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
  

Is this believable?
  
 
newhaven
Offline
• •
Upload & Sell: Off
p.2 #1 · Is this believable?


This tutorial shows how to color balance using color sampler points and individual color curves. I prefer to move the curve points with the up/down arrows. Also, to transfer the color sample targets to points on the curves - with the eyedropper tool active hovering over a color sample target, shift-ctrl-click. If you want to affect color without changing luminosity, change the adjustment layer blend mode to color. The color balance adjustment layer is better for fine tuning. Do not use point sample. Try to use 5x5 or larger.

http://projectwoman.com/2011/09/color-correction-in-photoshop.html



Sep 15, 2013 at 10:36 PM
ben egbert
Offline
• • • • •
Upload & Sell: On
p.2 #2 · Is this believable?


RustyBug wrote:
Mostly, I separate my pp into the different portions that the light is giving me ... dialing each to taste independently and then blending/balancing by eye.

Even though we have one image, I treat it as two different images because the different portions of the scene are really being lit by two very different sets of lighting. With the exposure DR variance between the portions of the scene, it is really impossible to make global adjustments.

If you have an incident exposure meter, you can take a couple of readings at say 3 PM to illustrate.

First reading:
Stand facing the sun and
...Show more

The entire process of adding dr is to accommodate this variance in exposure. I use less point by point methods than you. The steps I use are.

1. Use an nd grad.

2. Bracket.

3. In the recent case I use ACR to auto boost shadow and reduce highlights. In other cases I blend the brackets.

But what I hope I never ever do is make several versions of layers and then blend them point by point. That sounds like brush work or some even worse PS tricks.

Maybe I have it wrong. Of course I can and have worked my individual images before blending, but the blend needs to be no more complicated than a grad layer or my gaussian blend.






Sep 16, 2013 at 12:31 AM
ben egbert
Offline
• • • • •
Upload & Sell: On
p.2 #3 · Is this believable?


newhaven wrote:
This tutorial shows how to color balance using color sampler points and individual color curves. I prefer to move the curve points with the up/down arrows. Also, to transfer the color sample targets to points on the curves - with the eyedropper tool active hovering over a color sample target, shift-ctrl-click. If you want to affect color without changing luminosity, change the adjustment layer blend mode to color. The color balance adjustment layer is better for fine tuning. Do not use point sample. Try to use 5x5 or larger.

http://projectwoman.com/2011/09/color-correction-in-photoshop.html



This slipped past me as often happens in a long post, great info here and as soon as I have some time I am going to try it out. Packing for a trip this morning, but I appreciate this.



Sep 16, 2013 at 02:58 PM
RustyBug
Online
• • • • • •
Upload & Sell: On
p.2 #4 · Is this believable?


ben egbert wrote:
This slipped past me as often happens in a long post, great info here and as soon as I have some time I am going to try it out. Packing for a trip this morning, but I appreciate this.


You'll find this the same premise @ using color balance layers. It can be done with curves or color balance or levels ... they really are still all doing the same thing @ equalizing RGB channels. They are just different ways of doing the same thing with controls of your choosing/liking. The key though is selecting points that SHOULD be neutral or that your WANT to be neutral, relative to your lighting.

The mechanics are essentially the same either way, it is when you get into mixed lighting that you'll want to be more vigilant/selective in you application as different parts of your scene are illuminated by different colors of light. I used to just use levels and slide each channel endpoint independently, but now use color balance more often than not. Curves is just as viable ... and actually can garner you additional control to more selective/refined areas by placing sufficient points on the curve to hold areas you want UNCHANGED at neutral.

You'll notice the tutorial reference to Dan Margulis (again ... you really should pick up his book). and the reference to a 2-3 point spread (I strive for <5). Anyway, more than one way to skin this cat, but the crux of it is that neutral(s) simply equate to closely approximated RGB values. Where you choose to put them and how you choose to get them there ... that's your call.

Enjoy your trip. Looking forward to seeing what you bring back.



Sep 16, 2013 at 03:20 PM
ben egbert
Offline
• • • • •
Upload & Sell: On
p.2 #5 · Is this believable?


RustyBug wrote:
You'll find this the same premise @ using color balance layers. It can be done with curves or color balance or levels ... they really are still all doing the same thing @ equalizing RGB channels. They are just different ways of doing the same thing with controls of your choosing/liking. The key though is selecting points that SHOULD be neutral or that your WANT to be neutral, relative to your lighting.

The mechanics are essentially the same either way, it is when you get into mixed lighting that you'll want to be more vigilant/selective in you application as different
...Show more


Here is a try at the tutoral. Changing to 5 points made a world of difference, but I was not able to select these points in curves using Ctrl Click. I could select them with the eyedropper tool in the curves window, but it is not exact. I ended up with a cast even I could see,

So I used the points which I think display here. They were all within 5 points of neutral. I used color balance on shadow, highlight and mid and got them within 1 or so points spread. Here is the result with no other work done to the image. To my eyes, this looks pretty much like all the other versions I have done. At least in terms of color.






Per the tutorial




Sep 16, 2013 at 04:27 PM
newhaven
Offline
• •
Upload & Sell: Off
p.2 #6 · Is this believable?


