This is a sample chapter from Jim Harmer's book Improve Your Night Photography. Check out the book!
The success or failure of your night photography will likely depend on whether or not you have fully mastered your low-light shooting technique. A skilled night photographer can leave a shoot knowing that all of her images were shot correctly to attain an image with very little noise in the image, while an untrained photographer will usually leave a night shoot with nothing but frustration if he has not mastered his low-light technique. Any digital photographer knows that digital noise can be difficult to control in low-light situations, but what exactly causes noise?
Noise is like an audio cassette. When an audio cassette records a sound that is quiet, the listener often turns up the volume on the player to make the player more sensitive to the soft sound. When the listener does this, the player boosts the sound of the voice on the recording, but it also creates an undesirable fuzz that drowns out some of the voice on the recording. Just like audio cassettes, when a camera makes itself more sensitive to light by using a high ISO, unwanted fuzz enters the image. This fuzz, or digital noise, is created when stray electric signals on the camera sensor are interpreted as light.
The most common type of digital image noise that concerns photographers is called salt-and-pepper noise. This type of noise manifests itself with dark pixels appearing in light areas and light pixels appearing on dark regions of the photo. Dark regions of a photo show noise more plainly that lighter regions because the lightly colored noise sticks out on a dark background. This is of special concern to night photographers, because large portions of the image are often dark.
In addition to high ISO values, long shutter speeds also increase the amount of noise in an image. This is because the longer the image sensor attempts to collect the weak signal, the more stray electric signals it also collects. This creates a level of complexity for the night photographer, because it raises questions such as: “Will a 20 second shutter speed at ISO 200 produce more or less noise than a 10 second shutter speed at ISO 400?” This question is impossible to answer because it varies on every camera model. Cameras with better low-light performance will usually show less noise with a higher ISO value and shorter shutter speed, but other cameras will perform better with the opposite arrangement. Unfortunately, the best answer to this question is to come to know your own camera and how it performs.
You probably knew that high ISO values and long shutter speeds increase the amount of noise on the image before reading this book, but you may not know that these are not the only two factors which contribute to the amount of noise in an image. We’ll also consider ambient heat, camera heat, and photosite size.
Heat is a form of electricity and cameras record the electric signals from heat just as they record electric signals from light. Therefore, warmer temperatures create more digital noise than colder temperatures if all other factors are equal. This is a consideration for photographers when shooting long exposures at night. When traveling, keep in mind that warmer climates will produce slightly more noise than colder climates, so you will need to change your shooting style. Even the best-built DSLRs heat up as they are used. When long exposures are recorded, the camera must work for extended periods of time and will heat up quickly. To combat this problem, take breaks with the camera turned off between long exposures. If the lens can be removed momentarily without concern for dust entering the camera, this can cool down the camera quite quickly. While taking steps to cool the camera is usually not necessary, it can help to control noise in difficult shooting situations.
The last contributor to the amount of noise in an image is the photosites on the image sensor. You probably know that DSLRs produce less noise than point-and-shoots, but why? The sensor on a DSLR is larger. Just like a large umbrella touches more rain than a small umbrella, a large image sensor is able to collect more light than a smaller sensor. This light-gathering ability makes the light signal strong enough to differentiate it from the stray electric signals and results in less noise in the image. This is commonly referred to as the signal-to-noise ratio. Suppose a DSLR and a point-and-shoot both have 14 megapixels. On the larger DSLR image sensor, the photosites (the tiny pixels on the sensor that record light) will be spread further apart than the same amount of pixels on the point-and-shoot. Since each photosite is further away from its neighbor, the stray electric signals produced by the neighboring photosite is less likely to affect the other photosite. This is why a 35mm equivalent sensor (so-called “full frame sensor”) will generally perform better in low-light than an APS-C size sensor with the same number of megapixels. The full frame sensor allows each photosite to be larger and further from neighboring photosites. This principle also explains why photographers generally push camera manufacturers to reduce the number of megapixels on a DSLR, because it allows for larger photosites and better low-light performance. Camera manufacturers have been slow to bend to photographer’s demands in this area because so many novice photographers judge the quality of a camera by the number of megapixels.