Impossible objects

An optical illusion is an image or scene that creates a difference between its physical features and the way it is perceived. Context, perspective, contrast and visual adaptation can all contribute to the effect.

Vision interprets light using surrounding shapes, colours, depth cues and prior experience. Illusions reveal situations in which these normally useful processes produce surprising interpretations. Different illusions involve different mechanisms.

Through Mirror includes size and geometry illusions, light and colour effects, ambiguous figures, impossible objects, motion illusions and hidden contours. The collection also explores illusion art, architecture, fractals, animal camouflage and other senses.

Open an entry for its viewing instructions, explanation, history and sources. Gallery images can be enlarged, reversible pictures can be rotated, and animations play when selected. Some demonstrations link to their original source.

An architectural illusion changes the apparent shape, depth or extent of a real place through painting, perspective, geometry or reflection. Its effect often depends on where the observer stands.

The cathedral’s famous apparent dome is a painted illusion by Antonino Emanuele, dating to 1739. The painting makes a flat ceiling appear to rise into a dome from the intended viewpoint.

Forced perspective changes scale and alignment to suggest a different depth or size. Borromini’s corridor at Palazzo Spada uses a compressed construction that appears longer from its intended viewing position.

The collection explores Gary Fisher’s Optical Interactions as a laser transmission example, Matthew Schreiber’s Cassadaga (Spirit Pond) as a reflection hologram, and Stephen Benton’s Sphere as an early white-light transmission example. Diagrams explain each method, with links to the artists’ archives or original exhibition catalogue.

Optical holography encodes and reconstructs a light wavefront using an interference pattern. Pepper’s Ghost and Musion Eyeliner instead reflect an image into the audience’s view, making it appear to occupy the stage. Both can create a convincing presence, but their optical mechanisms differ.

A row of LEDs sits between a rear mirror and a partially reflective front panel. Repeated reflections create successive virtual images that appear farther away. Each loses light, so the apparent tunnel fades even though the frame is shallow.

Musion documents Gorillaz at the November 2005 MTV Europe Music Awards in Lisbon and lists the band’s 2006 Grammy Awards appearance in its portfolio. Eyeliner uses angled transparent foil and video imagery to create the stage illusion.

Projection mapping aligns light with a real object or building. Content that adds convincing perspective, shading or motion can turn that surface into a trompe-l’œil: it seems to gain depth or change shape. URBANSCREEN’s 555 KUBIK is an architectural example.

Her first appearance alongside Gorillaz was prerecorded footage presented through Musion’s reflected-video system. Musion director James Rock described filming it three weeks before the show. The performance then moved to Madonna appearing live with her dancers, so it combines a virtual appearance with a real performer.

Camouflage changes how an observer detects or recognises an organism or object. It includes matching a background, breaking up an outline, masquerading as something else, borrowing cover and confusing judgments of shape or direction.

No. Plants such as Lithops resemble stones. People use camouflage in clothing, face paint, netting, disguised structures and painted ships. Some designs also challenge how computer vision models detect people or faces.

Background matching reduces the visual contrast between an object and its surroundings. Masquerade makes an object resemble something the observer may ignore, such as a leaf, a stone or a tree.

Dazzle patterns were intended to confuse judgments of a ship’s form and course while it remained visible. Their historical effectiveness is difficult to establish; they are different from background matching.

No. Visibility depends on the background, light, distance, movement and the observer’s sensory abilities. A photograph illustrates appearance but does not establish effectiveness in every setting.

Face-like prints can act as decoys for a specific face detector, as explored by HyperFace. The original prototype used a scarf and targeted an older OpenCV detector. Producing false face detections does not prove that the real face is missed or that its identity cannot be matched. A generic face-pattern shirt has no established protective effect.

HyperFace adds face-like decoys around a person. Adversarial shirt research designs prints to change a person detector’s decisions; those prints need not resemble faces. Some studies use a chest patch, while others cover shirts, skirts and dresses with an all-over texture.

Face detection locates a face in an image. Face recognition compares a face with stored identities. Person detection locates the whole person. Confusing one of these tasks does not by itself establish that the other tasks have failed.

No universal effect is established by these examples. Results depend on the target model, camera viewpoint, movement, lighting and fabric deformation. The original HyperFace and CV Dazzle looks targeted older face detectors. The diagrams here explain the ideas and are not tested camouflage patterns.

A fractal is a mathematical structure with detail or related patterns across different scales. Some fractals repeat an exact construction; others show more varied forms of self-similarity.

Not necessarily. Through Mirror includes fractals because their repetition, intricate boundaries and nested patterns challenge visual intuition. Their geometry does not have to create a perceptual error.