Shift-ctrl-click with the eye dropper tool active.


Sep 16, 2013 at 05:37 PM
newhaven
Offline
• •
Upload & Sell: Off
p.2 #7 · Is this believable?


Why don't you put the .CR2 file on the web somewhere like yousendit or dropbox? The image you just posted has a strong blue cast.


Sep 16, 2013 at 05:53 PM
ben egbert
Offline
• • • • •
Upload & Sell: On
p.2 #8 · Is this believable?


newhaven wrote:
Shift-ctrl-click with the eye dropper tool active.



Thanks, the tut was cs3, I use cs6. I redid this using only 3 neutral points and got pretty close to the first version. Before learning about shift-ctl-clk.

By the way, I can find all the points with spreads under 5, and according to the tutorial, this indicates no color cast in the original raw.



Sep 16, 2013 at 05:57 PM
ben egbert
Offline
• • • • •
Upload & Sell: On
p.2 #9 · Is this believable?


newhaven wrote:
Why don't you put the .CR2 file on the web somewhere like yousendit or dropbox? The image you just posted has a strong blue cast.



I could do that, maybe when I get home, I am getting pretty close to departure time.



Sep 16, 2013 at 05:58 PM
newhaven
Offline
• •
Upload & Sell: Off
p.2 #10 · Is this believable?


The last image you posted has a clipped blue channel. I will go through the whole process starting from the beginning with a raw file.


Sep 16, 2013 at 06:17 PM
 

Search in Used Dept. 



ben egbert
Offline
• • • • •
Upload & Sell: On
p.2 #11 · Is this believable?


I attempted a link to dropbox, Doubt it works, I usually need half a day to figure it out.

Edit for another attempt.

https://dl.dropboxusercontent.com/u/77536572/130725-3877-5dm3.CR2



Sep 16, 2013 at 06:25 PM
newhaven
Offline
• •
Upload & Sell: Off
p.2 #12 · Is this believable?


That one looks fine. Where's the image with the blue cast - the last image you posted?


Sep 16, 2013 at 07:02 PM
RustyBug
Online
• • • • • •
Upload & Sell: On
p.2 #13 · Is this believable?


Here is where I'm looking at for my assessment to "neutralize" my WB.

The faces of the rock (assumed neutral) are orientate in three different directions. The light that is illuminating the scene is coming from overhead. The left and right faces of the rock will NOT be reflecting the same lighting that is illumination the scene.

This was just a "one click" using the gray dropper in levels. The key aspect isn't whether or not we use levels, curves, or color layers, but neutrals relative to our lighting source orientation. The specular spot on the top face of the rock is going to be reflective (AI=AR) of the light that is most illuminating the foreground as we really have little sidelighting influence in that portion of the scene. The bulk of illumination is coming from overhead.

So, if I am assuming it to be a neutral component of the scene (which is a guess), the if I balance it for the light it is most receiving, then the other areas that are receiving that same light will be balanced also. As you can see from the rock (pre-adjustment), depending on which portion of it you would select as your "neutral" basis, you would get different corresponding outcomes. Of course, if this were a "colored" rock, our assumption to use it as a neutral would be skewed. But, even if it is 100% neutral, it's three dimensional reflection of different portions of the sky will give us different values for the light falling on the scene.

















Sep 16, 2013 at 07:23 PM
skibum5
Offline
• • • • • •
Upload & Sell: Off
p.2 #14 · Is this believable?


micro-contrast on the orange mountain seems a bit flat and something seems a bit off with it



Sep 16, 2013 at 08:46 PM
skibum5
Offline
• • • • • •
Upload & Sell: Off
p.2 #15 · Is this believable?


ben egbert wrote:
Here is one taking some of your comments.

I followed the recipe (two images using ACR auto for exposure and WB), but I turned off saturation.

In CS, I made a black point by sliding each channel back to clipping. Then I added a linear curve and sharpened. That is all. No color balance, not added saturation.

It might still be too dark, and the way I brought down the curve was not done in luminance mode so it might have skewed the colors a bit.

This lacks punch and is too flat for my taste, it needs more contrast. But it was
...Show more

I like this one better, maybe you could add a little local contrast on the mtn to give it a bit more punch



Sep 16, 2013 at 08:49 PM
Shutterbug2006
Offline
• • •
Upload & Sell: Off
p.2 #16 · Is this believable?


curious80 wrote:
I think most of it looks believable. However the top part of the hill seems a bit overdone to me.


I think it might look better if the top part of the mountain was about what you see in the reflection now.

But that's just my opinion.

The first thing to come to mind is 'it's been photo-shopped" when I see saturated colors.

Nice shot Ben. As for printing that shot, I don't think you'd have to lighten it much, depends on the printer and paper used really.







Sep 17, 2013 at 04:56 AM
Shutterbug2006
Offline
• • •
Upload & Sell: Off
p.2 #17 · Is this believable?


Ben, have you ever tried "Organic Imaging"?

http://www.organicimaging.com/home

Download their software. You can process 250 images for free, after that it's 10 cents an image.

