Two Identical Grey Bars. The One on White Stripes is Darker.
Which patch is lighter?
You are looking at White’s illusion, discovered by the British vision scientist Michael White in 1979. Two identical grey bars. One sits embedded in the white stripes of a black-and-white striped pattern; the other sits embedded in the black stripes. The grey bar on the white stripes looks darker · which is the opposite of what classical simultaneous contrast would predict. This paradox broke a century of contrast-based theorising and kicked off a new chapter in brightness research.
What you are about to learn. What White’s illusion actually is, why it is a puzzle for classical contrast theory, the T-junction and grouping-based accounts that eventually explained it, why it matters for modern theories of brightness perception, and how it changed the way vision scientists talk about contrast.
What the Illusion Looks Like
Draw a regular pattern of black and white horizontal stripes. In one region, replace a segment of white stripe with a mid-grey. In another region, replace a segment of black stripe with the same mid-grey. Place the two regions side by side.
The grey bar on the white stripes reads darker. The grey bar on the black stripes reads lighter. The effect is strong · typically 15 to 20 percent in perceived brightness. And it runs in the opposite direction from what simultaneous contrast would predict.
The paradox. Classical contrast theory says: if a grey patch is mostly adjacent to white regions, it should look darker (subtract surround). If mostly adjacent to black regions, lighter. In White’s illusion, the grey bar on white stripes is partly flanked by black (above and below, the next stripes over) as much as by white, and similarly for the black-stripe bar. The naïve contrast calculation predicts a roughly neutral result, or even a small reversed effect. Instead, we get a strong illusion in what looks like the “wrong” direction.
Why It Works
The modern account, developed over the 1990s and 2000s, identifies the mechanism as perceptual grouping at T-junctions.
T-junctions are where a line ends against another line. In White’s figure, every place where the grey bar terminates, it forms a T-junction with the surrounding stripe. These T-junctions carry geometric information · they tell your visual system “this grey region belongs to this stripe, not that one”.
Grouping follows the T-junctions. The grey bar on a white stripe is grouped with the white stripes · your brain reads the grey as “a dimmer continuation of the white stripe”. The grey bar on a black stripe is grouped with the black stripes · read as “a lighter continuation of the black stripe”.
Brightness assimilation follows grouping. Once the grey bar is grouped with the white stripes, its perceived brightness assimilates toward the white · but relative to the white, it reads as dark. The relative-to-neighbour reading dominates. The grey bar grouped with the black stripes does the reverse.
This is assimilation, not contrast. White’s illusion is the flagship example of brightness assimilation · where a patch is pulled toward its perceptually-grouped neighbour rather than pushed away from it. The classical theories (lateral inhibition, centre-surround contrast) only account for push-away. White’s forced the field to accept that pull-toward is also a real mechanism. Modern theories of brightness perception integrate both.
Why White’s is a Big Deal
Before 1979, most brightness illusions (simultaneous contrast, Hermann grid, Mach bands) were explainable by lateral inhibition alone · a purely retinal, bottom-up mechanism. White’s illusion does not fit the pattern. Explaining it required invoking higher-level perceptual organisation · edges, junctions, grouping · that lateral inhibition knows nothing about.
The paradigm shift. The 1980s and 1990s saw an explosion of brightness-perception research specifically trying to reconcile White’s illusion with the older contrast-based framework. Models that could handle both · notably the ODOG (oriented difference-of-Gaussians) model, and later the grouping-based models of Gilchrist and colleagues · became the new baseline. White’s illusion is, in that sense, the figure that forced vision science to take perceptual organisation seriously in brightness perception.
A Clean Variant: Remove the T-Junctions
Here is a diagnostic experiment. If T-junctions are doing the work, then removing them should kill the illusion.
