Cambridge Festival Speaker Spotlight

Paul Bays is a Professor of Computation and Cognition in the Department of Psychology at the University of Cambridge. He investigates the nature of the internal representations our brains construct and sustain to interact with the world and the computations performed on them to achieve behavioural goals.

Paul will be speaking in "Seeing is Believing? The Secrets of Visual Illusions Unveiled", unveiling the secrets behind some of the most baffling visual tricks. As part of our Family Weekend, this talk will challenge young minds to question what they see and encourage them to look beyond the obvious.

It's fun when our eyes play tricks on us but why are optical illusions interesting to brain scientists?

Visual illusions have actually been a very important source of information about how brains work. Some illusions, like the dark spots you see after looking at a bright light - which we call afterimages - or the fact that objects seem to lose their colours in darkness, tell us things about the earliest stages of vision.

Specifically, they tell us about the different kinds of light-sensitive cells in the retina, at the back of the eye, and what signals they send to the brain. Other illusions, colour constancy for example, tell us about how the brain uses sensory signals to construct a sense of the external world.

OK, what is colour constancy?

You might have learned at school that the colour of visible light depends on its wavelength, with blue at small wavelengths, red at long wavelengths and a whole spectrum of colours in between. But that's really a very simplified view: the same wavelength of light entering the eye can be perceived as very different colours depending on the context it's seen in, like whether the scene is lit by morning or afternoon sunlight.

The colour we perceive isn't fixed by the wavelength of the light, it is more like the brain's best guess of the properties of the object out in the world that reflected light of that wavelength into our eyes - this means the colour of an object we see stays constant even when the illumination changes. I think this is one of the most significant discoveries ever made about the brain, that our perception is an "inference" about what is out there in the world, rather than just the sensations of light or sound or touch picked up by our sense organs.

Does this explain why people disagreed so strongly about "The Dress" a few years ago?

Yes that's right. This was a photo posted on Facebook in 2015 that went viral. The photo was of a dress that in reality was black and blue, but a large proportion of people who viewed it were convinced the dress was white and gold, including me.

It seems the reason was that the illumination in the photo was difficult to judge: the photo could equally well be of a blue or black dress seen in artificial light, which tends to be yellowish, or a white and gold dress in daylight, which contains more blue light. People seemed to have a strong and stable preference for one or the other interpretation, but it was very idiosyncratic as to which one.

Your talk at the Festival is illustrated with a kind of geometric image that seems to move jerkily as I'm looking at it. What is going on there?

This is an example of what's called a peripheral drift illusion. Most visual illusions have been studied for a hundred years or more, but this kind came to wide attention only relatively recently, thanks to Kitaoka Akiyoshi's "Rotating Snakes" example. We don't have a complete understanding of what causes it yet, but it seems to be related to the fact that visual parts of the brain respond to brighter parts of an image more quickly than darker ones. This process resets each time your eyes move: if you have the right kind of repeating pattern of patches with different brightness, the fact that some patches are perceived earlier than others can be mistaken for movement.

What is your favourite visual illusion?

It would be hard to pick just one, but an illusion I've always been fascinated by is the Wagon Wheel effect. This is where the wheels of a vehicle moving forwards at high speed sometimes look like they are turning slowly in the wrong direction, that is the direction they would turn if the vehicle was reversing. If you ask most people if they've seen that on the motorway they'll say yes, but they almost certainly haven't, they've seen it on TV instead.

That's because the illusion happens when the rotation speed of a wheel is almost but not quite a match to the frame rate of a video camera that's recording it. It can happen in films too, but until recently films were almost always filmed at a lower frame rate than TV, with 24 different images shown each second instead of 50 or more for TV.

A few years back some film directors made a big deal about shooting films you could watch at high frame rates, but it didn't prove popular - for a lot of people it made these very big-budget films look like daytime TV!

Is part of the experience of watching video or film an illusion then?

Yes! Video consists of static images presented one after another in a very rapid sequence: you see people and objects moving on the screen because of an illusion called the stroboscopic effect - nothing is actually moving. With modern LCD displays there is at least always an image on the screen, the old bulky cathode-ray tube TVs worked by scanning a point of light line-by-line over the screen to draw the image. Any particular bit of the screen was black for almost the whole time you were watching it!

The rate at which a light that switches on and off is seen as continuous is called the flicker fusion rate. LED lights also usually switch on and off very rapidly, so fast they seem constant, but if you move your eyes when looking at an LED light in darkness you might be able to see the light break up into a trail of dots as it is swept across your retina - a good time to see this is with the rear lights of the car in front when you're driving at night.

Published: 11 March 2025

_{The text in this work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License}