Influences of orientation on the Ponzo, contrast, and Craik-O'Brien-Cornsweet illusions

Abstract

Explanations of the Ponzo size illusion, the simultaneous contrast illusion, and the Craik-O'Brien-Cornsweet brightness illusions involve either stimulus-driven processes (assimilation, enhanced contrast, and anchoring) or prior experiences. Real-world up-down asymmetries for typical direction of illumination and ground planes in our physical environment should influence these illusions if they are experience based, but not if they are stimulus driven. Results presented here demonstrate differences in illusion strengths between upright and inverted versions of all three illusions. A left-right asymmetry of the Cornsweet illusion was produced by manipulating the direction of illumination, providing further support for the involvement of an experience-based explanation. When the inducers were incompatible with the targets being located at the different distances, the Ponzo illusion persisted and so did the influence from orientation, providing evidence for involvement of processes other than size constancy. As defined here, upright for the brightness illusions is consistent with an interpretation of a shaded bulging surface and a 3D object resulting from a light-from-above assumption triggering compensation for varying illumination. Upright for the Ponzo illusion is consistent with the inducers in the form of converging lines being interpreted as railway tracks receding on the ground triggering size constancy effects. The implications of these results, and other results providing evidence against experience-based accounts of the illusions, are discussed.

Keywords: Adjustment method; Craik-O’Brien-Cornsweet illusion; Orientation; Ponzo illusion; Simultaneous contrast.

Fig. 1

(A) The Ponzo illusion as originally demonstrated. The two horizontal lines have the same length but most observers perceive the upper one as longer. (B) Brightness contrast. Top: The two inset squares have the same luminance but the one embedded in a dark background is by most observers perceived as slightly brighter. Bottom: The luminance gradient providing a hint of a penumbra enhances the contrast illusion. (C) The Cornsweet illusion. Top: The original form. Bottom: As typically demonstrated with a horizontally oriented rectangular display, similar to the cut-out region of the original display with enhanced contrast. The left surface seems brighter than the right surface although they are same, which can be confirmed by covering the middle contrast with a pencil

Fig. 2

The illustrations show various shading scenarios given the light-from-above assumption. (A) The leftmost image is consistent with the 3D structure shown in the inset where the upper part is directly illuminated and the lower part is shaded. The rightmost image is an upside-down version of the left image and is consistent with the middle section being painted rather than shaded due to bulging (or as a sharp edge bulging outwards). The gray scales of the upper and lower parts of the surface now look more similar (from Purves et al., . Copyright 1999, Society for Neuroscience). (B) The crater illusion; top-most image is an upside-down version of the bottom image. (C) Direct illumination occurs more frequently on top parts of objects than bottom parts, which more frequently receive indirect illumination. (D) When the shadow is due to cast shadows it is not so obvious how the shadowed and directly illuminated areas are likely to be oriented relative to each other. Still, combining these two causes of luminance variations would result in a greater frequency of occurrences of brighter regions on top. All discs in (C) and (D) have the same luminance, but the ones in the shadowed regions usually appear brighter

Fig. 3

Stimuli used in the experiments, all shown with 0° orientation. (A) The Ponzo-illusion: the circular discs are the same size but people tend to perceive the upper one as larger. (B) Brightness contrast: The two small discs are equally bright but most people perceive the upper one as darker. (C) The Cornsweet illusion: The surface labeled a seems to be darker than the surface labeled b, although they are same

Fig. 4

Cross-section schematically showing the luminance profile of the Cornsweet illusion (top), and how the setting of luminance level of the adjustable side to obtain subjective equality with the fixed side influenced the slope of the luminance gradient on the same side

Fig. 5

Displayed are all stimulus orientations used for the three illusions. From top to bottom row: Simultaneous contrast, Craik-O´Brien-Cornsweet, and the Ponzo illusion

Fig. 6

The radar plots show the mean illusion magnitude expressed in proportions as a function of orientation in degrees from 0° to 315°. A positive sign indicates that the illusion is in the expected direction. Presented along the cardinal orientations (0°–180° and 90°–270°) are illustrations of stimuli. The shaded areas represent the 95% confidence intervals

Fig. 7

The average correlations between PSE (illusion strength) as a function of the absolute difference between orientations for each illusion. The 95% CIs are displayed

Fig. 8

Displays used in Experiment 2 showing the left-glare and right-glare conditions for the two illusions, here oriented 90° in all four displays. The two right-glare panels at the bottom are consistent with an interpretation of shading resulting from 3D shapes. The two left-glare panels on top are less consistent with a similar interpretation

Fig. 9

Illusion magnitude as a function of stimulus orientation and direction of incoming light from the simulated glare on the screen. The 95% CIs are displayed

Fig. 10

The radar plots show the mean illusion magnitudes for each of the orientations in the no-glare conditions. The shaded areas represent the 95% CIs

Fig. 11

The Ponzo illusion still manifests itself although the converging lines of the silhouette of the cones (partly illusory) are not perceived to recede in depth. The horizontal line segments are perceived at the same distance from the observer. For most people, although reduced, the illusion persists when turning the display upside down