Luminosity and Color (Hue and Saturation as separate channels of Color) the concept is exactly the same.

So in general NO, under limited circumstances YES.
Hope that helps
John Wheeler
 

Back
Top