Home · Register · Join Upload & Sell

Moderated by: Fred Miranda
Username  

  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
  

Archive 2014 · Fixing color in LAB

  
 
ben egbert
Offline
• • • • • •
Upload & Sell: Off
p.2 #1 · p.2 #1 · Fixing color in LAB


Bernie wrote:
One way to get rid of a color cast...

https://www.fredmiranda.com/hosting-data//5862/4945140307-6169-5dm3-w.jpg


I knew somebody would do this, thanks.



Mar 19, 2014 at 06:48 PM
ben egbert
Offline
• • • • • •
Upload & Sell: Off
p.2 #2 · p.2 #2 · Fixing color in LAB


Ok, after this, I still need to know if I have a cast or how to get rid of it. My first was off, my second was too blue.

Peters version and my attempt to copy it seem to be very close on color if not contrast and clarity.

I liked my original for everything but sky color so I took another swing at fixing just the sky. I used lab B and pulled it blue but only half as far as the other lab corrected version. I also used the L channel to darken the sky a bit.

This is the result and I want to know if the sky is within a believable range. If there is anything else about this version that you don't like, please let me know, but I think I could like a print of this.









Mar 19, 2014 at 09:54 PM
Peter Figen
Offline
• • • • •
Upload & Sell: On
p.2 #3 · p.2 #3 · Fixing color in LAB


Ben - Absent something that's obviously neutral in an image, it's really taking a quick look at the image on a well calibrated screen and asking yourself what your first impression of the image is. If it has a cast it will feel too green or too yellow, etc. You can't do this after laboring over the image for a long time. It needs to be a fresh peek. At first glance, most of the previous versions posted in the thread felt greenish to me in the tops of the secondary mesas to the left of the center of the image. My version felt a bit more brown than green, and for ME on this image, that felt better to my eye. But it's your image and not mine. That was just my interpretation, and mine is often as much about gut feel as anything else. This latest version, to my eye, feels cool overall. What happens if your very slightly warm the whole thing up. It would counter the somewhat cyanish upper left sky, which, in this type of image, might feel better with a hint more magenta in it. But then again...


Mar 19, 2014 at 11:34 PM
ben egbert
Offline
• • • • • •
Upload & Sell: Off
p.2 #4 · p.2 #4 · Fixing color in LAB


Peter Figen wrote:
Ben - Absent something that's obviously neutral in an image, it's really taking a quick look at the image on a well calibrated screen and asking yourself what your first impression of the image is. If it has a cast it will feel too green or too yellow, etc. You can't do this after laboring over the image for a long time. It needs to be a fresh peek. At first glance, most of the previous versions posted in the thread felt greenish to me in the tops of the secondary mesas to the left of the center of
...Show more

I don't know how you see these casts. Anyway, here is one with the 81 warming filter applied. 85 seemed to much, and 81 seemed ok without a fade.









Mar 20, 2014 at 09:01 AM
ben egbert
Offline
• • • • • •
Upload & Sell: Off
p.2 #5 · p.2 #5 · Fixing color in LAB


Peter, you said something about the absence of neutrals but this image has a big white slab of rock. I tried another go around by using the whitest patch of this rock to set WB. Then I made this version using the auto choices in ACR as a starting point. Auto makes very small changes to the sliders other than adding about 40 to whites.

I used shadow highlight on the L channel to get some pop, added some saturation, adjusted black point and thats it. No color correction required on this one. Using the color checker, the sky is in solidly blue territory.








Mar 20, 2014 at 10:15 AM
Eyeball
Offline
• • • •
Upload & Sell: Off
p.2 #6 · p.2 #6 · Fixing color in LAB


Ben - One technique that I use once in a while where casts are concerned is to crank the saturation all the way up and see what it tells you. You can do it with one of the saturation sliders in RGB or use more extreme diagonals in your A and B curves in LAB. They produce slightly different results.

On a related note - if I remember correctly Margulis has a section in his book where he talks about making excessive adjustments to the LAB curves and then adjusting back down with opacity. I think it is one of those things that can help you see and understand the interplay of color.



Mar 20, 2014 at 10:17 AM
ben egbert
Offline
• • • • • •
Upload & Sell: Off
p.2 #7 · p.2 #7 · Fixing color in LAB


Eyeball wrote:
Ben - One technique that I use once in a while where casts are concerned is to crank the saturation all the way up and see what it tells you. You can do it with one of the saturation sliders in RGB or use more extreme diagonals in your A and B curves in LAB. They produce slightly different results.

