When I began making a list of topics relevant to an article on night photography, I quickly realized that I could fill a book on the topic, not just one article, with chapters on technical, aesthetic, lighting, exploration, personal comfort, and many more issues. So I’ll begin by saying that I’m just going to focus on some key challenges specific to digital night photography, and save the rest for the book!
The Feeling of Darkness
Photography requires light. Night is essentially dark. Rendering an image, particularly a printed image, that creates the feeling of the nighttime viewing experience is incredibly challenging. Our perception is complex; we adjust dynamically to lighting conditions, and can see an extraordinary range of contrast. Cameras cannot handle nearly as much contrast, but through long exposures, they can gather much more light than our eyes can and reveal things we simply cannot see. The first example image was shot on the roof of a building in New York City after a snowstorm. The low cloud cover and the snow were bouncing and mixing the multi-colored city lights and the ambient skylight to create these astonishing pink and orange colors. It was barely apparent to the eye, but extraordinarily rendered even by the three megapixel digital camera I used in 2001.
Exposure and Shadows
In order to feel like night, your image must be dominated by dark tones. But how dark? That fundamental decision will determine your visual style as much as your compositions. Take as an example one of the undisputed masters of contemporary photography, Michael Kenna. He is particularly known for his night work. He makes liberal use of silhouettes, a fantastic technique to simplify a composition to pure graphical forms. Some of his images use an almost black-on-black palette with an incredibly narrow tonal range. Compare this image of the Rio de Janeiro skyline to this much darker one.
And herein lays the most formidable technical challenge to night photography: how to deal with shadow tones (silhouettes excluded). The three fundamental principles you need to understand are:
Digital sensors are much better at rendering good highlight detail than discreet shadow tones. The closer you get to black, which should be the absence of any electrical signal, the more spurious low-level electrical noise ends up appearing as pixels that aren’t black.
If you try to make your image look correct in-camera, which is to say basically dark, you will find it extremely difficult to get good shadow detail on a print. You’ll end up with muddy shadows, and have to brighten your RAW to make them printable. Brightening shadows will exaggerate noise, and because of a digital sensor’s non-linear response to light, you simply won’t have captured the very subtle distinctions in the shadows that separate one tone from another. In a nutshell, you need to expose brighter. You’re not trying to make the image look good on-camera; you’re trying to capture optimal data. Darkening a bright image will give you amazingly separated, clean shadow tones, as in the shot of the Manhattan Bridge at night. Lightening a dark image will produce muddy shadows. Needless to say, this is better achieved through lengthening your exposure time, not using a higher, noisier ISO. Use a tripod and the native ISO of your sensor. (View the bridge image on a black screen to see the detail that a white background will not permit your eye to perceive.)
If you ever plan to print your work, and I hope you do, you will need to capture and process printable exposure data. The translation from the transmissive medium of light on a screen to the reflective medium of subtractive inks on paper will always shift towards a less luminous, less contrasty image. So not only do you need to expose on the bright side, you’ll need to process an image that looks a bit too light onscreen to compensate for the translation to the print medium.
Unrelated to the shadow issue is your use of saturation. You should know that the cones in the human eye lose their sensitivity to color before we lose our ability to see tone, and the last color we see is blue before we lost color perception altogether. To create a more realistic feeling of night, try using a low level of saturation, and biasing your white balance a bit to the cool side.
That the lights themselves are clipped is not necessarily a problem; a filament is not an essential piece of highlight detail. The real trick is hiding the crossover from the clipped highlight to the glowing area right next to the bulb that does have detail. That transition will produce posterized tones – an obvious hard edge – that may require some hand-painting or selectively applied grain to hide.
All images courtesy Brian Dilg Photography LLC