Image Noise

The electronic sensors used in digital cameras do not always record light perfectly. As we discussed before, you can think of the sensor as being made up of millions of tiny light collectors, each of which is keeping a count of how much light hits the sensor. This count is not always 100% accurate, and will usually be a little bit higher or lower than the true amount of light seen. If enough light is captured then this small change doesn’t really affect the photograph, but in low light levels this shift can be enough to change the colour of the pixel.

Take a look at the image taken below, showing some fish swimming in a dark tank of water.   This was a challenging situation to shoot well – a short shutter speed is necessary to capture the fish without blurring, but it was not possible to increase the aperture enough to create a good exposure.  This made it necessary to try shooting in high ISO – a topic we will cover in more detail in the next lesson.  Unfortunately, as we will see, a high ISO can mean a noisy image.

Notice how the magnified image shows a kind of grainy structure with many pixels showing slightly different shades. Although we expect to see smooth dark water the camera has detected small amounts of light (either real or as a result of various electrical processes in the sensor) and has tried to represent this in the image. This has had the effect of lowering the overall picture quality.

Noise can be made worse in several ways, and we will run through the most important of these in this lesson.

 

Poor Light Conditions

The most important contributor to noise is low light levels. The less light the camera gets the more random fluctuations will dominate the light the sensor sees and the more noise will end up in the final image. To reduce the impact of this try to increase the exposure. The best way to do this is to increase the time over which the exposure is made. Alternatively increasing aperture, or introducing new sources of light (such as the camera flash) can help.

ISO

You might remember that there are three common ways to increase exposure – shutter speed, aperture and ISO. To reduce noise both shutter speed and aperture are useful, but increasing the ISO can actually make noise even worse. We will go into more depth about ISO in the next lesson, but essentially ISO makes the sensor more sensitive to light. Doing this can increase performance in low light levels, but it also means random fluctuations have a bigger impact. This is especially an issue at higher ISOs, and depending on your camera you may find that noise becomes a significant problem at ISOs above 1600.

Megapixels

Although camera manufacturers like to market their cameras by the number of megapixels, increasing megapixels is not always a good thing. The more pixels a sensor has the less light each pixel will receive (unless the sensor size increases in proportion). This means that each pixel will be more affected by noise – and so increasing the number of megapixels can result in increasing the amount of noise in your photographs! This can be a real problem with small compact cameras or with smartphone cameras, which often claim high megapixels but have small sensors, and so often struggle in poor light conditions.

Noise can to some extent be dealt with during post-processing. This involves smoothing the image, and lowers noise at the expense of image quality. However, the best way to reduce noise is to select the right combination of settings when shooting the photo – so keep ISO as low as possible and try increasing exposure through aperture or shutter speed. Alternatively, try to increase the light available, either by adding extra light sources or by making use of the camera flash.

Finally, higher quality professional DSLR cameras often contain features designed to reduce the impact of noise. This allows such cameras to use a larger range of ISO before noise starts to become such a problem.

No comments yet