Tag Archive: appearance

Aug 22 2010

More colours than the rainbow

This post is about making the bridge between how a physicist understands colour, and something a bit more useful.

Light is a collection of electromagnetic waves; for a physicist the most important property of a wave is its wavelength, its “size”. The wavelengths of visible light fall roughly in the range 1/1000 of a millimetre to 1/2000 of a millimetre. (1/1000 of a millimetre is a micron). Blue light has a shorter wavelength than red light.

Things have colour either because they generate light or because of the way they interact with light that falls upon them. The light we see is made of many different wavelengths, the visible spectrum. Each wavelength has a colour, and the colour we perceive is a result of adding all of these colours together.

The diagram to the right summarises this: it’s called a chromaticity diagram, the numbers around the edge are wavelengths in nanometres (a millionths of a millimetre), pure, single wavelength light falls on this line; any point inside the line is formed from the mixture of wavelengths. The line represents “all the colours of the rainbow”; colours inside the line are not in the rainbow. The chromaticity diagram is the “periodic table” for colour scientists, it’s iconic and it summarises the world of colour.
This chromaticity diagram is just a slice through a volume, we could draw another one a little bit dimmer, and a little bit dimmer than that until we reached black.

How do we get to this diagram? The central issue to understanding perceived colour is that although the light in the environment comes as a mixture of a multitude of wavelengths, our eyes are limited by the light sensitive cells they contain, known as “cones”. In humans cones come in three types, which are sensitive in three different ranges of the spectrum. Roughly there are red-, blue- and green sensitive cones. So the eye gives just three readings in terms of colour description. The chromaticity diagram comes from a calculation trying to predict these three values and combining them to fit on a flat page (which only gives you two dimensions to play with).

Some other animals don’t have three sorts of cones. Birds, for example, have four – this is known as tetrachromacy which sounds to me like some sort of wizardry involving chairs (I’m reading Terry Pratchett at the moment). Birds have an extra type of cone in the ultra-violet part of the spectrum – so they are sensitive to wavelengths which we are not. Most mammals are dichromatic, but other primates, like humans, are also trichromatic. Dichromatic animals will be able to perceive a smaller range of colours than us. The evolutionary implication here is that earlier mammals lost some colour sensing ability, possibly for a gain in low light sensitivity but some mammals subsequently regained the ability.

The chromaticity diagram is still something of a physicists play-thing, it’s useful for doing calculations. There are other ways of describing colour which are related to human perception, they are developments based on the first steps used in constructing the chromaticity diagram. The aim of these methods is to similar numbers to colours that look similar; make those numbers reasonably easy to explain from a conceptual standpoint and try to give numbers twice as big to colours that are twice as bright. My favoured system in this respect is the CIE LAB system. The colours are expressed as three numbers L, a and b: L tells you something about overall brightness, a tells you the point on a scale between red and green and b tells you a point on the scale between yellow and blue.

But all of this is a bit of a fraud, because actually the colours you’re seeing on your monitor aren’t the real colours I’m trying to show you. The problem is that display devices contain red, green and blue elements but they don’t fall anywhere near the extremes of the chromaticity diagram and we can only get colours inside the the triangle defined by the red, green and blue elements in the monitor. A typical monitor gamut is shown here.

All this is based on the study of ideal colour stimuli (little square patches) on grey backgrounds, things get an awful lot more complicated if we start to worry about context. This is best illustrated with an image:

As far as my computer is concerned squares A and B have the same colour, my brain and your brain are interpreting the scene context and giving them different colours. This is called Adelson’s checker shadow illusion.

Feb 02 2010

Why is that butterfly blue?

Some colours come from the properties of individual molecules, some colours come from the shape of things. This is a post about the colour from the shape of things – structural colour, like that found in the Morpho rhetenor butterfly pictured on the right.

To understand how this works, we first need to know that  light is a special sort of wave known as electromagnetic radiation, and that these waves are scattered by small structures.

For the purposes of this post the most important property of a wave is it’s wavelength, it’s “size”. The wavelengths of visible light fall roughly in the range 1/1000 of a millimetre to 1/2000 of a millimetre. (1/1000 of a millimetre is a micron). Blue light has a shorter wavelength than red light.