The junction-removal test. Replace the sharp stripe edges with smooth gradients. The grey bars now do not form T-junctions with clear bright/dark regions · they fade into a gradient. The illusion weakens dramatically. Now reinstate the sharp edges but add a thin gap between the grey bar and the stripe ends. The T-junctions are gone. The illusion weakens again. These controls are the strongest evidence that it is the T-junction geometry, not just the neighbour luminance, that drives White’s.
The ODOG Model
Anthony Blakeslee and Mark McCourt’s 1999 ODOG (oriented difference-of-Gaussians) model handles White’s illusion by postulating that brightness perception uses multiple orientation-tuned filters, not just an isotropic centre-surround. The key insight: because the stripes are oriented, a horizontal-stripe pattern activates the horizontal-orientation filters most strongly, and those filters have much weaker response to the grey bars (which are also horizontal and aligned with the stripes). The brightness calculation then comes primarily from vertical-orientation filters, whose response is dominated by the nearest stripe colour · which is the stripe the grey bar is embedded in.
Common misconception: “White’s illusion is just simultaneous contrast in disguise.” It is not. Genuine simultaneous contrast predicts the grey on white looks darker because of the total surround, including the black stripes flanking the grey bar vertically. White’s illusion magnitude is too big for that to be the full story. Controlled experiments with single-stripe and multi-stripe variants show that the direction and strength of the effect depend critically on how the grey groups with the surrounding stripes · a property that classical contrast theory has no machinery to represent.
A Harder Variant
Below is a White’s illusion figure at difficulty 3 · more stripes, sharper contrast. The grey bars are identical pixel values.
Which patch is lighter?
Cover adjacent stripes. Hold a piece of paper so it covers the stripes to the left and right of the grey bars, leaving only the immediate stripe each bar is embedded in. The two greys now appear much more similar · you have stripped away the context that was doing the grouping. Lift the paper and the illusion snaps back immediately. This is a live demonstration that the illusion depends on the extended stripe context, not just the nearest neighbour.
Where White’s Illusion Appears
- Textile design. Striped fabrics can carry coloured threads that, depending on which stripe the coloured thread is embedded in, look radically different. Designers working with complex striped patterns learn to compensate · or to exploit the effect deliberately.
- Architecture. Building facades with alternating light and dark horizontal bands (a common modernist motif) show the White mechanism wherever a window or panel is embedded. A grey panel in the light band looks darker than the same panel in the dark band.
- Typography on striped backgrounds. Text overlaid on a regularly-striped background (rare but visible in some magazine spreads) picks up a White-style brightness shift depending on which stripe each letter’s pixels land on. This is one of the reasons typographers generally discourage placing text over striped patterns.
- Visual cortex modelling. White’s illusion remains a benchmark test for computational brightness-perception models. If your model cannot produce the White result in the right direction, your model is incomplete.
Test Yourself on 50 More Illusions
White’s illusion is one of more than 50 classical illusions on PlayMemorize. Each round draws a deterministic SVG scene and asks one grounded question: which is larger, which is brighter, which is actually parallel. The reveal overlay shows the true geometry plus a one-line “why it works” caption.
- Keep playing White’s → · the standalone game, pinned to this one figure with fresh seeds each round
- Play Illusions → · spot the tricks across size, colour, orientation, and impossible figures
- Play Spatial → · train mental rotation and area estimation
- Play Matrix → · abstract pattern reasoning under time pressure
The takeaway. White’s illusion is the reason vision science no longer talks about brightness perception as a single, low-level mechanism. It is a reminder that what looks like a simple grey square is actually the output of multi-layered computation involving edges, junctions, groups, and oriented filters. When you see a grey bar on a stripe and cannot help but read it relative to that stripe, you are looking at perceptual organisation and brightness perception acting as one system. Michael White’s 1979 figure broke the old theory. The new theory is still being refined.
Illusions
Your eyes lie - the math knows the truth. Spot equal lengths, identical greys, and truly parallel lines across 57 classic optical illusions
العب الآن - مجاناًلا حاجة لحساب. يعمل على أي جهاز.
