Long Exposures at Night

Facebooktwitterlinkedin

Today’s Question: How do I capture long exposures at night without blowing out the windows of the buildings?

Tim’s Quick Answer: There are a few options here. The first is to set your exposure to retain detail in the brightest areas (in this case the windows of buildings). The second is to create a high dynamic range (HDR) image. The third is to not worry about the lost detail. And the fourth is to capture your images earlier in the evening before the contrast is so extreme.

More Detail: The problem with creating a night exposure that retains all detail in the bright lights (such as windows of buildings where the lights are on) is that the rest of the image will generally be extremely dark. As a result, if you want to capture only a single image you’ll need to choose between detail for the highlights versus detail for the shadows.

Creating an HDR image is a reasonable option that can greatly improve the overall tonal range represented in the final image. However, with long exposures there are some inherent challenges to HDR. To begin with, it can require considerable time to capture all of the images in the HDR sequence. On a related note, the longer time required to capture all of the images can result in problematic changes within the scene from one capture to the next.

My recommendation is to shift your night photography to earlier in the evening. If you capture your long exposures late enough that the lights of the city are on, but early enough that there is still some degree of illumination left in the sky, you’ll have a narrower range of tonal values to contend with, so you will be better able to capture the full range (or nearly so) with a single exposure.

Note, of course, that you can employ a neutral density filter to extend the exposure longer that would otherwise be possible when photographing in the relatively early evening. But in general, if you photograph as soon as the lights of the city are turned on, you should be able to achieve excellent results with a single frame, likely without the need for a neutral density filter.