spot 33

spot 33

Look at the rotating afterimage and see how the coloured spots disappear. Concentrate on the centre of the picture, look through the picture without moving your eyes or head and see how the rotating afterimage slowly develops while the motionless coloured spots fade.


A circling afterimage over a fading world

This animation consists of a sequence of 12 stationary images in each of which one of the coloured spots that form the colour wheel moves in turn towards the centre of the circle.

Various perceptual effects are created which provide information on how our visual system functions:

  1. The sudden disappearance of individual coloured spots triggers a non-existent negative afterimage on the retina where they were located.
  2. The displaced coloured spot is real (it physically exists) and rotates together with the neighbouring virtual spot (negative afterimage). The afterimage appears to glow in the complementary colour.
  3. he afterimages which have only been minimally displaced in both time and space trigger an illusion of motion in the brain's cortex (phi movement).
  4. If you manage to keep your head very still and suppress your eye movement, the rest of the colour wheel fades the longer you look. (Troxler effect)

The negative afterimages light up in the complementary colour of the missing object.

Take your time and observe in the second animation how the missing spots in the slowly rotating coloured gaps light up in their complementary colours. Compare these with the original colours of the displaced real spots as they rotate together.
The occurrence of the negative afterimage is a retinal effect. This simple optical experiment allows us to draw two conclusions:

  1. The afterimages also occur at relatively high rotation speeds so that there is very little time for a single afterimage to be created. Generating this effect in the cerebral cortex would be more sluggish and require more time. For example, memorizing a virtual coloured spot requires short-term memory to be activated in the cerebral cortex and needs longer (more than 100 milliseconds). When we watch the faster animation, it is hardly possible to memorize the individual colours as the data flow into our short-term memory is hampered by the rapid flow of new events. However, with the slower animation, the viewing length of the individual events is slow enough for us to register the colours of the individual objects and their afterimages.
  2. The location where the afterimage occurs is linked with the retina. Afterimages automatically shift when we move our head or turn our eyes. The individual colours of the afterimages do not correspond with those of Thomas Young's and Hermann von Helmholtz's three-colour system of our colour-sensitive receptors, but with the four-colour system operating in the retina. (The retina's nerve cells are really displaced parts of the brain which process visual signals in a multi-layered neural network.) The complementary colour of red is therefore not a lush green but a shade that moves into the light blue range. See for yourself.

Why do negative retinal afterimages occur?

The photopigments of the retina are bleached by bright light. This light excites the receptors. Because their photo-chemical substances have to be regenerated after extended exposure, the retina is temporarily less sensitive to the mix of wavelengths that reach the eyes. When we see grey, the wavelengths of the complementary colour have a stronger effect and create a negative afterimage on the retina at that location. Normally, no afterimages occur in our visual perception, because we unconsciously move our eyes three times a second (saccades). The bleaching of the photopigments or - in other words - the adaptation of the rods and cones has no effect because the retina has to deal with constantly changing stimulus patterns at this location.

The cerebral cortex treats the virtual objects that replace the missing spots of colour as real objects.

Fortunately, our cerebral cortex creates a continuous film from the discrete stationary images. This imaginary movement is known as the phi phenomenon and is discussed in greater detail in connection with Spot 26. The rapid film of afterimages that we see proves that our perception of movement in area V5 does not distinguish between real objects (the coloured spots that are moved in towards the centre of the circle) and virtual objects (the afterimages in the gaps in the colour wheel). Both are processed as if they were identical.

The brain also accepts the change in colour of both the “actors” in this film. In Nature, objects can also change colour when lighting conditions change, but this should not cause them to lose their identity. This so-called constancy of an object requires on-going correction for changes in lighting conditions. (Modern video cameras do the same thing by periodically performing a white balance). In addition, area V5, which is responsible for our perception of movement, is practically colour-blind and only processes the brightness values of the individual snapshots. How this discrete visual data is processed to create a continuous effect of motion is to a very large extent unclear and is therefore the subject of current research.

Stationary objects without clear contours fade in our perception

We have inherited our visual system from the animals from which we are descended. It is primarily intended for decoding the movements of objects rather than for looking at pretty pictures and stationary situations. As a rule, objects have clear contours, which are registered by the motion-sensitive neurons in V1 and tracked through time and space. The fovea (the tiny area of the retina centred around the optic axis) enables us not only to see sharply, but also to distinguish changes in colour more accurately and even to register stationary objects that are blurred and unsharp.

If we concentrate on the small grey circle at the centre of the image when we look at the animation, it is only this that is imprinted on the fovea As this keeps the fovea occupied, the unsharp spots of colour therefore have to be analysed in the peripheral area of the retina, which is less well equipped with colour receptors and has very poor acuity. In addition, we are only able to register objects with our animal eyes when they move in relation to our retina. If movement of both eyes and head is deliberately suppressed, our decoding of objects with the peripheral retina breaks down. The negative images that are superimposed neutralise the stationary images and the coloured spots fade as if they were not there (Troxler effect). Our brain then interpolates the gaps in the image that thus occur in the background colour and simulates an imaginary harmless situation without any coloured spots. The exception to this are the two rotating objects described in the previous section (one real and one virtual) because their stimulus patterns move relative to the retina and trigger neural activity in the motion sensors of the cerebral cortex.

A person whose head is held motionless and whose eyeballs are prevented from moving when the eyes are open is blind if the world around him does not move.

How this spot came into being

The “rotating pink dot“ animation with 12 magenta spots forming a wheel together with a rotating gap spread on the Internet during the course of 2005. The fascinating green afterimage that occurs in the gaps and the fading stationary ring sparked off a snowball reaction. In 2005 Google returns some 135,000 hits for the search term “rotating pink dot“.
The creator of this illusion was Jeremy Hinton, who discovered this dynamic afterimage effect quite by chance when he mistakenly omitted to erase a number of positions in a previous animation using a rotating disc.
Michael Bach parameterised the original version and made it possible to vary the speed and adjust the saturation and colour of the twelve spots.
In this present version by blelb, the colours in the wheel vary. In addition, a spot rotating at the same time enables a precise comparison to be made between the 12 basic colours of the colour wheel and their afterimages.