The Mandelbrot set is defined by repeatedly applying a mathematical rule to complex numbers and checking whether the results remain bounded. Images show its intricate boundary and related patterns at different scales.

Trompe-l’œil means deceiving the eye. In painting, carefully arranged perspective, shading and familiar objects can make a flat surface appear to contain real depth or objects.

Anamorphosis is a distorted image that becomes recognisable from a particular viewpoint or with an appropriate reflective surface. Holbein’s The Ambassadors includes a skull revealed by an oblique view.

Arcimboldo arranged recognisable objects such as fruit, flowers and books into the features of a face. Some works also reveal a second interpretation when the painting is turned upside down.

Geometric illusions change the apparent size, length, angle or alignment of shapes. Surrounding lines and perspective cues can make equal lengths look unequal or parallel lines appear tilted.

Yes. The horizontal boundaries are parallel, but the alternating arrangement of light and dark tiles makes them appear to slope.

Follow the entry’s instructions, then compare the relevant lengths with a ruler or cover the surrounding context. This can separate the physical measurement from the visual impression.

The visual system judges a surface in relation to nearby contrast and apparent lighting. The same image value can look lighter or darker depending on its surroundings.

Yes. The labelled squares have the same grey value. The depicted shadow and checkerboard context make them appear different; isolating the squares reveals the match.

Mach bands are apparent light and dark strips near changes in brightness. They illustrate how the visual system responds to contrast at boundaries.

A colour afterimage is a lingering visual impression after looking away from a coloured stimulus. Adaptation in the visual system can make a different colour appear against a neutral background.

Colour perception depends partly on surrounding colours and contrast. An unchanged patch can look different when placed in a different visual context.

Use the viewing instructions on the entry page. Look at the specified point, then shift to the suggested neutral area. You can stop whenever you wish; the effect varies between viewers.

An ambiguous figure supports more than one interpretation of the same image. A viewer may switch between the interpretations even though the picture stays unchanged.

The same outline can be read as a duck facing one direction or a rabbit facing the other. The shape that looks like a beak can also look like ears.

The central region can be seen as a vase, while the surrounding regions can be seen as two facing profiles. The shared boundary changes meaning as foreground and background switch.

An impossible object is a drawing whose parts suggest familiar three-dimensional shapes but cannot all connect as they appear to. The contradiction becomes clearer when you trace the whole structure.

It cannot exist as the ordinary solid the drawing suggests. A specially arranged model can create the appearance from one viewpoint, but its actual structure differs from the pictured object.

The drawing joins plausible stair segments into an apparent loop that continually rises or falls. In ordinary space, a closed circuit cannot gain height at every step and return to its starting level.

Certain arrangements of contrast and pattern can produce apparent movement, especially away from the centre of attention. A motion illusion does not always require an animated image.

No. The silhouette has ambiguous depth cues and can be interpreted as turning in either direction. Its perceived rotation does not establish a personality type or a dominant brain hemisphere.

No. Through Mirror initially shows still previews. Open a demonstration and select Play animation when you want to see it move.

Illusory contours are apparent edges that are not explicitly drawn. Surrounding shapes can make a viewer perceive a continuous boundary or a surface.

The bright triangle appears to have clear edges, but those edges are not drawn. The surrounding cut-out shapes suggest a triangle covering parts of them.

Vision uses the relationships between visible edges and surfaces to interpret partially hidden objects. Illusory-contour demonstrations show how a coherent shape can be perceived from incomplete cues.

An auditory illusion is a surprising perception of sound influenced by context, arrangement or interpretation. Examples include ambiguous melodies, apparent changes in pitch, and the virtual barber shop’s impression of a haircut happening around you.

Entries explain the effects and provide listening demonstrations. For Virtual Barber Shop, use stereo headphones with the left and right sides correctly placed, then play the embedded demonstration.

Spatial sound gives audio a sense of direction and distance in three-dimensional space. Binaural audio delivers this through two headphone channels that reproduce cues reaching each ear, including timing, loudness and filtering by the head and outer ears. The virtual barber shop uses these cues to place a barber around your head. Binaural beats are a different effect: nearby tones presented separately to each ear create a perceived beat.

A tactile illusion is a surprising experience of touch or bodily sensation. Adaptation, neighbouring stimulation, body position and expectations can influence what is felt.

Some demonstrations explore touch alone, while others involve interactions between touch and vision. The entry’s instructions explain which senses and conditions are involved.

A multisensory illusion involves interactions between senses. What a person sees, hears or feels can change how information from another sense is interpreted.

The McGurk effect occurs when seeing speech movements influences the speech sound a listener perceives. Combining mismatched visual and auditory speech can produce a different perceived syllable.