All digital-computer based systems represent real-world images as a series of discreet picture elements which are commonly known as pixels or sometimes pels. Each of the pixels is used to record the luminance and colour of the image at a particular point in space (x,y) and time (t) the idea being that if the samples are closely spaced enough when they are subsequently displayed on an output device they will appear to coalesce and form a continuous display.

The pixels can be loaded into the computer’s memory in one of two ways: either a real world scene can be converted into pixels using a device known as a scanner so called because it ‘scans’ the image line by line, or they can be generated mathematically using a knowledge of the desired stimulus. In many respects the mathematical description has advantages over using ‘real life’ images as it allows the researcher to know and control the important parameters of the stimulus such as the frequency content or the average luminance. Sometimes though, particularly when researching higher order cognitive functions, stimuli showing the sort of variations found in real-life may be preferable. Once inside the computer, the image must be processed ( if necessary) and then presented to the subject using a display device. No one has yet invented a display device that can take a digitised stimulus and transform it back into a continuous light source such as you might see from your window so we are reliant on systems that reproduce the pixels in a similar manner to that in which they are stored within the computer. In other words display devices can be considered as being made of thousands of little individual light sources each of which is modulated by the stored pixel data. Each of the light sources should be thought of as having a uniform emittance across its whole emitting surface. Luckily, as everyone knows from watching television, under certain conditions the observer can’t resolve the individual pixels and the image appears indistinguishable from one generated by a light source that is truly continuous in space and time. This of course begs the question How many pixels are needed to achieve this effect and how frequently must they be changed? The density of pixels per unit area is known as the spatial resolution whilst the rate at which the pixels can be changed in time is known as the frame rate. This term is derived from CRT display technology where one complete scan of the display is called a frame but should not be confused with a similar term used by TV engineers for one half of a pair of interlaced scans which is called a field. Note that input devices, computers and output devices can all possess resolutions in exactly the same way and in a well-designed system these will tend to mirror each other in order not to have unused (but paid for) capabilities in any particular component.

^ Back to top