The visual system of humans and other primates is highly developed. The large number of photoreceptors in the eyes allow for a very high resolution, i.e. the ability to see very small details. The system is also very sensitive. Under optimal conditions, a single photon can be perceived. This sensitivity is matched by an enormous dynamical range that allows us to see in moonlit nights and on a glacier in the sunshine despite an approximately 1 million fold change in luminance between these two situations. While the anatomy of the eye might suggest that the visual system can be thought of as a 'biological camera' this is not an appropriate analogy. Rather, the visual system strongly modulates the information falling on the retina. This process overcomes or ameliorates some of the deficiencies of the eyes, such as spherical and chromatic aberration. It also enhances some signal aspects in the image, such as the apparent contrast at luminance boundaries. Most prominently though, the visual system imposes a number of modifications on the transition between the image on the retina and perception. These modifications serve to compress the enormous amount of information received by the eyes, to apply some assumptions about the world when faced with ambiguous information and to recover the relevant information in the presence of spurious information or noise. A multitude of visual illusions have been described, many of which illustrate these modifications. The following list of websites shows a collection of these visual illusions that provide sometimes dramatic demonstrations of vision as a constructive process.

persci.mit.edu/gallery
www2.psych.ubc.ca/~rensink/flicker/
www.simonslab.com/videos.html
michaelbach.de/ot/index.html
www.psy.vanderbilt.edu/faculty/blake/demos.html
escherdroste.math.leidenuniv.nl

The most frequently used visual stimulus in our laboratory is a random dot pattern (RDP). RDPs are created by randomly plotting a specified number of small dots within an invisible boundary. Normally, we use a circle as this edge. Computer monitors work by refreshing the display about 60 to 100 times a second. This means that a new image or a change to an existing image can be made about every 10 to 15 ms. To create moving RDPs we therefore systematically change the position of every dot in our RDPs to create a series of pictures, much like a movie in the cinema consists of the presentation of many still images in quick succession. If the refresh rate of the display or the movie projector is high enough we perceive a smooth motion rather than individual images. Below you will find a little animation that gives you an idea of what a moving RDP looks like. Since the internet is not well suited for presenting highly controlled stimuli this is only an approximation of what we use. In our laboratory the motion will be much smoother. Even though the idea behind the RDPs is very simple it allows for a large number of variations. By changing the rules of how dots change their position from image to image one can not only generate the homogenous uni-directional movement shown below but rotating, expanding, contracting and many other forms of motion.