Home · Register · Search · View Winners · Software · Hosting · Software · Join Upload & Sell

Moderated by: Hendrik, Fred Miranda
Username   Password

  New fredmiranda.com Mobile Site
  New Feature: SMS Notification alert
  New Feature: Buy & Sell Watchlist
  

FM Forums | Post-processing & Printing | Join Upload & Sell

1
       2       3       end
  

Fixing color in LAB
  
 
ben egbert
Online
• • • • •
Upload & Sell: On
p.1 #1 · p.1 #1 · Fixing color in LAB


The sky in the first image was said to have a color cast identified as green. I did not see it myself but I believed it was probably true. There are two problems here for those of us with color vision issues.

1. How to discover a cast.
2. How to fix it.

Reading Photoshop Lab color by Dan Margulis, chapter 8 provided two clues. I want to show what I did and then solicit other ideas. But just remember, I need ideas that work on objective things like color numbers, because my color vision is highly suspect.

The first thing I got from the book is to open the color picker, and then use the eye dropper to click in the upper left corner sky area. This returns a color on my system at R=43 G=57 B=90. Now I can see in the color swatch that this is very near the green zone. This is not something I could see without this comparison.

Next I went into LAB, and selected the darker blues using “select color range” I then opened a curve layer and adjusted the B channel by clicking at the center and dragging it to the left by .10 (one small square in the grid). This moved the blues to what appears to be a more blue hue. In LAB I get L=29 A=-14 B=-47. Note I did not touch the L or B. These numbers mean nothing to me yet as I don’t understand LAB math.

I now went back to RGB and see the blue is at R=40 G=57 B=115. So I see I added blue, had no effect on G and a bit on R. But it looks better to me.

The second image is the result. How would you do it?






image with cast in sky







image corrected in LAB




Mar 19, 2014 at 03:06 PM
Bernie
Offline
• • • •
Upload & Sell: Off
p.1 #2 · p.1 #2 · Fixing color in LAB


When doing landscapes, there are no correct numbers -- especially with a CP. Because of clouds, sunrise / sunset colors, leaves, clouds from forest fires, etc "correct" color can be whatever is pleasing to your eye for the circumstance that you are capturing -- and the mood you are portraying. JMHO

Also, I don't understand why you are bothering with LAB? Stay in rgb and you can work with HSB sliders if working with removing color casts. You may also have better results with some of the photo filters. Or just set the temperature to your favority daylight setting when converting.

Also try various gray points in your image... You might be surprised how they don't work well... (so back to what pleases you)



Mar 19, 2014 at 03:38 PM
ben egbert
Online
• • • • •
Upload & Sell: On
p.1 #3 · p.1 #3 · Fixing color in LAB


Bernie wrote:
When doing landscapes, there are no correct numbers -- especially with a CP. Because of clouds, sunrise / sunset colors, leaves, clouds from forest fires, etc "correct" color can be whatever is pleasing to your eye for the circumstance that you are capturing -- and the mood you are portraying. JMHO

Also, I don't understand why you are bothering with LAB? Stay in rgb and you can work with HSB sliders if working with removing color casts. You may also have better results with some of the photo filters. Or just set the temperature to your favority daylight setting when converting.

Also
...Show more

I think my 3 stop reverse ND grad caused the shift, I did not use a CP.

I am using LAB because of the very strong recommendations from several on this board. I like it very much so far.

You did not say if the rework was good bad or in between. Its not what pleases me but the landscape forum that matters. I am happy with my prints, but when I post to a forum, it is specifically to please at least the more experienced members.



Mar 19, 2014 at 03:46 PM
howardm4
Offline
• • • •
Upload & Sell: On
p.1 #4 · p.1 #4 · Fixing color in LAB


I'm at work on an un-cal'd monitor but to be honest, I hate the rework. IMO, you'd be better off modifying the A channel and moving it slightly toward to magenta to counter the green and use a mask. From what I see, the un-reworked version has much more drama of impending weather whereas the rework screams overworked color in the sky.


Mar 19, 2014 at 04:11 PM
ben egbert
Online
• • • • •
Upload & Sell: On
p.1 #5 · p.1 #5 · Fixing color in LAB


howardm4 wrote:
I'm at work on an un-cal'd monitor but to be honest, I hate the rework. IMO, you'd be better off modifying the A channel and moving it slightly toward to magenta to counter the green and use a mask. From what I see, the un-reworked version has much more drama of impending weather whereas the rework screams overworked color in the sky.


Thanks a color judgement is what I was looking for. I am using a mask however.

Here is one with the A shifted .5 small square to the right (magenta). More was not going to work, even to my poor eyes.

I personally like my first fix bette but perhaps half as much added blue, or maybe desaturate the blue a bit.








