Explore illusions.im
colour.illusions.im
Every fact

88 facts, every one sourced

Every fact on this site in one list, each linked to its source. Our build stores the source text and checks that the key words and figures in each fact really appear in it; see the sources page.

  1. 1 In February 2015 one photo split the internet: was the dress blue with black lace, or white with gold lace? WIRED
  2. 2 Michael Bach calls it a social illusion: it only gets interesting when you compare notes with other people. Michael Bach
  3. 3 It has nothing to do with colour blindness. Michael Bach
  4. 4 The colour of a non-glowing thing like a dress depends on two things: the colour of the light falling on it, and how its material reflects each colour of light. Michael Bach
  5. 5 Your visual system tries to work out the colour of the light and ‘subtract’ it. This is called colour constancy, and it usually works well. Michael Bach
  6. 6 That is why a photo taken on a glacier at noon can look surprisingly blue when you look at it at home. Michael Bach
  7. 7 Bach blames a poor photo: badly lit, overexposed on one side and showing only the dress, so the brain has too little to go on when it guesses the light. Michael Bach
  8. 8 Colour names play a part too: a bright, bluish colour counts as ‘white’ for some people and ‘blue’ for others. Michael Bach
  9. 9 Look at two stripes through small holes, cut off from the rest of the picture, and the argument mostly goes away: both sides tend to see brownish and light bluish. Michael Bach
  10. 10 It has been called the ‘Necker cube of colour vision’: two readings are about equally likely, but unlike the Necker cube it is hard to flip between them yourself. Michael Bach
  11. 11 Bach’s answer to the real colours, well lit with white light: black and blue. Michael Bach
  12. 12 WIRED’s verdict in 2015: the dress is blue. WIRED
  13. 13 Neuroscientist Bevil Conway ran the pixels through Photoshop and decided the dress was actually blue and orange. WIRED
  14. 14 Conway, of Wellesley College, suggested that people discount either the blue end or the gold end of daylight’s colours, and so see white and gold or blue and black. WIRED
  15. 15 Daylight itself changes colour, from pinkish red at dawn to blue-white at noon and back to reddish at twilight. WIRED
  16. 16 Jay Neitz, who saw white and gold, said that in 30 years of studying colour vision it was one of the biggest individual differences he had ever seen. WIRED
  17. 17 Pixels that some people saw as gold measured R 93, G 76, B 50: on their own, a kind of orange. WIRED
  18. 18 Conway joked that night owls might be more likely to see blue and black. WIRED
  19. 19 Akiyoshi Kitaoka’s famous picture shows a strawberry tartlet with a blue-green tint, and yet the strawberries still look red. Michael Bach
  20. 20 There is not a single red, or even reddish, pixel in it. Michael Bach
  21. 21 Mask off just the ‘reddest’ parts and the red mostly disappears. Michael Bach
  22. 22 Swap the strawberries for ovals of the same colours and the ovals still look reddish, so it isn’t only that we know strawberries are red. Michael Bach
  23. 23 Expectation does play some part: there is evidence that it affects colour perception to some degree. Michael Bach
  24. 24 Kitaoka made it in 2015, and Bach calls it a strong demonstration of colour constancy. Michael Bach
  25. 25 The mostly green-blue picture shifts what your eye takes as white, so grey and white areas look reddish. Michael Bach
  26. 26 In 1824 the chemist Michel-Eugène Chevreul became director of dyeing at the Gobelins tapestry works, and soon got complaints that the colours looked dull. Britannica
  27. 27 He found the problem was not chemical but optical. Britannica
  28. 28 His law of simultaneous contrast: colours placed side by side affect each other, each pushing its own complementary colour onto its neighbour. Britannica
  29. 29 He set it out in 1839 in his most influential book, The Laws of Contrast of Colour. Britannica
  30. 30 He described ways to get colour effects, such as massed dots of single colours. Britannica
  31. 31 He arranged all the colours of the spectrum in a circle to show how they relate. Britannica
  32. 32 Before all that, he patented a new kind of candle, made with stearic acid from fat, in 1825. Britannica
  33. 33 In Bach’s eye demo, a grey eye with a red tint around it looks bluish, while the identical grey eye on the other side looks neutral. Michael Bach
  34. 34 The reddish surround pushes the grey towards its opposite colour, which for red is blue-green. Michael Bach
  35. 35 Bach calls it simultaneous colour contrast, probably helped by colour adaptation. Michael Bach
  36. 36 In Ted Adelson’s checker shadow, a square in the shadow of a cylinder looks brighter than a square outside it. Michael Bach
  37. 37 In fact the two squares have identical luminance: the same amount of light comes from each. Michael Bach
  38. 38 Our eyes are not light meters: they take context into account. Michael Bach
  39. 39 Your brain estimates the lighting and judges the material: is this a light or a dark square, whatever the shadow? Michael Bach
