Author Archives: Stuart Anstis

About Stuart Anstis

Professor of Psychology at UC San Diego. Ph.D. from Cambridge, supervised b Richard Gregory. Published about 150 papers on visual perception, especially illusions. A Humboldt Fellow, a visiting Fellow at Pembroke College, Oxford, and recipient of the Kurt Koffka Medal for outstanding contributions to vision research.

El Greco

Was El Greco astigmatic?

Why did El Greco (1541-1614) paint such elongated figures?  Could he have suffered from a visual astigmatism that optically stretched his visual field?  Art historians doubt it, and logicians ague that this is a fallacy because any visual defect would affect sitter and painting equally and would cancel out.


I converted a volunteer into an ‘artificial El Greco’ with an experimental telescope that expanded the world horizontally.

When asked to copy a square, she drew an exact square copy, but when asked to draw a square from memory, she drew a tall, El Greco-style rectangle. This might suggest that El Greco’s portraits from life would be normal, but his portraits from memory would be elongated. However, the volunteer adapted over two days to the visual distortion; a series of her drawings of a square from memory gradually became perfectly square. So even an astigmatic El Greco could have painted in normal proportions if he chose. His elongations arose from his mannerist style, not from defective vision.

Motion-Based Position Shift with Patrick Cavanagh

    What do you see?

[quicktime width=”600″ height=”400″][/quicktime]

Motion undershoot. Bar rotates through 180°, from 12 to 12 o’clock.  But it appears to move only from 1 to 11 o’clock.

[quicktime width=”600″ height=”400″][/quicktime]

Red and green dots are in identical positions, vertically oriented. But when flashed at the motion endpoints of the rotating ring, they appear to be offset.

[quicktime width=”600″ height=”400″][/quicktime]

Same as the ring but for linear motion. Red and green bars are in the same position but appear to be offset. Try tracking them with your eyes; your eyes feel as if they move, but they really don’t!!

[quicktime width=”600″ height=”400″][/quicktime]

Same idea here! Right-angled cross looks wonky because moving sector edges shift the cross arms more than moving middles of sectors.

Reverse Phi

 Do you see what I see?

[quicktime width=”350″ height=”350″][/quicktime]

The four spots move back and forth in exact synchrony, in the direction shown by the arrow.  The two upper spots are correctly seen as moving in the direction of the arrow.  However, the two lower spots change their polarity between black and white as they shift.  These are perceived as moving backwards, toward the earlier stimulus and opposite to the true displacement.  This is reverse phi.  It is consistent with Ted Adelson’s motion energy model (JOSA 1985).

[quicktime width=”300″ height=”300″][/quicktime] [quicktime width=”300″ height=”300″][/quicktime]

Both movies are identical and both rotate clockwise.  But in the right movie the dots are black and white on alternate frames, and appear to rotate counterclockwise.  This is reverse phi (Anstis 1970: Anstis & Rogers 1975), in which the motion energy does go counterclockwise.
Gaze at the centre of each movie for 20s, then stop the movement.  Which way does the movement aftereffect go?  CCW in the left-hand movie of course.  But CW in the right-hand movie, appropriate to the perceived motion direction, not to the physical dot displacements.
This shows that reverse phi does adapt neural motion detectors; possibly in brain area MT (V5).

[quicktime width=”600″ height=”400″][/quicktime]

In this reverse phi movie, made by PATRICK CAVANAGH, the spokes reverse their polarity on every movie frame.  Thus the inner ring actually steps counterclockwise (track a spoke with your eyes to check this) but it seems to rotate clockwise.  The opposite is true for the outer ring.  Adapt to the motion for 20s, then stop the motion.  In the motion aftereffect, the outer ring appears to move CW and the inner ring CCW — appropriate to the illusory reverse phi, not to the physical displacement.

Reverse Phi: Four-Stroke Cycle

Scroll down and click on each four-stroke cycle movie. The objects appear to move continuously without changing their average position.

[quicktime width=”600″ height=”400″][/quicktime]
[quicktime width=”600″ height=”300″][/quicktime]
Vertical movement
[quicktime width=”600″ height=”300″][/quicktime]
Horizontal movement
[quicktime width=”600″ height=”300″][/quicktime]

Each movie is four frames long, in the sequence positive-positive-negative-negatives.

All Kinds of Motion

When the black and white bars switch places, on a dark surround (left) the white bar appears to jump, but on a light surround (right) the black bar appears to jump. The bar with the higher contrast wins out. The mid-grey at which the motions balance is the arithmetic (not geometric) mean of the black & white, suggesting linear, not logarithmic processing of luminance. (Anstis & Mather, Perception 1986).

[quicktime width=”500″ height=”400″][/quicktime]
Ambiguous apparent motion. The two spots move either vertically or horizontally. Can you control the direction by willpower?

[quicktime width=”600″ height=”400″][/quicktime]
Proximity: Motion is seen between nearest neighbors, horizontally on the left, vertically on the right. Shorter motion paths win out.

[quicktime width=”600″ height=”400″][/quicktime]

The motion path changes gradually from a tall, skinny rectangle to a wide, flat rectangle. Perceived motion is always along the shorter side of the rectangle. Proximity wins.

[quicktime width=”600″ height=”400″][/quicktime]

Visual inertia drives ambiguous apparent motion. Each spot appears to follow a horizontal path, not jumping up or down halfway across. Straight motion paths are preferred to going round corners.

[quicktime width=”600″ height=”400″][/quicktime]

Do these all move together or do they move individually?

[quicktime width=”600″ height=”400″][/quicktime]

The center dot simply flashes on and off but it gets entrained by the other dots and seems to disappear and reappear from behind the green square. [V.S. Ramachandran]

Kinetic Edges

[quicktime width=”500″ height=”400″][/quicktime]

Although the three windows are actually aligned vertically, the central window appears shifted to the left or right, in the direction of the drifting dots that it contains.

[quicktime width=”500″ height=”400″][/quicktime]

Eight circular windows, arranged in a circle, contain random dot textures that move counterclockwise. Although the windows themselves are not moving, they appear to rotate together like a ferris wheel. This is a stronger version of the illusion demonstrated above — it introduces continuous illusory movement, not just a static illusory shift. Also, after fixating for a while, you may perceive the windows fade out and disappear.

Binocular Brightness

When two eyes see different grays:

Free fuse the two columns of gray squares in (a), so that the left eye sees column L and the right eye sees column R (or vice versa: it doesnt matter), to give a single vertical column of gray squares. Note the perceived brightness of the squares. On a light background (a), the middle square (arrowed) is the same to both eyes and probably looks lighter than the squares above and below. However, on a black background (b) the middle square probably looks darker than those above and below. Squares in (a) and (b) are actually identical, only the backgrounds differ.

What s going on? The stimulus presents a light square to one eye and a dark square to the other eye, so that the two luminances always sum to a constant. So if your eyes (and your web browser) were linear, all fused squares would look the same brightness. However, the visual system is non-linear, and systematically overweights the square with the higher contrast (not luminance), favouring dark squares on a light background in (a), and favouring light squares on a dark background in (b). Careful measurements have shown that the weighting function is quadratic for light squares (spatial increments) but square, or winner take all, for dark squares (spatial decrements), as shown in the graphs.

Adaptation to Flicker with Debbie Giaschi

Adaptation to Flicker and Spatial Frequency

[quicktime width=”500″ height=”400″][/quicktime]

Run the movie & fixate the red cross. Both gratings are the same, but following adaptation to spatially uniform flicker, the upper grating looks apparently finer. Reason: Adaptation of transient pathways that are tuned to high temporal and low spatial frequency.