Use one of the 250 freebies to see what their software would do to your image. It's pretty amazing, and actually does a pretty good job, though it might not be the look you are shooting for.

If you do check it out, I'd like to see what their software does to that image.






Sep 17, 2013 at 05:37 AM
Camperjim
Offline
• • •
Upload & Sell: On
p.2 #18 · Is this believable?


I re-read this and the other believability thread and I think I have come to some conclusions.

1. The idea of some sort of automated or software solution to color balance issues seems absurd. As I stated before using a gray card and custom white balance makes no sense when the lighting is unusual. A gray rock will not appear gray when illuminated by warm sunset or sunrise lighting. This does not work at the time of capture. Likewise you cannot use an eyedropper or other WB technique during post processing for the same reasons. I know when I take sunset pictures, I fight with my camera to achieve reasonable exposure and WB. The first thing I need to do is dial down exposure compensation by about -2 Ev. Otherwise the camera wants to turn a sunset into daylight exposure.

2. Trying to fine tune colors on an internet forum is an exercise in futility. Many of us have differences in eyesight and color acuity. In addition there can be huge differences in how images appear on our monitors. I have been playing with my Color Munki and as a result have become very aware of the differences I see at different times of the day with different ambient lighting. I understand even the browsing software we use can alter the appearance of colors.

3. When it comes to color and dynamic range, our expectations and preferences can vary a lot. We all see countless printed images during our lifetimes. The appearance of those images influence how we think an image should appear. HDR is a good example where individual expectations can vary. Look at magazine advertisements. Most of these have HDR images. The general public typically likes the impact of HDR images. We can even make the argument that HDR is closer to what our eyes see. Unfortunately I have seen too many pre-HDR images. I also know how much manipulation is needed to convert the OOC images into HDR images. For me anything with more than a trace of HDR processing appears manipulated. Maybe with time and continued bombardment with HDR images, my tastes will change.

4. Ben intentionally picked some really tough examples for a discussion of believability. If you have never seen an alpine glow, I can tell you that it does not look believable even in person. There can be an intense orange glow that only lasts a few minutes. Ben also gave us an unusual sunrise image where the lighting remained flat and instead of a warm glow, the lighting remained bluish.

5. Finally, I have no belief that any color balance matches the "truth." I do think that subtle alterations can have a big impact on our perceptions. It seems absolutely necessary to tinker with color balance in order to match what we believe we saw. It is also natural to want to add to the impact that we remember. Those enhancements can get us in trouble. A little enhancement can help retrieve the impact we remember. But then we get used to that level of enhancement and want to add more and more. When it is time to print an image all bets are off. Printer/paper profiles are supposed to help us so that what we see on the monitor is reflected in the print. That may or may not work with regard to colors but we can never expect a monitor image to be optimal when printed. There are just too many differences in the media, paper types, gloss, the presence of brighteners, metallic, etc.

Yes the more I think about this subject the more complications I find. I am convinced that this issue needs both an intellectual, technical approach and a final artistic interpretation.




Sep 17, 2013 at 02:04 PM
RustyBug
Online
• • • • • •
Upload & Sell: On
p.2 #19 · Is this believable?


Camperjim wrote:
A gray rock will not appear gray when illuminated by warm sunset or sunrise lighting.

image where the lighting remained flat and instead of a warm glow, the lighting remained bluish.

I have no belief that any color balance matches the "truth."

we can never expect a monitor image to be optimal when printed.

I am convinced that this issue needs both an intellectual, technical approach and a final artistic interpretation.

+100 @ this ^

We can adjust the color to make the gray rock neutral "as if" it had been illuminated by a balanced color of light, or we can allow the warm/cool of the light falling on it to be revealed ... it's a matter of choice.

Sunrise may be the time of day, but if the light coming directly from the specular source (i.e. sun) is blocked and the subject is being illuminated by the overhead sky ... flat/soft and blue is the nature of the light illuminating the subject. (See "choice" above at render cast or correct/neutralize.)

Natural ambient is highly varied and our eye/brain accommodation alters the "truth" into perception.

Monitor and print are two different animals ... one at additive, the other at subtractive.

Gotta study the light/subject relationship, know how to make the changes ... however it is that you want them to be.



Sep 17, 2013 at 02:59 PM
ben egbert
Offline
• • • • •
Upload & Sell: On
p.2 #20 · Is this believable?


RustyBug wrote:
Here is where I'm looking at for my assessment to "neutralize" my WB.

The faces of the rock (assumed neutral) are orientate in three different directions. The light that is illuminating the scene is coming from overhead. The left and right faces of the rock will NOT be reflecting the same lighting that is illumination the scene.

This was just a "one click" using the gray dropper in levels. The key aspect isn't whether or not we use levels, curves, or color layers, but neutrals relative to our lighting source orientation. The specular spot on the top face of the rock
...Show more

Yes, I used the rock for a neutral. I am not sure I see any real difference between this and my most recent ones, and in fact, all of them are acceptable to my eyes in terms of color. The variances between the versions in this post that catch my eye are the differences in brightness and contrast. To some extent saturation. I like more saturation than most people here, but the comments on color cast are mostly things for which I have to take your word for.



Sep 19, 2013 at 02:56 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