On a related note - if I remember correctly Margulis has a section in his book where he talks about making excessive adjustments to the LAB curves and then adjusting back down with opacity. I think it is one of those
...Show more

Thanks finding color cast is my big challenge and any tips to spot them is good news for me. I wilt try these.



Mar 20, 2014 at 10:20 AM
Peter Figen
Offline
• • • • •
Upload & Sell: On
p.2 #8 · p.2 #8 · Fixing color in LAB


Ben - I think you need to be careful about the white rock. It's not really white and you probably don't really want to make it neutral. There's a lot of color in it. Actually warm AND cool at the same time. Now, there is a tree just to the right that does look fairly neutral, but again, do you want it to actually be neutral or, because it's in the shadows, should it have a cast? You could go either warm or cool with it and be right. It just depends on the mood you're going after.

A lot of what I look for are color harmonies and color contrasts, and whether they enhance the viewing experience or detract from it. That's why those areas that appeared every so slightly green in your original version bothered me. They contrasted with the red rocks rather than harmonized with them. And in THIS case, that didn't work for me. It's also why a seemingly insignificant move toward a warmer color made the image feel more homogenous. In an urban landscape, you might have wanted the color contrast. Your latest version is much better overall.

Off to jury duty...




Mar 20, 2014 at 10:53 AM
ben egbert
Offline
• • • • • •
Upload & Sell: Off
p.2 #9 · p.2 #9 · Fixing color in LAB


For Eyeball.

Looking at the raw as shot versus the one with the WB changed and then running hue saturation at 100%. This is the differences I note.

The as shot has a green cast just above the horizon. This is just a thin line above another yellow line that seems natural. The green does not seem natural. There is not much green anywhere else other than trees. The white rock is mostly blue with patches of yellow. The upper left sky is dark blue.
The WB corrected shot has no green cast in the sky, the upper left is solid blue. The rock is mostly light yellow with some blue patches and some red.

Using the steep LAB curves, I see yellow green above the horizon on the as shot RAW. The WB corrected image has very strong saturation but all areas seem the correct color just highly over saturated. This seems like a powerful check.


Edited on Mar 20, 2014 at 12:19 PM · View previous versions



Mar 20, 2014 at 10:58 AM
ben egbert
Offline
• • • • • •
Upload & Sell: Off
p.2 #10 · p.2 #10 · Fixing color in LAB


Peter Figen wrote:
Ben - I think you need to be careful about the white rock. It's not really white and you probably don't really want to make it neutral. There's a lot of color in it. Actually warm AND cool at the same time. Now, there is a tree just to the right that does look fairly neutral, but again, do you want it to actually be neutral or, because it's in the shadows, should it have a cast? You could go either warm or cool with it and be right. It just depends on the mood you're going after.

A
...Show more


I zoomed in 100% and found a very white patch some is yellowish but I avoided that and also any that were not well lit.

You have not said if any of my last two images are better. Ok I see it now, thanks

I know what I like with my yellow filter eyes and thats the WB version, but then I liked my first version until I was told otherwise. I don't know what I don't know.

Good luck on JD








Mar 20, 2014 at 11:02 AM
Bernie
Offline
• • • •
Upload & Sell: Off
p.2 #11 · p.2 #11 · Fixing color in LAB


ACR conversion (edit: punched up a bit)


https://www.fredmiranda.com/hosting-data//5862/4945140307-6169-5dm3-w_1-med.jpg

Edited on Mar 22, 2014 at 09:41 AM · View previous versions



Mar 20, 2014 at 05:32 PM
ben egbert
Offline
• • • • • •
Upload & Sell: Off
p.2 #12 · p.2 #12 · Fixing color in LAB


Looks pretty good, what settings did you use? Any saturation, clarity and vibrance? The image responds pretty well to ACR only until you try to punch up a bit.




Mar 20, 2014 at 06:42 PM
Peter Figen
Offline
• • • • •
Upload & Sell: On
p.2 #13 · p.2 #13 · Fixing color in LAB


A lot depends on what you do both in the raw processor and in Ps after. This latest "ACR" version, whatever that means, seems flat and lifeless to me. It might be a good starting point, but it needs a lot of massage to make it shine.


Mar 20, 2014 at 07:30 PM
Ho1972
Offline
• • • •
Upload & Sell: Off
p.2 #14 · p.2 #14 · Fixing color in LAB


Ben, I used this method to find neutral gray in your image. This is the result with a slight tweak or two. Note that this trick works better on some images than others. I refrained from making it punchy.







