The shadow blister effect is the appearance, when two shadows fall close to one another, that one of them bulges or “blisters” outward toward the other just before they touch. It looks like an optical illusion and is often described as one, but the bulge is really there: it is a genuine feature of the light distribution, produced by the overlap of the two shadows’ penumbrae when the light source is extended rather than a point. The effect was analysed rigorously by James A. Lock in Applied Optics in 1998, and his ray-theory treatment explains it without appealing to diffraction, nonlinearity or the quirks of the eye.[1][2]
What you actually see
Set up two objects between a broad light source and a wall, with one closer to the wall than the other, and slide them so their shadows approach each other. Before the shadows meet, the shadow of the object nearer the wall develops a rounded protrusion pointing at its neighbour, as though something were pushing out from inside it. The neighbouring shadow does not bulge in return. When the shadows finally merge, the protrusion has done its work and the effect disappears into a single dark region.[3]
Why it happens
An extended source is not a point, so each object casts two distinct regions: an umbra, where the source is completely hidden, and a penumbra, where it is partly hidden and the illumination falls off gradually. The nearer an object is to the wall, the narrower its penumbra; the nearer it is to the source, the wider. In the classic case, object a sits close to the wall with a narrow penumbra and object b sits further back with a wide one.[4]
The mechanism becomes clear if you imagine walking across the wall, from the lit side into the shadows, and asking at each point how much of the source you can see. In the absence of b, you would lose sight of the source’s edge at a particular position, and that position would mark the boundary of a’s umbra. But where the penumbrae overlap, b is still blocking that same edge of the source. You have to walk further before any part of the source becomes visible past both objects. The umbra of a therefore extends further toward b than it otherwise would, and that extension is the blister. Lock’s phrasing is precise: the continued obscuration of the source by b, at positions past where the edge of a’s umbra would fall in b’s absence, causes the apparent bulge.[5]
The shadow is not bending toward its neighbour out of any affinity between them. It is simply that the second object is standing in the way of the light that would otherwise have revealed the first object’s edge.
The conditions that matter
Lock derives the geometry and identifies when the effect is strong. It vanishes if the two objects are the same distance from the screen, because their penumbrae then behave symmetrically. It fades if their separation along the light path is very large, because the near object’s penumbra becomes too narrow for the overlap to have any visual weight. It is most pronounced when the longitudinal separation is comparable to the distance from the objects to the screen — in Lock’s notation, when x/R is of order one. The bulge also grows with the angular size of the source, with the distance from object to screen, and with smaller objects, since the blister is then a larger fraction of the shadow it deforms.[6]
The laboratory check
The paper is not only theoretical. Lock built the effect on a bench using a long wooden board with a straight edge and a cardboard disc of 13 cm diameter, lit by a 40 W tubular bulb with a 10 cm filament that approximates a line source, and photographed the shadows with the objects placed at 145 cm from the source, 70 cm apart, and 100 cm from the screen. That arrangement gave a source of 1.8° angular diameter at the screen — more than three times the Sun’s 0.5°, and therefore a wider penumbra and a more obvious blister. Theory predicted an aspect ratio for the bulge of 0.125 at the moment the umbrae touched; measurement of a contrast-enhanced photocopy gave 0.125 ± 0.005. The effect behaved exactly as ray optics said it should, including reversing its direction when the disc was moved nearer the screen than the board.[7]
What it is not
The effect is often explained as diffraction, or as a nonlinear response of the eye, or simply as two penumbrae adding up to something darker that the brain misreads as shadow. None of these is needed. Ray theory alone accounts for the geometry and predicts the size of the bulge, which a diffraction or contrast explanation could not do quantitatively.[8]
Perception does, however, amplify what is physically there. In casual viewing people largely ignore penumbrae and treat the visible shadow as though it were the umbra, so a change in the umbra’s boundary reads as the whole shadow changing shape. And the narrow lit gap between two nearly touching umbrae is perceived as brighter than it measurably is, which sharpens the apparent separation and makes the bulge more conspicuous. The effect is therefore a real optical phenomenon that the visual system presents more forcefully than the raw intensities warrant — a useful reminder that “illusion” and “real” are not always opposites.[9]
How to see it
Lock gives two demonstrations that need nothing but sunlight. Indoors, with a low sun coming through a window, stand in the middle of the room and move your head until its shadow on the opposite wall almost touches the edge of the window frame’s shadow; your head’s shadow will appear to bulge toward the frame. Outdoors, stand a few metres from a tree trunk or the edge of a wall on a sunny day and move until the shadow of your head nearly touches the shadow of the trunk or wall on the ground. Both work because standing roughly midway between the aperture and the wall puts the geometry in the regime where the effect is strongest.[10]
Why it is worth knowing
Three audiences care about this. Vision scientists care because shadow blur is a depth cue: a shadow with a soft edge tends to look nearer than one with a hard edge, an observation that goes back to Hering in the nineteenth century, and the blister effect shows how much geometric information the visual system can extract from penumbrae that most people never notice.[11][12]
Photographers and compositors care because they spend their working lives trying to make synthetic shadows behave like real ones, and a renderer that ignores penumbra overlap will produce shadows that do not merge the way a viewer expects. And teachers of optics care because the effect is a clean, cheap demonstration that an extended source is not a point source, that umbra and penumbra are different things, and that a great deal of classical optics can be done with a lamp, a disc and a wall.[13]
Select a year and month above to explore cold archived discussions.
Please select an archived topic above to view the discussion logs.