With the proliferation of digital photography and digital post-processing came a corresponding belief that shadows were to be avoided. Now that it was easy to capture shadow detail, shadows should be avoided at all costs. This nascent phase of digital photography also introduced us to the tone-mapped HDR photograph trend which is now history. Unfortunately, some of the biases against shadows still remain from that era.
Say it with me: shadows are OK!
Although it is now easy with digital camera sensors to capture the full dynamic range of a high contrast scene, does it mean you always should? Whether or not to show shadow detail in the resulting image is a creative decision. There is no rule in photography that says there needs to be detail everywhere, and certainly not just because it's easy to capture. In fact, I would often argue the opposite: that too much detail can take an otherwise dynamic scene and turn it into a flat muddled mess.
Silhouettes take this philosophy to the extreme, and often render their subjects with no detail at all - shadow areas can be nearly or even completely black! Why would anyone want to throw away all that precious detail?
Why Photograph Silhouettes?
Silhouettes can be used to reduce complex subjects to their essence, namely their shape. For many trees and plants, this shape is distinctive and enough to identify the subject without providing so much information as to detract from the rest of the scene.
The shadow areas might not be that interesting. By making them black (or nearly so), you eliminate distracting elements and highlight the more interesting portion of your photograph, such as the light shining through. The uninteresting shadow areas can merge together into one tonal range rather than compete with each other and distract the viewer.
How to Photograph Silhouettes
Silhouettes often work best with strongly backlit scenes, so face into the light source (most usually the sun). This can often result in undesirable lens flare, so you may need to shield your lens with a lens hood and/or your hand. If your hand gets in the frame you can zoom out a little bit, and then crop your hand out of the resulting photograph so that there is no lens flare at all and you retain your original desired composition.
From an exposure perspective, silhouettes are actually easier to photograph than scenes where you want to show a full range of detail. If you do not care about shadow detail, you do not have to expose for it, and instead can just concentrate on making sure that the brighter parts of the scene are properly exposed. Most metering modes will try and "average" the light through the scene, which usually works great but not so much for silhouettes where such an exposure would render the shadows too bright. So use manual exposure, or dial back the exposure compensation a few stops so that the shadows are properly dark.
That being said, you may wish to capture all or most of the detail in the field (exposure bracketing if necessary) and then later decide how much of it to keep or throw away during post-processing. This approach is more flexible, though usually if I know I want a silhouette I keep it simple and do a single exposure.
How to Post Process Silhouettes
Silhouettes are usually easier to process compared to photographs that show a full range of detail. A silhouette photograph can usually be divided into two parts: the silhouetted areas and the highlights/mid-tones. For the highlights and mid-tones, use your normal processing strategies for adding contrast, saturation, etc.
For the shadows, there are a few specific techniques you can use:
In your RAW editor (such as Adobe Camera Raw or Lightroom) bring down the shadows, blacks, or exposure sliders. These are all different adjustments and have different effects, so experiment to see what works best for your image. By default these adjustments are global, meaning they affect the entire image. Often this is exactly what you want, but if you need finer control, you will need to use Photoshop to use masks to fine tune your adjustments.
In Photoshop, you can selectively burn (darken) areas of the photo that might still be too bright or otherwise distracting. To do this, I create a dodge/burn adjustment layer, and paint black through a darks luminosity selection (more information on this "luminosity painting" technique can be found in Sarah's Ebook/Video on processing in black and white, or on Tony Kuyper's website). I generally find this to be the most controlled way to selectively darken portions of an image (regardless of if it is a silhouette or not).
Also in Photoshop, you can use a levels adjustment layer and move the black point and/or midpoint to the right to darken the shadows, or you can use a curves adjustment layer and drag the shadow areas down. This technique works best if most of the shadow areas are in the same tonal range. If they aren't, you will need to fine tune with layer masks.
"Inverted" Silhouettes
You can also change it up and do what I call inverted silhouettes, where the majority of the photo is in shadow but there is only a small portion of light, as in this photograph looking up from a slot canyon:
Conclusion
Don't be afraid of the shadows! Silhouettes are one tool, among many, that every photographer should have in their photographic toolbox. While silhouettes are not necessarily appropriate for every scene, they are appropriate for many scenes and can often be used to make a more powerful image.
While this post concentrated on silhouettes (the elimination of shadow detail), the same process can be applied for high key images which often throw away highlight detail. Many photographers are increasingly fine with throwing away shadow detail but bristle at the thought of throwing away highlight detail. My philosophy, as always, is that it depends on the image. Do not unnecessarily box yourself in! If you are unsure capture the full range of detail in the field but be open to throwing parts of it away in post processing. The result may be a stronger and more impactful photograph.