Mar 20, 2014 at 07:36 PM
ben egbert
Offline
• • • • • •
Upload & Sell: Off
p.2 #15 · p.2 #15 · Fixing color in LAB


Thanks, I will give it a try. This seems pretty red.

Ok, I tried your method on my last image. The clouds just above the horizon and right of center were the 50% grey spot. I would have to live with this a while to decide if I want it this warm.

I will give it another go with a raw tomorrow.







Mar 20, 2014 at 09:22 PM
ben egbert
Offline
• • • • • •
Upload & Sell: Off
p.2 #16 · p.2 #16 · Fixing color in LAB


Ok here is one using the 50% grey started from raw. I included my tricks for getting some snap into the canyon.

I found 4 places in the clouds that all show up at the same time and each gives a different result. A couple were rose colored so I avoided them and stayed with areas that looked grey.

When I finished, I reset it to grey as the processing had added some cast. Here is the result.







Mar 20, 2014 at 09:57 PM
Ho1972
Offline
• • • •
Upload & Sell: Off
p.2 #17 · p.2 #17 · Fixing color in LAB


I'm a bit surprised our images are so different. Maybe the fact that I processed the raw in Photo Ninja and we each added our own tweaks helps explain it. At any rate, maybe you'll find this technique useful going forward.


Mar 21, 2014 at 03:44 AM
ben egbert
Offline
• • • • • •
Upload & Sell: Off
p.2 #18 · p.2 #18 · Fixing color in LAB


Ho1972 wrote:
I'm a bit surprised our images are so different. Maybe the fact that I processed the raw in Photo Ninja and we each added our own tweaks helps explain it. At any rate, maybe you'll find this technique useful going forward.


I think I have seen this trick before but got talked out of using it. The problem I found is that it is very hard to add snap and drama and avoid changing the cast. Maybe that the beauty of LAB, but I need more snap than I can get just with the A&B channel, and other methods add saturation. Even doing Shadow Highlights in LAB on L changes the saturation.

In the end, this is a fools errand to attempt to please everyone on color cast.

I am the sort of guy who will discard dozens of praise and listen to the one critic. I work hard to make my image free of technical flaws so that the image stands on its intrinsic merits. But after years of fighting focus and sharpness and exposure and post processing, I am still not there.

I want my images to work or not work based on the scene and composition and light.

In the end, this is a failed image. I have printed the first one with the supposed color cast and the later one I did that has no cast. I have a hard time telling them apart in print, and so does my wife.

Of all the versions in this thread, I still measure them against the first one and while I now believe it has a cast, the drama and clarity make it the one to beat. The one that does for me is the WB corrected version. In fact I went back and added some blue just above the horizon.

But I don't think I will print this. I have come to dislike the image itself. I don't like the white rock, and I don't like the crop to eliminate it, and I don't like that it needs so much PP. My favorite prints were ones that were good right out of camera.






Mar 21, 2014 at 09:57 AM
ben egbert
Offline
• • • • • •
Upload & Sell: Off
p.2 #19 · p.2 #19 · Fixing color in LAB


I printed the following at 8.5x11 and when side by side, they are much closer in appearance than here on the web.

The left is one corrected for WB on the white rock. The right side uses the cloud to set 50% grey in RGB. I also tried setting it in LAB and got nearly identical results.

The point here is that the 50% grey version is much warmer on the web than in print. I am using Canson photorag with a Canson profile for printing.







Mar 21, 2014 at 01:25 PM
John Wheeler
Offline
• •
Upload & Sell: Off
p.2 #20 · p.2 #20 · Fixing color in LAB


Hi Ben
Adding a comment late to the party yet here are some minor thoughts to consider as I did not see them mentioned and you may already have it covered. Staying in 16 bit mode vs 8 bit mode has two advantages especially when jumping between color spaces or into LAB mode and back
- There is a subtle posterization that goes on as 8 bit has insufficient precision to prevent it
- Each time shifting back from LAB to RGB in 8 bit mode Photoshop throughs in a little bit of 1 bit noise dithering (to hide the banding) unless the Use Dither option is unchecked under the Conversion Options section of Edit > Color settings > More Options command.

These items are subtle and would not be noticed in the "land" portion of a landscape image yet has the potential to introduce some banding if the sky is a slowly varying gradient.

Multiple round trips to LAB space can cause this noise dithering to be cumulative with each round trip. Not a huge issue yet thought it was worth mentioning.



Mar 22, 2014 at 12:42 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       Or Reset password



This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.