Colour Separation Using LAB Colour Space in Photoshop
I would like to thank Dan Margulis for the technical expertise behind today’s article. He gives a convincing argument of LAB colour space’s hidden power. If you would like to find out more about LAB I can recommend Dan’s book ‘Photoshop LAB Color: The Canyon Conundrum and Other Adventures in the Most Powerful Colorspace’
Today’s tutorial is all about giving an image that has a very shallow tonal range, a much greater colour separation without creating an unnatural look. In the original image below, the clouds have a very compacted tonal range. There are colours other than blue present but they are very subtle. Using the power of the LAB colour space it is possible to enhance the colour separation in a natural way.
A very brief explanation of the LAB colour space follows.. For an in depth description I would suggest reading this article.
Lab colour space is an alternative to the various RGB or CYMK colour spaces. Lab consists of 3 channels.. L is the lightness or luminance channel and contains no colour information. This makes it very useful for sharpening or improving contrast as the colour of an image is not altered. The A and B channels contain only colour information and no luminance. The A channel contains red and green details while the B has yellow and blue.
The unique way that Lab handles luminance and colour separately allows some very useful manipulation of the image not possible in other colour spaces.
The first step in this tutorial is to load the photo into Photoshop and convert to the Lab colour space. See the image below for where to find Lab.
Next we create a curves adjustment layer.
Now the fun begins.. In the L channel we can boost the contrast of the image by creating a classic S curve as seen below.
Using Lab for contrast ensures that colours remain constant.
Moving to the A channel, you now need to find a part of the image that is neither more red than green or vice versa. Put a pin in the curve at that point (for me this was in the distant mountains). This pin is your centre point that will remain neutral, while reds will become redder and greens greener. To achieve this, drag the curve from the bottom left to the right until the curves ‘flattens out’ in the top right as seen.
Do not worry that the colours have gone really funky, this will be corrected later. Now move to the B channel and repeat the same process but looking for a centre point that is not too yellow or too blue.
As this stage, you could be thinking that I’ve ruined the image but all that is required now is to reduce the opacity of the curves layer until you are happy with the results.
As you can see below, the original image is very monotone compared to Lab curves version.
The interesting thing is that this separation of similar colours is something that the human eye does naturally but we can simulate it using Lab.
As always, please fell free to ask any questions or comment below.
For Contrast adjustments, color shifts can be avoiding when in RGB mode by changing the Blend Mode of the Curves Layer to Luminosity.
Do you believe that working in LAB gives better results for Contrast adjustments?
I think that “better” is subjective. LAB contrast adjustments are just another way to perform the same task. There are benefits to using LAB for contrast, such as the colour is not distorted when you boost contrast on just the Luminance channel. Whether this is better or worse depends on you.. Which method produces the results you prefer? Which method is quicker? Which method are you more comfortable with? These determine which is better, and that is a personal decision