Making a good digital black and white photograph involves much more than just using “Desaturate” in Photoshop. I have spent quite some time working out a good, efficient workflow for my B/W images which I’d like to share here.
My workflow is broadly based on Petteri Sulonen’s, and I suggest you read his essay, as it’ll make my elaborations easier to understand. One particular point in Petteri’s Essay I must stress here again: Expose for the highlights! Make sure you have detail in the highlight regions – don’t let them blow out.
The complete set-up for my workflow is contained in a single Photoshop action which you can download here. The nice thing about this action, compared to Petteri’s way of working, is that it is completely non-destructive, yet is nearly as full-featured, and, at least for me, much more efficient.
The action assumes a single layer called “Background”, which it doesn’t modify. Instead, it builds the adjustment layers “Darken”, “Brighten”, “Channels”, “Levels”, “Contrast”, and “Tint”, in addition to a copy of the background layer called “Colorize”, which is disabled by default. If you are not completely comfortable working with adjustment layers I recommend watching the “Essential Adjustment Layers” Video Tutorial on The Radiant Vista (click through the pages to find it – it was posted in February 2006).
The “Darken” and “Brighten” layers provide dodge and burn functionality. Paint with a white brush (preferably with a soft edge) on the layer mask and the painted-on area will turn brighter or darker, depending on the layer. I often work with opacities less than 100 percent to achieve more differentiated dodging and burning. You can also play around with the curves of these two layers if you want to modify the extent or look of the effect. Non-linear (i.e. “curvy”) curves often look better – try it out!
Use the “Channels” layer to select a channel mix for the “raw” B/W image. Use a balanced mix (with about equal representation for each channel) to get the look of a normal panchromatic B/W film or change the balance to achieve the effect of using a color filter in B/W shooting.
The red channel gives human skin a “glowing” infrared-kind look and almost magically removes skin blemishes whereas the blue channel represents skin in rather dark tones and shows even the slightest blemish. The green channel is somewhat in between, so it’s often good to use a mix of red and green for portraits. Skies look very boring and uniform in the blue channel, but clouds really come out in the red channel, where the cloudless parts of the sky get very dark. In terms of noise the green channel is usually the best – most digital cameras have as many green pixels as they have red and blue pixels combined (the so-called Bayer pattern). The blue channel usually has the worst noise of the three.
The next two layers, “Levels” and “Contrast”, modify the brightness range and contrast of your image as a whole. Usually you’ll want your B/W photos to span the whole tonal range from deep black to pure white. The “Levels” layer helps with this goal. Use it to bring the darkest pixels in your photo to black and the brightest pixels to white. Having done that, modify the “Contrast” curve to achieve the image contrast you need. Spend lots of time playing around with this channel – it’s worth it! It is not strictly necessary to do use the “Levels” layer since you can achieve the same result by modifying the “Contrast” layer accordingly, but I find the workflow to be easier this way – when I use “Contrast” I can rely on my brightest pixels being on the right and the darkest on the left instead of having to “search” for them.
That’s it as far as brightness is concerned – the last two layers reintroduce color to the image. Just turn them off if you want a neutral B/W image.
The adjustment layer “Tint” generates a monochromatic tint, adjustable, via the the layer opacity, in strength and, via the hue setting in the adjustment layer options, in hue. By default it is set to a moderate but quite visible sepia tint. The layer mask of the “Tint” layer is set up so that bright pixels are tinted less than dark pixels, i.e., white is pure, untinted white while the darker tones are tinted more. Some people prefer this kind of toning, while I find the difference to be neglible. In addition, the layer mask does not automatically adjust to modifications in brightness you might have done, resulting, in extreme cases, in distracting color “shifts”. I recommend disabling or deleting this layer mask. Use it only if you know exactly what you’re doing!
Finally, the “Colorize” layer, which is disabled by default, allows you to bring the original colors back into the image. You can do this for the whole image or, by using a layer mask, only for parts of the image. The layer’s opacity is set to a low value by default, introducing only a hint of the original colors. Setting it to full opacity usually produces a look reminiscent of hand-colored B/W photos. Use at your own risk!
As with everything concerning digital photo editing I recommend that you play around a lot and try different approaches. Regard my workflow as a starting point or suggestion, not as the last word in digital B/W. If you have any suggestions or improvements to it I’d be happy to hear from you!