News & Events in the World of Photography, Printing and Design
Digital Photography Tutorials – Image Noise And Noise Reduction
14th Oct 2008
Anyone who owns a digital camera will be familiar with image noise (that grainy distortion that spoils pictures shot in low light conditions or when high ISO settings are selected) but not all will know that it is called " image noise", where it comes from and what can be done to prevent it from happening.
It all stems from the fact that all electronic devices generate "noise", not the type you can hear (at least not all the time, the background hiss of a radio being an exception) but scattered electrical interference that causes sounds and images alike to be recorded and played back in a distorted manner. In fact all electronic signals contain of a certain level of noise as a background to the actual useful information being generated or transmitted, the relative level of information to background noise being called the ‘signal to noise ratio'.
Digital cameras suffer noise from a variety of sources, some being internal noise which is generated by imperfections in the electronic components of which the camera is made, while others are a by-product of normal operation. The electronic components in the camera can also be affected by environmental noise, such as the electrical fields and electromagnetic radiation that constantly surround all of us.
The circuit noise in a camera can be minimised by efficient design, superior manufacturing techniques and rigorous quality control, however, as you can imagine many cheaper camera brands, and the budget camera ranges from the well known brands, often use components that are built "down" to a price rather than up to a standard. The result is that such cameras produce noisier images than their more expensive brothers.
In Digital Cameras the image is captured by a Sensor and this is the other main source of image noise, which is most cases this is unavoidable, the job they are doing being incredibly complex. Still every day advances are being made and in the long run, at least for the expensive models, Sensor noise will be a thing of the past.
The size of the sensors in most cameras (just 1/2.3 inch) now commonly have over 10 million individual photocells which are all crammed into an area of less than 30 square millimetres. As these light sensing cells are so small, it means that in low light conditions they might only be collecting a few thousand photons (the individual 'particles' that make up light) during an exposure. This low level of data and together with even low levels of electrical "noise" mean that that random statistical fluctuations in photon density cause degradation to the end image. This is the main reason that physically larger sensors are much better than the smaller ones, as, as the individual photocells are larger and collect more photons during the exposure any statistical variations (and electronic noise) have a lower effect on the image, the signal noise being proportionately higher than the noise ratio.
Due to the fact that the engineers have done the maths (and can thus cancel out noise when there is plenty of "signal" - light - about), it is only when light levels fall that problems start. You see when you are shooting in low light the level of the signal drops nearer to the constant noise level, thus producing a lower signal to noise ratio. And when the light gets too low, the signal can be entirely drowned out by the noise and all you get is a "snow" picture..
This problem is made worse when shooting at higher ISO settings, as when you set a higher ISO number, what you are really doing is increasing the amount by which the signals from the sensor photocells are amplified. As this amplification process can't tell the difference between an image signal and a noise signal, the noise gets amplified as well and the image just gets worse and worse. This is why high ISO images are always more noisy than ones taken at lower settings.
For more details and ways to work around the issues, please visit
Back to Photographic News Listings
