• Welcome to Operation Photo Rescue's Online Community.
 

Non-destructive curves color adjustment tutorial

Started by cmpentecost, February 13, 2009, 04:18:46 PM

Previous topic - Next topic

cmpentecost

Through some email discussions with Margie, I offered to do a curves color correction tutorial, based on instruction from Moose Peterson (www.MoosePeterson.com ~ I highly recommend his blog), and Ben Willmore (www.digitalmastery.com).

The photo example I am using was restored a long time ago but a good example on using the curves color correction.  I should also note that this tutorial is based on using CS3.



You notice in this image the yellowish color cast and that the histogram is not touching either end of the box.  It is bunched up, more on the left side, which means it is too dark.  Anything on the far left is black, anything on the far right is white. 

Step 1:  Open your image, click the Create New Adjustment Layer at the bottom of the layers panel and select Curves.  Click the show clipping check box when the dialog appears.  You'll notice your image goes all white.  Once you start moving the white slider (see below) the box turns black.



Click and drag the white slider (far right white triangle) and side it to the left until meaningful white appears.  It could be red, yellow or black.  Turn the Show Clipping checkbox on and off to see where the whites you're seeing appear in the photograph.

Once you've located a meaningful white, zoom in (if necessary) on the noticeable pixels by pressing Ctrl (Command) + (to move back out, it's Ctrl (Command) -.  Press the spacebar to make the hand tool appear and then you can move your image around if necessary.  Next, shift-click the sample to set your white point.   



When you are finished setting the white point, click the black slider located at the left of the histogram, above the word "input".  Your image will go all black (make sure you still have the show clipping box still checked).  Gradually slide the black slider to the right until you find meaningful black.  As above, zoom in if necessary and shift-click to set your black point.



Now that you have your black and white points set, uncheck the "show clipping" box.  Your image appears.  Now, take your white eye drop and click on your white point.  Next, click on your black eye drop and click on your black point.  Click OK to close the curves dialog box.



As you can see, the color cast is removed, and by looking at the final histogram (flattened), the histogram stretches more of the length of the histogram box and there are no large spikes to indicate blown out colors.

I hope this makes sense, but if you have questions, please let me know.

Christine




kiska

kiska
Photoshop 2021, MacPro

Hannie

Hi Christine,

For a long  time I have been wanting to learn more about this subject but it just never seemed to happen.
When I saw the title of your topic I was so pleased, I can't wait to try it out for myself!

Thanks so much!

:hug:

Hannie
Hannie Scheltema
Distribution Coordinator
[email protected]

cmpentecost

This technique won't work perfect for every photo, especially if there is heavy, heavy damage.  You have to watch out that the white point is set on something actually in the photo, and not damage.  Sometimes I reduce the opacity of the curves adjustment if it's too bright.  I use this technique on a lot of my photography, and it usually works well.

Christine

Mhayes

Hi Christine, thanks for the great tutorial! This was easy to follow your instructions and the screen prints were great! I am definitely going to try this out.  :up: :up:

Margie
"carpe diem"

Margie Hayes
OPR President
[email protected]

Candice

Candice

Atlantis

It is almost similar to the other curves colourcorrectionmethod first described by Glennab and later by Mitzs : http://www.operationphotorescue.org/forum/index.php/topic,1811.0.html but taking a different and faster route.

I do like the suggestion in the other method to set the white to 96 and black to 4 to get more definition in the blacks and whites.

I think it would be interesting to take one photo and use all methods to see how different the results may be.
The only way to get better is to figure out what I did wrong.