The Spectrum of visible light (Image from

Things have colour either because they generate light or because of the way they interact with light that falls upon them. The light we see is made of many different wavelengths, the visible spectrum. Each wavelength has a colour, and the colour we perceive is a result of adding all of these colours together. Our eyes only have three different colour detectors, so in the eye a multiplicity of wavelengths is converted to just three signals which we interpret as colour. The three colour detectors are why we can get a full colour image from a TV with just three colours (red, green and blue) mixed together. Some other animals have more colour sensors, so they see things differently.

The problem with viewing the small structures that lead to the blue colour of the butterfly wings is that they have interesting features of a size about the same as the wavelength of light, and that means you can’t really tell much by looking at them under a light microscope. They come out blurry because they’re at the resolution limit. So you resort to an electron microscope, electrons act as a wave with a short wavelength so you can use an electron microscope to look at small things in much the same way as you would use a light microscope except the wavelength of the electrons is smaller than that of light so you can look at smaller things.

So how to explain resolution (how small a thing you can see) in microscopy. I would like to introduce you to a fresh analogy in this area. Summon up in your mind, a goat (tethered and compliant), a beachball (in your hands), and a ping-pong ball (perhaps in a pocket). Your task is to explore the shape of the goat, by touch, via the beachball, so proceed to press your beachball against the goat. The beachball is pretty big, so you’re going to get a pretty poor tactile picture of the goat. It’s probably going to have a head and a body but the legs will be tricky. You might be able to tell the goat has legs, but you’re going to struggle to make out the two front legs and the two back legs separately. Now discard the beachball and repeat the process with the ping-pong ball. Your tactile picture of the goat should now become much clearer. The beachball represents the longer wavelength of light, the ping-pong ball the shorter wavelengths of electrons in an electron microscope.

And now for scattering; retrieve your beachball; step back from the goat. You are now going to repeatedly throw beachball and ping-pong ball at the goat and examine where the balls end up having struck the goat. This is a scattering experiment. You can see that how the ball bounces off the goat will depend on the size of the ball, and obviously the shape of the goat. This isn’t a great analogy, but it gives you some idea that the shape of the goat can lead to different wavelengths being scattered in different ways.

So returning to the butterfly at the top of the page, the iridescent blueness doesn’t come from special blue molecules but from subtle structures on the surface of the wings. These are pictured below, because these features are smaller than the wavelength of light we need to take the image using an electron microscope (we are in ping-pong ball mode). The structures on the surface of the butterfly’s wing look like tiny Christmas trees.

Structures on the surface of a morpho butterfly wing (scale bar 1.8 micron)

These structures reflect blue light really well, because of their shape, but not other colours – so the butterfly comes out blue.

Another example of special structures that interact with light is this is a *very* white beetle:

Cyphochilus beetle (Image by Peter Vukusic)

The cunning thing here is that the beetle manages to make itself very white, meaning it reflects light of all wavelengths very efficiently, using a very thin scales (5 micron). This is much better than we can achieve with synthetic materials. The trick is in the detail, once again the scales have a complicated internal structure as you can see in this image from an electron microscope:
Cross-section of a beetle scale (scale bar is 1 micron, Image by Peter Vukusic )

It turns out that the details of the distribution of the scale material (keratin) and air in the scale conspire to make the scale highly reflective. Making things white is something important to a number of industries, for example those that make paint or paper. If we can work out how the beetle does this trick then we can make cheaper, thinner, better white coatings.

Finally, this is something a little different. If you’ve got eyes, then you want to get as much light into them as possible. The problem is that some light gets reflected from the surface of an object, even if it is transparent – think of the reflection of light from the front surface of a clear glass window. These structures:

The surface of a butterfly’s eye (scalebar 1micron, Image by Peter Vukusic)

known an “anti-reflective nipple array”, are found on the surface of butterfly eyes. The nipples stop the light being reflected from the surface of the eye, allowing it instead to enter the eye. Similar structures are found on the surface of transparent butterfly wings.

In these cases animals have evolved structures to achieve a colour effect, but more widely we see structural colours in other places like rainbows, opal, oil films and CDs. The sky is blue for a related reason…

Sources
The work on butterflies and beetles was done by a team led by Peter Vukusic at Exeter University: