One of the rules of beginning photography is that you never shoot into the light. One of my suggestions is that you learn the rules but then you learn to break them creatively and this is one of those situations where you can get dramatic images by understanding what’s going on with backlit scenes.
NASA gantry at sunrise
Film and digital cameras do not have the dynamic range to be able to capture the brightest brights and the darkest darks all at the same time in one exposure. They have a range of lightness values that they can capture while still resolving the many slight brightness variances across the entire range. This array of lightness values is called the dynamic range. There are techniques you can use to either reduce the dynamic range of the scene or grab multiple exposures that capture it all.
Let’s start with the easiest scenario. One in which you don’t really want to capture it all. Backlit scenes offer wonderful opportunities to capture very dramatic images. As in the example above, I exposed for the mid range and bright areas in order to let the lows go dark and make a nice silhouette. These are pretty easy to do. They almost take themselves since the bright background will almost certainly dominate and cause you to underexpose the foreground completely. And if it’s not dark enough, you can always make adjustments in post-processing.
By now, I hope everyone is already shooting in RAW rather than the default JPEG format. RAW allows you to work with the data that the camera captures, which is 16-bit rather than the reduced 8-bit format. What that means is that your brightest and your darkest values are divided into 65,535 different shades of each of the three colors, red, green, and blue versus the 256 levels that you get with 8-bit data. Even though your output to a screen or a printer is going to be limited to 8-bit, operating in the full 16-bit allows you to decide how that large amount of data gets squeezed into that narrow output. You can recover apparent saturations and clipped data where detail would otherwise be lost.
So, what can you do to change the dynamic range of the scene? The easiest is to use light modifiers (reflectors) or flash to add light to your subject if it’s not too far away.
Gratiot Lake Road lit from behind
In this image, by Wendy Podmenik, the subjects were heavily backlit by the strong sunlight coming through the fiberglass panels behind them. I held a gold reflector off to the photographer’s left, to direct some of that light back on to their faces. Without the reflector, the subjects would have been very dark and although they could have been lightened up in software, noise would quickly start to increase, thereby lowering contrast and saturation.
The most common method of reducing the dynamic range of a scene is to use flash. I’m not a big flash user. Unless it’s done properly, it can look harsh and unnatural. There are people who specialize in flash photography and I’ve certainly done my time in a studio under controlled conditions, but I prefer to use available light which almost always looks more natural.
A more complex method of compressing the dynamic range of a scene is to take multiple exposures and then combine them in software to create a result that essentially lowers the highs and raises the lows from what is essentially a much wider dynamic range image than any one by itself. There is a lot of software out there to do this. I use PhotoMatix Pro with limited success. Sometimes it just brings out too much noise for my liking. But when it works, it works well.
A scene captured with three different exposures and then combined in software
So, what do you do when you can’t modify the scene and you can’t take multiple exposures? You do your best to place the dynamic range of your sensor within the dynamic range of the parts of the scene that you want to capture. I have a lot of experience with doing exactly that from shooting live music. This is, by far, the hardest type of photography I have ever shot. Not only is the lighting often backlit, but it is almost always changing in color and intensity (faster than you can make decisions about it and react), it’s either way too bright, or no where enough, your subject is almost always moving, and there is often fog generated which reduces the scene contrast.
For my live music shooting, I use a Canon 6D body. It has the best high ISO performance of almost any other camera on the market and surely the best of any in its price range. I have found that for so much of the backlit scenes on a live stage, I like to shoot in aperture priority with my lens wide open at f2.8. I set my ISO to automatically adjust as needed in order to keep my shutter speed up as high as possible. High is a relative term because many of my shots are made with a shutter speed of less than 1/50 and that’s at ISO values of 6000 all the way up to 12800.
Since I’m wanting to keep as much of the source light from saturating as possible, I often underexpose by 1 to 2 eV levels using my exposure compensation dial. This does often underexpose my subject but the highlights are less likely to saturate because of it.
Kenda "Obi Wan Kendabi" Legaspi of The Creepshow, lit primarily from behind
Here is such an example. There were bright lights behind and to the left, coming over her shoulder as the primary illumination, with almost no light from the front. Had I not added exposure compensation, my camera would have metered for the dark face and the lights behind would have been horribly overexposed. What’s deceiving is that I lowered the highlights, making the scene look evenly lit. By keeping the lights within or close to the dynamic range of the camera, I was able to bring up the darker areas, while taking advantage of its superior low noise performance.
Pretty silhouettes are actually quite easy to shoot. It’s when you want to preserve detail in both the highlights and the darks that you need to think ahead by either modifying the existing light or exposing in such a way that you can recover seemingly lost values during software processing.