The reason why I still haven’t written more about LAB is simple. I got hooked! The more I read about LAB the more I am intrigues by the possibilities and the more I read about it. It’s a vicious circle.

I am now working on my 3rd attempt of the first example (yes, the usual-luminosity-suspect ) and I am still not quite satisfied. My recent discovery is how stupidly easy one can fix a color-cast in an image with LAB. With this technique you can even get rid of light fog or haze.

However, creating, and annotating, the example images is very time-consuming. And until end of summer I am still very busy, so I will have to postpone my posts a few months

Stay tuned!

The examples are done in The Gimp! It should also be possible to adapt those examples to Adobe Photoshop as it also supports the LAB colorspace. Details may vary though.

A boring example of brightening up an image will follow soon.

To get you started:

0 <= L <= 100
-128 <= A <= 128
-128 <= B <= 128

L denotes the lightness such that 0 is black and 100 is white.
A separates magenta/red and green such that -128 is green, 0 is neither red nor green and 128 is red.
B separates cyan/blue and yellow such that -128 is cyan, 0 is neither cyan nor yellow, and 128 is yellow

A fun fact: With LAB you can specify impossible colors. Let’s say, you can crank up B to the maximum positive value, to obtain yellow. What happens if you reduce the L-value? Well it becomes “dark-yellow” which does not exist. Dark-Yellow looks brown to the human eye. Yet you can specify it. The same way you can specify “bright-green”.