Illusions observed in striped patterns have been investigated extensively but the origins of perceptual distortions remain unclear. Illusions in striped patterns were first recorded in the last century, with descriptions such as the oscillations of lines, appearing as waves on closer inspection (Purkinje, 1823; pg. 122; cited in Wilkins, 1995), and lines losing their straightness, with the spaces between such lines becoming coloured (Brewster, 1832; pg. 170; cited in Wilkins, 1995).
Patterns of striped lines can induce seizures in patients with photosensitive epilepsy and when these patients view patterns of this kind, epileptiform electroencephalographic activity are often seen (Wilkins, Nimmo-Smith, Tait, McManus, Della Sala, Tilley, Arnold, Barrie & Scott, 1984). When viewing similar patterns of striped lines, individuals not suffering from epilepsy report illusions of colour, shape and motion giving rise to headaches and eyestrain (Wilkins & Nimmo-Smith, 1984).Wilkins (1995) observed a 14-year-old girl with epilepsy, who suffered absences only when she was exposed to striped patterns. An electroencephalograph (EEG) was used to determine specific types of pattern that induced the electroencephalographic precursors of seizures.
Further investigation suggested that pattern sensitivity in epilepsy was not as rare as first thought and subsequentially led to studies that determined the stimulus characteristics responsible for the EEG abnormalities in patients with epilepsy (Wilkins et al., 1975, 1979a, 1980, 1981; cited in Wilkins, 1995). The pattern parameters found to increase the probability of epileptiform activity became the basis of research into parameters that induce illusions (Wilkins et al., 1984).Until 1984, there appeared to be no study that examined the spatial parameters of patterns to determine the properties that induced the documented visual illusions.
As a result, experimentation by Wilkins et al. (1984) set out to show that the spatial properties that induced illusions and headaches were quite specific.Wilkins et al. (1984) presented participants who did not suffer from epilepsy, with a booklet of patterns having various spatial parameter measures, and instructed them to look at each print in turn and indicate which illusions they had seen from a list of 'red', 'green', 'blue', 'yellow', 'blurring', 'bending of stripes', 'shimmering', 'flickering', 'fading', 'shadowy shapes', and 'other'. It was found in the first series of experiments that a spatial frequency of 3 cycles/degree (when each stripe subtends about 10 min of arc at the eye (Wilkins, 1993)) produced the most illusions.
The length to width ratio of the patterns was also varied, to investigate the effects of check length. Results showed that patterns with longer stripes elicited more illusions than more checkered patterns. The relative width and separation of the bars of the pattern, as measured by the duty cycle, was varied to determine the duty cycle producing the most illusions. A preference was shown towards gratings with extreme duty cycles and illusions were reported most in gratings with a duty cycle of 50%, where the width and spacing of bars were approximately equal. Wilkins et al.
(1984) also demonstrated that the number of illusions reported increased sharply when the Michelson contrast was increased in the range of 5-30%.Wilkins & Nimmo-Smith (1984) suggested that text could also cause eyestrain and headaches in susceptible individuals and hypothesised that interrupting the stripes of text could reduce the induction of headaches that form the unpleasant pattern. Reading aids were developed using this rationale and participants who suffered from eyestrain and headaches were recruited to evaluate the effectiveness of such aids. Direct questioning of participants revealed that covering the lines of text above and below what was being read was a helpful aid and one third of the participants benefited from the use to the extent they planned to continue using the aid. Wilkins & Nimmo-Smith (1984) also found a tendency for participants more susceptible to visual illusion to report planned continuance of using the aid.
The parameters identified as inducing illusions can be present in text, as the lines of text and interline spacing cause the text to take on the properties of stripes (Wilkins, 1993). In a pilot study Wilkins & Nimmo-Smith (1987) revealed that participants did indeed see the visual illusions as identified in the pattern gratings in the previous experiments (Wilkins et al., 1984). Consequentially, Wilkins & Nimmo-Smith (1987) compared books of printed text to ascertain the properties that result in perceived clarity, and investigate how text could be made clearer. The participants selected samples of text judged to be clear and less clear and the spatial properties of the text were measured. The amalgamation of the results obtained across the studies suggested that the clearer text was critically dependent on larger spacing between the lines rather than the density of lettering on the page, inter-letter spacing or letter width.
Increasing the spacing between the lines also increased the duty cycle, as well as reducing the spatial frequency, critical measures in patterns
