More about Lab Color and Other Tricks

A few days ago, I shared a before and after pair of images of a Maine lighthouse photo shot by my friend Jim Rowson. I loved his composition and the lighting in the shop. Since I had been learning some new photo enhancement techniques with my friend and photo buddy Doug Kaye, I asked Jim if it’d be OK for me to share his image here on my blog along with an enhanced version. That was OK with him!

By way of reference, here’s Jim’s original image:

New england 24

and here’s my edited version:

New england 24 lab color edit

In my earlier post I really didn’t provide any details about the enhancements steps I took to adjust and, hopefully, improve this great image. In this post, I will share with you what I can remember about this process.

My work here is the result of lot of concentrated learning over the last month or so – beginning with a workshop that Doug arranged with Mark Lindsay. From there, I’ve learned a whole lot more – and I’ll include a list of the most important resources I’ve found at the end of this post.

As an aside, one of the great shortcomings of today’s photo editing tools is how it’s so difficult to actually log the changes one makes to an image. You can, of course, choose to keep the fully edited version (in Photoshop these can be very large PSD files) – or, if you’re like me, you can just complete your editing in Photoshop, commit your changes (by merging down the layers), and save the final result back into Lightroom. Too bad there isn’t a better way to do all of this – one which doesn’t eat up huge amounts of disk space while still providing the opportunity to look back and learn from a particular editing experience. Seems like an opportunity here for providing a much better experience!

So much for my lament, let’s get back to Jim’s image – and my editing of it.

Because I started with a relatively small size JPEG version of the file from Jim (and not the original RAW file out of his camera), I completely skipped the otherwise important workflow steps involving the initial RAW processing of an image I might have shot in the field. I switched to Lightroom (from Apple’s Aperture) a couple of years ago and it’s been my photo database and RAW processor since then. As Doug described in his recent workflow post, I do the basics in Lightroom (Lens Calibration, Camera Profile, White Balance, basic Exposure, etc.) but then do an export of the image to Photoshop where the bulk of my serious image editing takes place. For most images, Lightroom’s editing tools are more than adequate – but for those special images need more serious blending and “chording”, a trip to Photoshop always results.

I’ve included below an annotated snapshot of my layers in Photoshop from editing Jim’s original image. Let me walk through each layer in turn – from the bottom to the top – as I added them in Photoshop while editing this image.

PhotoshopLayers 2

The original image comes into Photoshop as a Background layer. I’ve recently learned (from Mark Lindsay, Lee Varis and Dan Margulis among others) the techniques associated with improving images using Lab color and other techniques. These techniques typically begin by examining the three color channels of the RGB image (Red, Green, Blue) and selecting the one with the most interesting contrast to use as the basis for the second layer.

To get there, I created a new layer – the second layer from the bottom in the diagram – somewhat confusingly labeled as Layer 1 in the diagram – and then used the Apply Image commend to select the Blue channel and load it into the layer. I then changed the layer blend mode for that layer to Luminosity – so that only the tonality of the image is adjusted by the Blue channel. (Note that for most images the Green channel is actually the best one to use for this first step – but in looking at all three channels of this image I decided that the Blue channel provided the most interest).

Next, for the third layer (labeled Layer 2 in the diagram) I created another new layer, this time also creating a Clipping Mask to link this layer with the second layer. You can tell that a Clipping Mask is applied by the indentation of that particular layer in the Layers panel. I then used Apply Image to load the Green channel into this layer and selected the Multiply blend mode to darken the image. Multiply typically goes too far, however – so I reduced the opacity of this channel to 20% which, to my eye, looked about right.

For the fourth layer, I duplicated the third layer and applied Overlay blend mode along with a Clipping Mask. I only wanted the Overlay effect to apply to the brighter portions of the layer – so I used the Blend If option of Layer Styles to eliminate the effect on tones darker than neutral (128).

For the fifth – and final tonality adjustment layer – I added another new layer with a Clipping Mask and used Apply Image to load the Red channel (the brightest of the three). I used the Color Dodge blend mode to have this brighten the sky area. I applied a layer mask using the gradient tool to block the effect of this layer on all of the areas below the sky.

Next, I duplicated this image – with the five layers so far – and selected Merge Layers while duplicating. I switched to the duplicate image – now a single layer – and immediate changed the image into Lab color. Next, I added a new layer in Overlay mode. I used a Layer Style option to eliminate the L channel from the layer. I then merged the two layers into one, used the Move tool (with shift) to move the image back on top of the original image. I made sure the new layer was the topmost layer of the original image and released any Clipping Mask that might have remained. I then changed the blend mode of the layer to Color and adjusted the opacity to my taste.

For the seventh layer, I brought in some luminosity from the Lab version’s b channel – using Overlay blend mode and adjusting opacity to taste – ending up at 45% opacity. This is totally subjective and done according to taste. Depending upon the colors in the image, the a or b channel from the Lab version might be better – you can try one or the other – or decide just to skip this layer.

Next, I duplicated the image once again – again merging the layers. I switched this duplicate’s mode immediately to CMYK color. Using the Channels menu, I selected the Black channel and applied a contrast enhancing curve to the black channel. This is a bit tricky – you need to be sure the black channel is the only one selected, etc. The goal is to have a stark contrast in this channel.

I then came back to the original image and added a Solid Color adjustment layer – using neutral grey (128 with no saturation). I applied a layer mask and loaded it using Apply Image from the Black channel of the CMYK image. I then switched the image to Color blend mode. This step is intended to reduce any inappropriate color tones from the shadow areas of the image.

At this point, I’ve completed the core adjustments in Photoshop. At this point, I selected all of the layers and flattened the image to a single layer.

Next, I used a couple of Photoshop plug-ins to complete the image. First, I used ALCE 2.1 (at 50 percent in Turbo mode) to add local contrast to the image. Second, I used the Margulis Picture Perfect Workflow panel (free download) to sharpen – specifically using his Sharpen 2012. And, finally, I used Nik’s Dfine to reduce any noise remaining in the sky portion of the image.

That’s it. Each step takes you along the path from a nice image to a much more interesting version.

Resources

We stand on the shoulders of giants someone once said – and, certainly, when it comes to image adjustment in Photoshop that’s very true. In my particular case, I benefited greatly from the following:

What’s Next?

I feel very much like I’m just on the cusp of understanding what’s happening here. Photoshop is such a complex program – with a user interface that adds complexity while enabling great power.

I’m hoping some of these techniques “stick” with me – based on a more fundamental understanding of what they’re about, how they work in the context of what’s being manipulated, and great familiarity with using Photoshop.

Over the summer, I felt like I had finally broken through some sort of plateau in my understanding of image manipulation. The last four weeks have kicked that appreciation into “turbo mode” – here’s hoping things stick and settle down going forward!

Please let me know if any of this helps you in your struggles like mine towards better images! I’d love to hear from you!