Building using LAB Mode

This is the discussion thread for the September 2010 Class.
mikemeister_admin
Posts: 4927
Joined: Fri Sep 20, 2013 8:29 pm

Postby mikemeister_admin » Sun Sep 19, 2010 3:52 am

Here is my attempt using LAB.  I sharpened it a bit and cloned out the flag.  I don't have the skills yet to do any better.  I think the version in Photoshop only looks better so far.  I'll be anxious to go back at the end of the class and work on it some more.  Any suggestions will be appreciated.
Attachments
building-using-lab-jpg
building-using-lab-jpg (999 Bytes) Viewed 6665 times

mikemeister_admin
Posts: 4927
Joined: Fri Sep 20, 2013 8:29 pm

Postby mikemeister_admin » Sun Sep 19, 2010 4:55 am

Layer1  Lab neutral on water tower,saturation boost,cut back sky lightness
Layer2  multiple neutrals(you can do it in RGB)
Layer3  filter>sharpen  GregM
(these were added to your image)
Attachments
ellisfllab-jpg
ellisfllab-jpg (264.58 KiB) Viewed 6665 times
ellisflrgb-jpg
ellisflrgb-jpg (257.01 KiB) Viewed 6665 times
ellisfle-jpg
ellisfle-jpg (230.49 KiB) Viewed 6665 times

mikemeister_admin
Posts: 4927
Joined: Fri Sep 20, 2013 8:29 pm

Postby mikemeister_admin » Sun Sep 19, 2010 9:47 pm

Thanks, Gremurphoto--I like your result.  I see what you did in RGB, but I don't quite understand how what you did related to what you said about "multiple neutrals."  The points on the various curves don't make sense regarding neutrals, especially the ones in the combined RGB channel, which look high for neutrals.  In the Red, Green and Blue channels you also have a point set high in each curve.  Were those points to stabilize or "tack down" the curve?  I'm surprised that such a little adjustment in the color channel curves made a noticeable difference.  If I have time, I'll try it myself.  (Hmmm--in continuing to ponder the points on the curve, do the ones in the ALL box  in the RGB mode reflect a culmination of the "tack down" points in the other curves?  If so, it does make sense.)  I hadn't thought about doing the CM in a layer--do you always do this?  How is this accomplished?  Thanks again.

ggroess
Posts: 5342
Joined: Wed May 24, 2006 2:15 am
Contact:

Postby ggroess » Sun Sep 19, 2010 10:37 pm

Rhonda,
Please remember that a neutral is different from a mid-tone. 
In RGB you can have multiple neutrals based on Brightness.  The shadow Highlight and "Mid-tone" can all be neutral. 
In the color correction "world" neutral in RGB is all three colors RG and B being the same number on the hue clock or on the numbers pallet. 

We leverage the RGB color spaces ability to have multiple neutrals to help us remove color cast from a shadow or highlight without messing up the mid-tones.  In fact in advanced work flows you start in RGB for color correction then switch to LAB for color boost.  You get the best of both worlds.

Generally we talk in terms of 1/4 tone, mid-tone, and 3/4 tone; to mean the density and tonality.  This is separate from color.  Once you separate these in your mind it gets easier to understand the curves you are working on.

CM in a layer is a great way to use the tool.  It allows you to adjust the correction if you get back to PS and find the correction is over powering the image.  To use CM on a layer just make a copy of the background layer and have it selected. Open CM and work on the the image.  Your changes will apply only to the layer and you can adjust the opacity to suit your needs.  Another great use for this is to make adjustments in a layer and set it to  "Luminosity" mode and use curves to adjust contrast and luminosity.

Points on the RGB master curve and most likely for tonal control not necessarily pinning the curves to a specific point.  They can be for tacking down the curve but in general Greg M usually does not work in that fashion...

Greg

mikemeister_admin
Posts: 4927
Joined: Fri Sep 20, 2013 8:29 pm

Postby mikemeister_admin » Mon Sep 20, 2010 2:13 am

Greg,thanks for doing the  technical explaining.The reason I used two neutrals in RGB was that I figured  on the shadow side of the building  they would be two levels of grey and would push the overall color in a certain direction.GregM

mikemeister_admin
Posts: 4927
Joined: Fri Sep 20, 2013 8:29 pm

Postby mikemeister_admin » Mon Sep 20, 2010 7:22 pm


Rhonda,
Please remember that a neutral is different from a mid-tone. 
In RGB you can have multiple neutrals based on Brightness.  The shadow Highlight and "Mid-tone" can all be neutral. 
In the color correction "world" neutral in RGB is all three colors RG and B being the same number on the hue clock or on the numbers pallet. 

Greg


Greg, this is something that should be explained at the very beginning.  When we first use the wizard in CM, we are finding the "Shadow", then the "Highlight", then the "Neutral".  When try to find the neutral, we are looking for something gray, so I took that to mean that neutral is gray.

Then as we go along, we are told that we can set a hue clock on an object....say the water tower in the above picture, and from the hue clock, we can tell that it has a slight color cast, so to remove the color cast, we need to adjust the numbers to get them neutral.  To me that meant getting them to be gray, since gray is neutral, but that's not the case. 

The term neutral needs to be defined so that we know what is being talked about.  Neutral in one context means gray, whereas in another context it can mean any color.  That's confusing.....well, to me it is.(It doesn't take a lot to confuse me.) :)

ggroess
Posts: 5342
Joined: Wed May 24, 2006 2:15 am
Contact:

Postby ggroess » Mon Sep 20, 2010 8:04 pm

I'll look at the language use in the lesson.  We can certainly make this more clear.
Ok...All Gray is neutral BUT....
Not all neutrals are Gray....

My examples are Black and White...Black is 0,0,0 and White is 255,255,255 in the RGB scale.  They would be neutral since the values are equal.  Visually everything else with equal numbers will appear a shade of gray.

In Lab the L channel can have any value...the A and B have to be 0 otherwise you have color.

The trouble lies in the sample size used by CM.  If you have CM set to 1 pixel for the sample size you would have to find a single neutral pixel and that would be it.  You might also go nuts trying to find that one pixel. 

I have my sample size set to 3X3 that samples 9 pixels and averages then to be a color value.  I have set it as high as 11X11 for a specific reason but that is off the lesson...There are some color contrast moves you can do with a larger sample size.

Greg



Return to “Curvemeister 101 September 2010”

Who is online

Users browsing this forum: No registered users and 29 guests