Relativity (detail). Lithograph print, MC Escher, 1953

Everyone loves an optical illusion, right? Remember how the Internet exploded with the blue/gold dress? Sure you do…

Well I’m not going to, um, address that here (pro-tip: it’s about our differing ways of compensating for ambient light, or ‘white balance’ as we photographers call it) but I am going to look at some other popular illusions and explain why they’re not really illusions at all. Kind of.

We can put optical illusions into a few different boxes: those that come about through perspective distortion, wherein our brains struggle to resolve apparent inconsistencies in depth or space by way of distorted vanishing points and converging parallels. I have to insert a ditty here from my undergraduate days which is of no particular relevance…

‘Parallel Lines meet at infinity’
Euclid repeatedly, heatedly urged.
But then when we died and reached that vicinity
He found that the damn things diverged!

Anyway, in this box, we could put the enduring, if over-exposed work of Dutch artist, MC Escher and his impossible drawings of concurrently ascending and descending stairs, and his endlessly one-sided ‘Mobius Strips’.

Another category would be three-dimension illusions which use distortions in perspective to convince us we see greater or lesser depth than is actually present in the scene. A famous example is the Ames room, after Adelbert Ames, Jr. wherein we appear to see a rectilinear space but which is in-fact a ‘polyhedron’ whose far side is not parallel to the observer.

AmesRoom

The Ames Room, invented by American ophthalmologist Adelbert Ames, Jr. 1934

A sub-category of these would be those amazing 3-dimensional amazing pavement drawings which come to life when viewed at the right angle, fooling our minds into believing the road is being consumed by a flaming dragon or revealing a portal to Hell. Etc. 

Luminance and Tone

Of more interest to my research though are those that involve colour, and among the best-known of these is neuroscientist, Edward Adelson’s checkerboard illusion. Adelson’s image is constructed of dark and light-grey squares whose shades differ but not quite in the way we think they do. The cylinder appears to cast a shadow over part of the board. On closer inspection, we see that the square marked ‘B’ which we perceive to be lighter in tone to that of ‘A’, is in-fact precisely the same shade of grey.

Adelson Adelson Solved

Even when employing the artist’s trick of squinting at the image through half-closed eyes, it is hard to convince ourselves that this is true. And yet, when the squares are ‘joined’ by a rectangle of constant tone, it becomes clear they are indeed the same.

StripedChestTrans

As with the Adelson illusion, the grey bands of this object appear to differ in a manner which is optically misleading. The connected version reveals the deception.

StripedChestSolution

Now, Adelson himself explains this by suggesting that our visual system compensates for the perceived shadow by assuming that the stripe in the lower half the object (or the B square in his example) must be lighter in tone. Interestingly, there is a whole body of research around the shadow itself, revealing as it does, our evolving understanding of light though its very absence.

Regardless of the shadow though, we assume (that is, our ‘visual system’ assumes) that given the proximity of the darker lines either side, this centre line must be lighter than its neighbours. And indeed it is but not lighter than the darker lines above the fold.

Both these examples relate to tone or lightness rather than the other two properties of the ubiquitous hue, saturation, lightness (HSL) colour model but this next example is a little different…

Blue and Green Illusion

Spiral Illusion – technically, there is no blue in this picture!

Amazing huh? Take a look at the details below. It may be a little hard to read but the RGB values in both are the same (8, 248 and 152).

Spiral Illusion Numbers  Spiral Illusion Detail

This one really goes in another box again – that of ‘colour comparison’ which reveals our perception of colour as one of comparison. We see colour not in isolation but by comparison to the surroundings. That is, we tend to make allowances for relative hues and compensate accordingly, seeing a different colour when the surrounding colours are different. The same goes for that pesky blue/black, gold/white dress: we compensate for the surrounding, or in that case, incident light and see the net results of our automatic adjustment. That particular example (which I promised I wouldn’t discuss) is of interest because the results are not consistent between individuals. Hence the debate…

Constant Craving

Here’s another one which relates to ‘colour constancy’ – the uncanny ability we trichromats have of perceiving consistent colour despite changes to the physical nature of a scene due. A process which, of course, lets us down sometimes…

ColourConstancyGirls

Each of these girls have the same colour in both their eyes. Their left eye being ‘true’.

These examples from the extensive work of Akiyoshi Kitaoka of Ritsumeikan University in Japan. Likewise the one below.

Andersoncolorillusion8

Both disks are the same – as you can see below

Andersoncolorillusion Masked

most interpretations of this deception are based on the idea of colour ‘constancy’ or our neurological disposition to en