Mar 19, 2014 at 04:14 PM
ben egbert
Online
• • • • •
Upload & Sell: On
p.1 #6 · p.1 #6 · Fixing color in LAB


By the way, here is a link for the RAW

https://dl.dropboxusercontent.com/u/77536572/140307-6169-5dm3.CR2



Mar 19, 2014 at 04:53 PM
Eyeball
Offline
• • • •
Upload & Sell: On
p.1 #7 · p.1 #7 · Fixing color in LAB


The first thing I would question is whether the assertion that there is a green cast is correct.

On my calibrated monitor, the only place I can notice a green cast by eye is the neutral/beige outcropping at the bottom just left of center. Even there it is very subtle and hard to be sure if it is "real" or just a psychological effect of being surrounded by so much blue, orange, and magenta.

The blue cast seems much more pronounced to me and I think that takes us into subjective territory as to whether the removal of atmospheric haze and blue sky reflection is desired or not.

I'll try to take a look at the raw when I get a chance.



Mar 19, 2014 at 05:50 PM
ben egbert
Online
• • • • •
Upload & Sell: On
p.1 #8 · p.1 #8 · Fixing color in LAB


Eyeball wrote:
The first thing I would question is whether the assertion that there is a green cast is correct.

On my calibrated monitor, the only place I can notice a green cast by eye is the neutral/beige outcropping at the bottom just left of center. Even there it is very subtle and hard to be sure if it is "real" or just a psychological effect of being surrounded by so much blue, orange, and magenta.

The blue cast seems much more pronounced to me and I think that takes us into subjective territory as to whether the removal of atmospheric haze and blue
...Show more

Looking forward to it. I took another stab at this and made about half the blue correction and also darkened the blue a bit then desaturated it. I won't bother showing it until I see a few more comments and or other peoples take.

I am not sure there is a cast in the RAW, but there is an intense blue brought about by the filter. As I recall, I desaturated the blue and maybe went too far and that may be whats causing the complaints.

When a few top forum members say something has a cast, there is no recovery from that.




Mar 19, 2014 at 06:38 PM
Picture This!
Offline
• • •
Upload & Sell: On
p.1 #9 · p.1 #9 · Fixing color in LAB


Color balancing is to a certain extent a subjective art. Having said that, I see more of a magenta and blue cast in the image than green. Here's my attempt at color correcting the image. This was done in RGB mode with thresholds for whites and blacks and a grey fill later set to difference with a threshold to get 18% gray for midtone correction. Resized, sharpened with luminosity masks and that's it.





Edited on Mar 19, 2014 at 07:59 PM · View previous versions



Mar 19, 2014 at 06:44 PM
howardm4
Offline
• • • •
Upload & Sell: On
p.1 #10 · p.1 #10 · Fixing color in LAB


Ben, just curious; what is the exposure and focal length used?


Mar 19, 2014 at 07:05 PM
 

Search in Used Dept. 



Peter Figen
Offline
• • • •
Upload & Sell: On
p.1 #11 · p.1 #11 · Fixing color in LAB


Here's a quick five minute crack at it. No Lab corrections though. This image seemed not to need that.




  Canon EOS 5D Mark III    17mm    f/8.0    1/10s    100 ISO    +0.3 EV  




Mar 19, 2014 at 07:33 PM
ben egbert
Online
• • • • •
Upload & Sell: On
p.1 #12 · p.1 #12 · Fixing color in LAB


Thanks Picture Perfect. I appreciate your work for sure. And while color cast is subjective, my mild cataracts adds a yellow filter and means I need to find some other way to get it right. I have my camera profile via Color Checker and usually trust that and I try to make all my post processing color neutral. But if the raw has an issue I am cooked.

I will need to take your word for it on the magenta blue cast as I did on the previous green cast. I could certainly live with your rendition, but I would go for a lot more contrast. Contrast I can see.

This image is interesting, I used a 17TSE perfectly leveled and then upshifted for composition. Some complained about that white ledge and suggesting pointing the lens up to avoid it. But I can spot distortion as easy as others spot color cast. Even corrected for horizon, I would see the side walls of the canyon as distorted.




Edited on Mar 19, 2014 at 08:13 PM · View previous versions



Mar 19, 2014 at 07:33 PM
ben egbert
Online
• • • • •
Upload & Sell: On
p.1 #13 · p.1 #13 · Fixing color in LAB


Peter Figen wrote:
Here's a quick five minute crack at it. No Lab corrections though. This image seemed not to need that.


Peter, in my motel room I was using ACR and letting it auto correct and was getting similar results. Sometimes we overthink things.

I love yours, but my quick methods did not return the clarity or color yours has. If I added Topaz pop, it gets close, but Topaz frequently adds a cast so I typically change it to luminosity and use an HSL/HLB to get back some saturation.

You are correct, it hardly needs LAB, but I have started using LAB to apply shadow highlight in the L channel (per the book) and also use a sharpening action in Lab L channel at the end.

For this image, I only bumped the A and B slightly in straight lines after the shadow highlight.

I can process this image in 5 minutes, but at the same time I can do 25 different 5 minute process with each slightly different. And of course after I read a new chapter, I have new stuff to try.

