Home Page

Jump Page

Hall of Illusions


Applet design: Al Seckel  Programing: Anthony Azevedo 

Cafe Wall©1997 IllusionWorks, L.L.C.

 Cafe Wall

     Right now there is no illusion. Set the mortar thickness to 1 and move the shift bar so that it is directly over the middle of the mortar button 2 (between the button and the number 2). You should now see illusionary wedges. The horizontal lines are no longer perceived as parallel.


     One can have a dramatic effect on the strength of this illusion by adjusting a few critical parameters.


     Use the controls to change the luminosity of the mortar (or grout). You will find that the illusion is strongest when it is in the mid-range of luminosity. The illusion is still present at either end of the luminosity scale, but it is considerably weaker. Reset the luminosity to the mid-range.

Mortar thickness:

     Use the controls to vary the thickness of the mortar. This also has a dramatic effect on the illusion. When the mortar is thin, the illusion is strong, and when the mortar is thick, the illusion breaks. Reset the mortar so it is thin.



     First notice the direction of the wedges. Then use the controls to adjust the alignment of the bricks. What happens to the direction of the wedges when you align the bricks over one half cycle? How does the alignment of the bricks affect the illusion?


Effect of viewing:

     The distortion is greater in the peripheral vision or in foveal vision when the display is somewhat blurred. Use the controls to blur the image.

So What's happening:

     The café wall illusion differs from the Münsterberg illusion in an important way. In the Münsterberg figure the black horizontal borders are isoluminant with the black regions. In the café wall illusion, the rows of tiles are separated by a thin line of mortar (or grout), which, for greatest effect, should be midway in luminance between the luminances of the black and white (or blue and yellow) tiles. The illusory effect obtained is that of a tilt to the mortar lines, with alternate mortar lines being tilted in opposite directions. The width of the mortar lines also affects the strength of the illusion. The strongest effect occurs when the mortar is narrow.

     In contrast with many other types of distortion illusions, the café wall illusion occurs very early on in the visual system, when the position of edges and brightness differences are encoded. This type of processing occurs prior to the cognitive processes of object recognition. Illusions that involve higher cognitive processes are not likely to be affected by stimulus changes that do not alter the informational content of the picture or object.

     The café wall illusion has been the subject of considerable theoretical interest and empirical investigation. In 1979 visual psychologists Richard. Gregory and Pricilla Heard advanced a "border-locking" hypothesis to account for the illusionary wedges. In 1983 M. E. McCourt offered his alternative "brightness induction" hypothesis. While both approaches were of considerable merit at the time, and were consistent with the data, they might not be necessary if it turns out that the effect occurs in the early stages of low-level visual processing.

     In 1986 Morgan and Moulden showed that twisted cords are revealed in the mortar lines if the café wall figure is processed with a band-pass spatial-frequency filter. In 1993 D. Earle and S. Maskell used a difference-of-Gaussians filter and achieved an even stronger effect. It has been suggested by D. Marr and E. C. Hildreth that this type of filter closely describes certain early stages of low-level visual processing.

     The café wall illusion demonstrates the effect of some simple image processing occurring at the retina combined with some complex processing in the cortical cells of the striate cortex. The incoming image is filtered by the center-surround operator of the retina (similar to a Laplacian or difference-of-Gaussians operator). If the receptive field of the operator is larger than the mortar width, a set of diagonal bands, similar to Fraser's twisted cords, will appear in place of the mortar line. One can see that the long wedge is actually composed of many individual wedges that are integrated across to produce a long continuous wedge.

     These findings lead one to conclude that the apparent tilt of the mortar lines is caused by orientation-sensitive simple cells in the striate cortex. These locally tilted components cause a global impression of tilt along the mortar lines. The cells interact with one another to interpret the diagonal bands produced by the retina as a single continuous line, tilted in the direction of the diagonal bands.

     The approach originally suggested by Morgan and Moulden and later amplified by Earle and Maskell suggests that the key to understanding the Münsterberg figure, and related twisted cord illusions, is that band-pass spatial-frequency filtering of the image produces locally tilted elements along the mortar lines.



     The café wall illusion is named after a nineteenth-century café located in the city of Bristol, U.K., which has this pattern of tiles. As in the related Münsterberg figure, the horizontal lines are parallel and the wedges are illusory.


Anstis, S. and C. Tyler, (1980) "Induced tilt from checkerboards: Edges vs. Fourier components," Investigative Ophthalmology and Visual Science, 19, 165

Bressan, P., (1985) "Revisitation of the family tie between Münsterberg and Taylor-Woodhouse illusions," Perception, 14, 579-585

Coren, S. and J. S. Girgus, (1978) Seeing is Deceiving: The Psychology of Visual Illusions, Hillsdale, New Jersey, Lawrence Erlbaum Associates

Cowan, T. M. (1973) "Some variations of the twisted cord illusion and their analyses," Perception and Psychophysics, 14, 553-564

Day, R. H. (1978) "A note on the Münsterberg or café wall illusion," Perception, 7, 123-124

DeValois, K. K. and R. L. DeValois, et. al. (1979) "Responses of striate cortex cells to gratings and checkerboard patterns," London Journal of Physiology, 291, 483-505

Earle, D. and S. Maskell, (1993) "Fraser cords and the reversal of the café wall illusion," Perception, 22, 383-390

Fraser, J., (1908) "A new illusion of direction," British Journal of Psychology, 2, 307-320

Gregory, R. L. and P. Heard, (1979) "Border locking and the café wall illusion," Perception, 8, 365-380

Gregory, R. L. (1977) "Vision with isoluminant color contrast: 1. A projection technique and observations," Perception, 6, 113-119

Grossberg, S. and E. Mingolla, (1985) "Neural dynamics of form perception: Boundary completion, illusory figures, and neon color spreading," Psychological Review, 92, 173-211

Haig, N. (1989) "A new visual illusion, and its mechanism," Perception, 18, 333-345

Luckiesh, M., (1922) Visual Illusions: Their Causes, Characteristics and Applications, New York, Dover Publications, Inc., 1965

Marr, D. and E. C. Hildreth, (1980) "A theory of edge detection," Proceedings of the Royal Society of London, Section B, 207, 187-217

McCourt, M. E., (1983) "Brightness induction and the café wall illusion," Perception, 12, 131-142

Morgan, M. J. and B. Moulden, (1986) "The Münsterberg figure and twisted cords," Vision Research, 26, no. 11, 1793-1800

Moulden, B. and J. Renshaw, (1979) "The Münsterberg illusion and 'irradiation,'" Perception, 8, 275-301

Münsterberg, H., (1894) Pseudoptics, New York, Milton Bradley

Münsterberg, H., (1897) "Die verschobene Schachbrettfigur," Zeitschrift für Psychologie, 5, 185-188

Robinson, J. O., (1972) The Psychology of Visual Illusion, London, Hutchinson University Library

Stuart, G. W. and R. H. Day, (1991) "The Fraser illusion: Complex figures," Perception and Psychophysics, 49, No. 5, 465-468

Stuart, G. W. and R. H. Day, (1988) "The Fraser illusion: Simple figures," Perception and Psychophysics, 44, No. 5, 409-420

Taylor, S., and J. M. Woodhouse, (1980) "A new illusion and possible links with the Münsterberg and Fraser illusions of direction," Perception, 9, 479-481

Wade, N., 1982. The Art and Science of Visual Illusions, London, Rouledge & Kegan Paul

Woodhouse, J. M. and S. Taylor, (1987) "Further Studies of the café wall and hollow squares illusions," Perception, 16, 467-471


Entire web site©1997 IllusionWorks, L.L.C.