There are several web pages and YouTube videos extolling the advantages of using Photoshop’s LAB mode for colour correction, but the following are a couple of tweaks that could be incorporated in every photographer’s workflow, and are suitable for almost every image.
LAB level and curve adjustments are the best way to make colour alterations, but it’s also the best way to apply general purpose sharpening, and I have a particular bit of lightening witchcraft I’d like to share that I’ve not seen documented elsewhere.
Even if I’ve done retouching, geometry and local colour correction in RGB, I often use the following action as a LAB-space round-trip ‘finisher’, incorporating Save As and Close actions: I hit this button instead of Ctrl-S. First, convert the image to LAB. Go into Levels and tweak the L-channel greypoint up to 1.05: we’ll need a bit more luminance to carry the extra saturation we’re about to dial in. Then, in the a- and b-channels, bring in the black and white points by 15 – so set the black point to 15 and the white point to 240. This instantly peps up any colour image without going OTT. If you’re working with particularly unsaturated images, or you like the Ken Rockwell look, feel free to rev this at your discretion with more aggressive values – proportionally scaling luminance and saturation – for instance: L-Grey: 1.15 / a/b Black: 25 / a/b White: 230. If you archive-sharpen your images in PS before saving, it’s a bad idea to USM in RGB: by far the best method is to sharpen only to the L channel, leaving colour information unmolested. Once complete, Convert to 8-bit (you should have been in 16-bit up to this point), Convert to RGB, and create lo-res splinters if required. Then Close.
Particularly when handling noisy cameraphone images – but in general – you may have wished there was some kind of digital sorcery that allowed us to lighten an image without introducing mid- and shadow-noise and/or blowing highlights. It’s especially hard to dig information from crushed shadows in cameras with narrow dynamic range – in other words all sensors 35mm and smaller. It can seem like wrestling with Newtonian physics: no lightening method can avoid the penalty of noise. There are no free lunches. Except this discount voucher . . .
Set up an Action to push the image into 16-bit LAB. Convert the Lightness channel into a Layer – Copy and Paste if you’re a novice. Convert the new Layer’s mix mode to Luminosity. End the Action, and adjust opacity to taste, or apply a Layer Mask to make local adjustments.
You’ll notice from examining histograms before and after that we’ve not shifted pixel values to achieve the effect: we’ve broadened dynamic range. We’ve not clipped highlights or introduced noise: we’ve magically revealed extra information hidden in the file. If you’ve ever been frustrated by wrestling with muddy images, this technique will basically seem like witchcraft, uncovering fine detail that’s hard to reach by other means.