Sun Moon Lake

[mikemeister_admin]

Did you say "easy in Lab?"  Here are my RGB and Lab versions.  The Lab version has what would appear to be a yellow cast, but correcting the yellow only seemed to create a corresponding increase in a blue cast.  This was my best effort but, I don't think I was able to eliminate it entirely.  The RGB version . . . well, what can I say?

[ggroess]

I can tell by your comments that you are not satisfied with this image.....

I'm here to tell you that is "ok"  your image(s) look just fine. 

The goals here are to increase the contrast and recover some of the color.  In RGB with Color and Lightness tied together you will always have more trouble getting this image figured out. 

Your lab image has better detail in the bright highlights and overall less color shifting.

One thing to note...
This is one image where your interpretation is as good as mine.  This image could be sub-titled "Subjective" because of the wide range of ways you can adjust it.  Myself I prefer it to be dark and moody, others have made it light and bright.  either one is good as long as you have full range in the image and you bring out some color. 

As a challenge see if you can bring out the temple on the top right...it's in there and it's pretty tough.

Greg

[mikemeister_admin]

Hi Greg:

I did spend quite a bit of time on these images (honest) and spent some time working on that temple trying to pull it out of the dark, but the more I tried to draw it out, the more I seemed to lose other parts of the image.  Can you show me what your curve would look like?

Also, how about the yellow mist?  To me it looks unnatural, although I had a hue clock posted up in the clouds and kept them white (which I think means the hue clock just registers a centered dot). How do I get rid of the yellow mist in Lab without creating a blue color cast?

Brian

[ggroess]

Oh I have no doubt you have spent some time on this...It is very subjective and a tough image to be satisfied about.

The Hue clocks show you what is going on under the marker.  So sometimes you need to set multiple hue clocks.  The sample for the hue clock is set in the user settings dialog under the little wrench Icon.  Default is 5X5 sample size.  You can change it as needed to expand the range or narrow it to meet your needs.

In LAB the biggest trouble I have is seeing the color "area" so many times I'll open the single channel in CM and make the curve window more than 1/2 of my screen.  Then I place the mouse over the part of the image I am having toruble with and right click.  I choose "mark"  This places a marker on the curve line and that is where I want to center my adjustment.

The trouble with this image is the fog is layered by brightness...so as you adjust you bring more or less of the fog out...That is why this is a great LAB image.  Color is not linked to brightness...

My B curve for this image is always changing but here is an example....
I have also included the ACV file so you can load my curve and look at the B curve.  Drag the window open so it is quite large.  The adjustment is very close to the center of the curve.  I create a "flat spot" in the curve which makes the yellow more of a neutral.  In the LAB space horizontal parts on the curves create neutral spots in the color range.  They can be very bad...or in this case very helpful.

Greg

[mikemeister_admin]

Thanks.  I will try that.

Also interesting where you placed your neutral.  I wouldn't think of putting it there.  Is that just the result of moving it around and judging what it did to the colors?  As I recall I tried fiddling with a neutral for a while but couldn't find a satisfactory spot.

Also, as I recall, in my trial and error, I placed a shadow point on the boat and a highlight point above the clouds in the corner, thinking that might keep them from blowing out.  Good idea or bad idea?  it doesn't look like you used highlight or shadow points.  What was your thinking on that?

[ggroess]

Because I was not sure where the "Significant Shadow" was I grabbed the shadow edge of the curve and pulled it in until the little black specks started showing up.  For the highlights the far left is blown out so I had no room to move there.  My solution was a steep lizard tail.  I wanted to bring back some highlight details in the clouds.  Both settings are a compromise I acknowledge, but they get the job done...when you are not sure about the shadows/highlights.. you can threshold using the method above. 

I would say the boat should be above the shadow line so you get all of the detail and color that might be there.  I went for a few specks in the boat but certainly not the entire boat.  As Greg M pointed out there is red in the net...

The neutral was shopped around for kind of like the Honest Abe video.  I did know that the yellow color in the mist could be really there and I wanted to find a part of the image where it would be naturally reduced. It was not completely reduced out where it is but it is much better. 

The sharp bend in the B curve kills the yellow in the larger parts of the mist.  Because the color lives in a very narrow area of the B curve I had to enlarge the curve window to be able to see the area for adjustment.  It is actually 1 thing that the histogram does help you figure out.  If you look at the B curve histogram it is just a small spike.  There is little room on the curve for adjustment.  So you have to enlarge the curve to see into the adjustment area.

Great questions...
Greg