I would love to hear your formula, just in broad outline. I am especially interested in how you got the clarity and saturation.



Mar 19, 2014 at 07:40 PM
ben egbert
Online
• • • • •
Upload & Sell: On
p.1 #14 · p.1 #14 · Fixing color in LAB


howardm4 wrote:
Ben, just curious; what is the exposure and focal length used?


This was at f8, ISO100 and 1/10 second. I used a Singh Ray 3 stop reverse grad above the canyon horizon.



Mar 19, 2014 at 08:14 PM
Peter Figen
Offline
• • • •
Upload & Sell: On
p.1 #15 · p.1 #15 · Fixing color in LAB


Now I have to think... First move was an overall curve to set approximate black and white points. Then I think I did another curve for contrast on the bottom half. That also added some saturation. Then I probably used a bit of Hue/Saturation in the sky. Then Merged Visible layers to a new layer on top, duped that layer and added HiRadiusLowAmount USM selectively to add pop to the image. That also added saturation. Added another Hue/Sat layer and slightly desaturated the reds in the canyon, as they were getting overboard. A final very light USM for web display = 75, .3, 0, but History brushed in so as not to include the horizon, which would have had too much of a halo. That's more or less it.


Mar 19, 2014 at 08:41 PM
ben egbert
Online
• • • • •
Upload & Sell: On
p.1 #16 · p.1 #16 · Fixing color in LAB


Peter Figen wrote:
Now I have to think... First move was an overall curve to set approximate black and white points. Then I think I did another curve for contrast on the bottom half. That also added some saturation. Then I probably used a bit of Hue/Saturation in the sky. Then Merged Visible layers to a new layer on top, duped that layer and added HiRadiusLowAmount USM selectively to add pop to the image. That also added saturation. Added another Hue/Sat layer and slightly desaturated the reds in the canyon, as they were getting overboard. A final very light USM for web
...Show more

Thanks Peter, I also noted it was easy to overdo red in the canyon, especially the left side. You must be getting most of the contrast with your curves work.

By the way, to my eyes, I think your colors are closer to my rework (second image) than to any of the others here.

I think I am stalling out at Chapter 9 of the book. He has too many unexplained Photoshop steps mentioned that are not in my repertoire. For example loading the B channel into a curves mask, or using the blend if technique. I can do what he shows in that respect, but when I start splitting the sliders, the effect is null. That is the layer I want to blend looses its effect.

I had to look all over the place to figure out how to invert a layer for example. Also I see his version of shadow highlights is missing the later additions. But still, this has been a very big leap in my post processing skills. The trick will be to not overuse them.



Mar 19, 2014 at 09:14 PM
ben egbert
Online
• • • • •
Upload & Sell: On
p.1 #17 · p.1 #17 · Fixing color in LAB


Here is one I did following Peters recipe. It's clear I am not as good with curves as he is.







Mar 19, 2014 at 09:58 PM
Peter Figen
Offline
• • • •
Upload & Sell: On
p.1 #18 · p.1 #18 · Fixing color in LAB


"Also I see his version of shadow highlights is missing the later additions"

I think that tool has pretty much remained the same from its inception.

As great a book as that is, I also have found that I would digest it for a few weeks then put it aside for six months to a year and then come back to it. That process worked, and is working well for me. I've probably forgotten a bunch of steps and techniques by now and it's time to go back once again and see what else I can absorb. Don't worry about trying to get it all at once. Just get what you get and have fun with your images.



Mar 19, 2014 at 10:09 PM
Bernie
Offline
• • • •
Upload & Sell: Off
p.1 #19 · p.1 #19 · Fixing color in LAB


One way to get rid of a color cast...








Mar 19, 2014 at 10:37 PM
ben egbert
Online
• • • • •
Upload & Sell: On
p.1 #20 · p.1 #20 · Fixing color in LAB


Peter Figen wrote:
"Also I see his version of shadow highlights is missing the later additions"

I think that tool has pretty much remained the same from its inception.

As great a book as that is, I also have found that I would digest it for a few weeks then put it aside for six months to a year and then come back to it. That process worked, and is working well for me. I've probably forgotten a bunch of steps and techniques by now and it's time to go back once again and see what else I can absorb. Don't worry about
...Show more

Good advice. I am pretty sure I will keep doing shadow highlight in lab and also sharpening there as well.

The shadow highlight he shows does not include the color correction and midtone contrast area. I see, I am using it with show more options checked.

The next chapter is on changing colors green Corvette to red or such. I have never wanted to do this, but it might have other clues. My real problem with Photoshop tutorials and books is that the author often assumes I have the entire menu memorized which is about like memorizing the IRS code.




Mar 19, 2014 at 11:47 PM
1
       2       3       end




FM Forums | Post-processing & Printing | Join Upload & Sell

1
       2       3       end
    
 

You are not logged in. Login or Register

Username   Password    Retrive password