This optical illusion has been making the rounds on the Internet recently, and most people are astonished to find that the A and B squares are the same shade, to the extent that they consult Photoshop to confirm that A is not darker than B. The explanation for this is simple – the eye is better at distinguishing sharp boundaries rather than shallow gradients of shade or places of even shade.
This satisfies most people. But why is this so? Why can’t the eye do that, and also be able to quantitatively compare the shades of two spatially separated areas?
It all comes down to space, or rather, lack thereof. There are roughly 130 million photoreceptor cells in the retina of the eye, each of which individually making measurements of the amount of light falling on it. However, if you look at the optic nerve bundle that conveys information from the retina to the brain, you’ll find that it contains only 1 million nerve fibres. That’s a contention ratio of 130 to 1, and the physiology of the situation dictates that one neurone cannot possibly contain all the information produced by 130 photoreceptors. There is a good reason for this, to do with the wiring of the retina and neural bandwidth limits, but you’ll just have to take my word for it for now.
As a result of all of this, there’s a significant loss in information from the total amount gathered by the photoreceptors, and that which is sent to the brain. The optimal solution would be to transmit the information that is most important to the survival of the organism, and that happens to be edges – sharp changes in shade – that you see, and of course that’s what the eye does. Each photoreceptor is linked to adjacent photoreceptors via ‘higher’ cells (still in the retina), and these higher cells perform a bit of processing called lateral inhibition.
Lateral inhibition is a relatively simply process that enhances edges between areas of different shade, and it’s mediated by a mechanism called centre-surround antagonism, which you can see in this Mach Bands demonstration.
Ultimately, a much better solution would be to have a whopping great big optic nerve that had 130 million nerve fibres in it, one for each photoreceptor in the retina. Alas, as I said earlier, it’s a question of space and there just isn’t enough room in our heads for such a big optic nerve, so we have to make do with a smaller one that causes us to have so much fun looking at visual illusions. Human vision isn’t perfect, but it is ‘good enough’ which is really the story of evolution.