Illusion Test Results

A record of which illusions I experience and which I don't, organized by the level of visual processing they're generally associated with. This is self-reported and not clinically verified — I'm recording it as honestly as I can, including results that don't fit a clean pattern.

Low-level

Retinal / early visual cortex — brightness and contrast processing. These are generally attributed to local, early-stage processing, before object recognition or interpretation comes in.

Simultaneous contrastIdentical grays on different backgrounds

I see it

Koffka ringRing split across two backgrounds

I see it

Ordinary afterimagesLingering image after looking away

I see them

High-level

Object recognition, contour integration, predictive processing. These are generally attributed to higher-level interpretation — the brain's predictions about objects, faces, depth, and scale.

Kanizsa triangleIllusory triangle from notched circles

I don't see it

Ponzo illusionTwo equal bars between converging lines

I don't see it

Ebbinghaus illusionCenter circles ringed by different sizes

I don't see it

Moon illusionMoon looks larger near the horizon

I don't see it

Hollow-Face illusionRotating concave mask

Complex

On first viewing, I correctly see the concave (hollow) side as hollow. If I keep watching the same mask rotate repeatedly in one sitting, it starts to look convex partway through. After a break it resets — the next time I look, I see the hollow side correctly again, until repetition triggers the illusion once more.

Gestalt collapseA character stops looking coherent

Doesn't happen

Because I don't think I perceive the character as a unified whole in the first place, there's nothing to "collapse."

Rubin's vaseFigure-ground reversal: vase / two faces

I see both at once

I don't experience the usual one-or-the-other switching — both the vase and the faces are present at the same time.

Needs further testing

Results I can't draw conclusions from yet, because the conditions weren't properly controlled.

Color afterimageInverted-color photo, then blank surface

Inconclusive

Stayed black-and-white, never turned to color — unlike the ordinary afterimages above. Conditions (gaze duration, image contrast, viewing surface) weren't controlled, so this needs to be retested properly before drawing any conclusions.

None of this is offered as proof. It's a record, kept as honestly as I can manage, including the results that don't fit the pattern — the Koffka ring and simultaneous contrast, in particular, behave the way most people report, not the way the rest of this list would predict.