  40. 40 Adelson published it in 1995. Michael Bach
  41. 41 In White’s illusion, identical bars look different in brightness depending on which stripes they sit on. Michael Bach
  42. 42 Michael White described it in 1979 and 1981. Michael Bach
  43. 43 Lateral inhibition, the usual explanation for contrast effects, predicts the opposite of what you see. Michael Bach
  44. 44 Mach bands: where dark ramps up to light, you see a bright band at the top of the ramp and a darker band at the bottom that are not really there. Michael Bach
  45. 45 They have been known since 1865. Michael Bach
  46. 46 They appear where a gradient of brightness, or colour, bends. Michael Bach
  47. 47 In Bach’s assimilation demo three discs behind coloured grids look cyan, magenta and yellow, but all three are white (strictly, a light yellow). Michael Bach
  48. 48 This is assimilation, first described by Bezold: each disc takes on the mixed colour of the lines crossing it. Michael Bach
  49. 49 Assimilation is the opposite of colour contrast. Michael Bach
  50. 50 Bach admits he doesn’t understand the brain mechanism behind assimilation. Michael Bach
  51. 51 One idea, from King in 1988: we see contrast when patches look like separate things, and assimilation when they look like one whole. Michael Bach
  52. 52 In the Munker illusion two stripes look green and blue, yet there are only three colours in the picture: both stripes are the same green. Michael Bach
  53. 53 It shows how much the colours around something change how we see it. Michael Bach
  54. 54 Hans Munker wrote about coloured grids in 1970. Michael Bach
  55. 55 In the watercolour illusion, shapes edged with a dark wavy line and a lighter coloured line inside look faintly tinted, though the inside is plain white. Michael Bach
  56. 56 Baingio Pinna discovered it in 1987. Michael Bach
  57. 57 Even with pale lines the illusory colour is quite strong. Michael Bach
  58. 58 Bach suspects map colourers knew about it centuries ago. Michael Bach
  59. 59 Neon colour spreading: a coloured star set inside black lines seems to glow, so it looks a little like a coloured disc. Michael Bach
  60. 60 Any gap between the coloured and black lines destroys the effect. Michael Bach
  61. 61 Bach says Van Tuijl was the first to describe it, in 1975. Michael Bach
  62. 62 Bach’s own reading list starts earlier, with D. Varin’s 1971 Italian paper on contrast and colour spreading. Michael Bach
  63. 63 It works with greys too. Michael Bach
  64. 64 An afterimage is an illusion in which the picture on your retina lingers after the thing itself has gone. Britannica
  65. 65 It can be positive, matching the original, or negative: less bright, or in the complementary colours. Britannica
  66. 66 A common one is the spot you see after a camera flash. Britannica
  67. 67 In the lilac chaser, a ring of magenta blobs circles a cross, and one blob at a time briefly disappears. Michael Bach
  68. 68 Keep your eyes on the cross and you see a green spot running round the ring. Michael Bach
  69. 69 With a really steady gaze the magenta blobs fade away completely, leaving only the green spot. Michael Bach
  70. 70 Bach notes that the opposite colour of red is not green, as many textbooks say, but blue-green. Michael Bach
  71. 71 Its inventor, Jeremy Hinton, says it shows Troxler fading, complementary colours and negative after-effects. Michael Bach
  72. 72 We normally move our eyes about three times a second, so afterimages rarely build up in everyday life. Michael Bach
  73. 73 Research has traced the afterimage to retinal ganglion cells, in the eye itself. Michael Bach
  74. 74 An afterimage normally fades within a few seconds. Michael Bach
  75. 75 Your brain joins the afterimages, one place after the next, into a single moving green disc. Michael Bach
  76. 76 The McCollough effect: after a few minutes of looking at coloured stripes, black and white stripes look faintly coloured. Michael Bach
  77. 77 Most aftereffects last a minute at most. This one can last hours. Michael Bach
  78. 78 Ten minutes of adapting is said to give an effect lasting months. Michael Bach
  79. 79 Bach warns that trying it changes your brain for a long time, up to months. Michael Bach
  80. 80 Benham’s top is a black and white disc that, when it spins, shows faint colours. Michael Bach
  81. 81 Take a photo of it spinning and the colours are not there. Michael Bach
  82. 82 Spin it the other way and the colours swap places. Michael Bach
  83. 83 In 1894 the toy maker Charles E. Benham sold his disc as an ‘Artificial Spectrum Top’. Michael Bach
  84. 84 Bach says Gustav Fechner was the first to describe the effect, in 1838. Michael Bach
  85. 85 The psychologist Richard Gregory traced it back further, to a French monk, Benedict Prévost, in 1826. Michael Bach
  86. 86 Prévost saw colours ‘like a heavenly light’ on his fingers when he waved his hands about in the cloisters. Michael Bach
  87. 87 The effect was forgotten and rediscovered no fewer than twelve times. Michael Bach
  88. 88 Nobody knows for certain why the colours appear. Michael Bach