ISO is perhaps one of the most difficult aspects of photography to understand. Its origins are based in film photography, where it was used to rate the sensitivity of film. The term has crossed over to the digital age as a way of rating the how the sensor reacts to light. Unlike in the film era, where one particular type of film would have a specific ISO, these days digital cameras have one sensor that can be many ISO ratings. In this video, Dylan A Bennett sets out to explain how a sensor works in relation to ISO:
Using a simple blackboard style of graphics, Dylan shows us that the sensor is basically a grid of microscopic sensors that we know as pixels. Taking a single pixel as an example, he explains in a bar graph, how the light is read by the sensor, by breaking it into the primary colours of Red, Green and Blue. The ratio between the three represents the final colour the pixel will show in the image.
The very top of the graph, 100% represents the maximum amount of light the sensor can receive. With a new graph, Dylan shows us that when the sensor is receiving a lower amount of light, the percentage of Red Green and Blue, compared to the previous graph, is much lower resulting in a dimmer image. To brighten the image, the camera, artificially lowers the 100% level to nearer the top of the RGB values, effectively scaling up the sensitivity of the sensor to light.
Dylan then goes on to explain that there has to be a trade off to this scaling up of sensitivity. As each sensor is electronic, the circuit that connects it gives off a magnetic field. As the sensors are closely packed, and the camera contains many electronic circuits, the magnetic fields, interfere with each other resulting in tiny changes to the signal.
To demonstrate the effect of this we are shown that even in a graph representing a completely dark room, the effect of the interference means the sensor believes that there is some Red, Green and Blue light hitting it. When recording something with actually light in it, the sensor adds on this tiny amount of interference, changing the color of the image very slightly from the original and adding a tiny amount of noise to it.
When the light levels are low and the sensor scales up the light, it also scales up the interference and hence the noise levels in the image. Dylan goes onto explain that this is where the term ISO is used in digital cameras. The optimum ISO is when the sensor is producing the minimum amount of interference or noise. In the example 100 ISO is the optimum and Dylan shows us the effect of how scaling up to 400, 800 and 1000 introduces progressively more noise into the image.
Dylan’s interesting explanation goes a long way to show how smaller sensors with many pixels will have more noise than larger sensors with fewer pixels.
Like This Article?
Don't Miss The Next One!
Join over 100,000 photographers of all experience levels who receive our free photography tips and articles to stay current: