Archive for the ‘Vision and Colour’ Category

What would things sound like if we had six ears?

January 26, 2014
If we detected sounds in the same way detected colour we would have six ears. What?

If we detected sound in the same way we detected colour we would have six ears. What?

We see things and hear things in quite different ways.

We have two ears so we can sense the direction from which sound emanates. And each ear has two dimensions of hearing. What do I mean by that? Well the first dimension is loudness and quietness. The second is pitch.

As a pure sound tone increases in frequency we detect the sound as changing ‘pitch’. But a pure tone at 440 Hz (‘concert A’) is not qualitatively different from a pure tone at 256 Hz (‘middle C’).

With light things are more complicated. Each eye gives us an image of the world but each small region within that image elicits an experience we call ‘colour‘. We have the experience of brightness which is akin to loudness. But the sensation of ‘colour‘ is quite different from the sensation of pitch.

As the frequency of an electromagnetic wave reaching a patch in your eye increases from:

  • 400, 000, 000, 000, 000 Hz to
  • 1000, 000, 000, 000, 000 Hz

..our sensation changes qualitatively. Let’s call 1,000, 000, 000, 000 Hz a terahertz (THz) so this range is from 400 THz to 1000 THz.

As the frequency increases, the light first elicits the sensation of red, then yellow, then green and finally blue and violet. All that has changed is the frequency of the electromagnetic wave, but our sensation has changed qualitatively: red is not just a ‘low blue‘: green is not a shade of red. These are a completely distinct sensations.

The reason is that at daylight levels of light intensity, each single frequency of light stimulates not one, but three ‘sensors’ (called cone cells) at each location. Simplifying considerably, each type of cell when stimulated individually elicits one of the three basic ‘colour‘sensations: red, green or blue.

As the frequency changes each of the three sensors is excited to different extents, and our overall sensation of ‘colour‘ at each location in the image is a combination of the three qualitatively different basic sensations.

But what if sound worked like that?

Well if our sensation of sound pitch worked in a  similar way to our ‘colour sensations: red, green or blue, then we would have 3 ears on each side of our head – or at least three sensors inside each ear. Let’s stick with the 6-ear idea because frankly it is more dramatic.

However each ear each would just respond to a single range of frequency. The ranges would have to overlap otherwise there would be some frequencies that would elicit no sensation at all i.e. we would be deaf to those frequencies. But stimulating each sensor would elicit just a single ‘note’ or ‘tone’ or ‘pitch’ no matter what the actual frequency. The notes would be a kind of audio-red, audio-green and audio-blue.

A single pure tone of sound would then elicit an audio-‘colour‘ depending on the relative stimulations of the single sensor within each of the three ears on each side of our head.

Now of course, this whole idea is nonsense. But it did strike me as interesting that we have evolved such distinct ways of seeing and hearing. On reflection one can imagine reasons why the different detection mechanisms might make sense.

  • For light the range of frequencies we can detect ranges from 400 THz to 1000 THz. This is a ratio of just over a factor of two, and if these were musical tones they would cover only just over a single octave. And yet our sensation of colour can detect millions of distinct colours in this small frequency range, giving us phenomenal ability to discriminate between subtly different colours.
  • For sound the range of frequencies we can detect ranges from 20 Hz to 20,000 Hz. This is a ratio of around 1000, or just under ten octaves. The sensor we have in our single our ear needs to have as wide a range as possible to let us hear the sounds around us.

Anyway. As Forrest Gump might have said: “That is all I have to say about that”.

A look at sideways lightbulbs

November 27, 2011
A tungsten halogen light bulb

A tungsten-halogen light bulb. Notice the inner bulb within the outer glass enclosure

I have just replaced a light fitting in our living room in which light bulbs of every type failed to thrive:

  • Conventional light bulbs would blow after a couple of weeks;
  • Compact fluorescent bulbs would fade prematurely;
  • And even expensive halogen bulbs dimmed, and then died.

What was going on? Well, I think the answer has to do with temperature and gravity. As evidence, I present a picture of a failed halogen light bulb which I examined after I had removed it from the fitting. The key feature of the fitting was that it held three light bulbs sideways – and looking at the picture below it is clear that the tungsten elements have sagged until – instead of being isolated in the middle of the inner bulb – they are lying on the surface of the inner bulb. At this point the filaments would be very likely to fail.

Detail of a failed tungsten halogen light bulb.

Detail of a failed tungsten halogen light bulb. Notice how the filaments have 'sagged' to the point where they lie along the surface of the inner bulb.

But why does the light bulb have that complicated structure with the filament so close to the inner bulb? To answer that I have tell you about the amazing extreme physics that goes on inside these devices. There are 3 key points

  1. Tungsten-halogen light bulbs run hotter than normal light bulbs – around 2800 °C rather than 2500 °C. This makes them brighter, and converts a larger fraction of electrical energy into light than conventional – colder – light bulbs.
  2. However, increasing the operating temperature increases the rate of evaporation of the tungsten – which causes darkening of the light bulb – and failure of the filament as it thins to the point where the current density causes it to melt.
  3. This is where the halogen gas – usually iodine  – trapped in the inner bulb comes into operation. If the temperature of the bulb surface exceeds roughly 250 °C – then amazingly the iodine reacts with the deposited tungsten forming WI (No! Not molecules of Women’s Institute! Tungsten Iodide) which is a gas at these temperatures. However, when the WI molecules come close to the heated filament, they decompose and the tungsten is re-deposited on the filament! Wow! Is that a clever trick or what!

In order to keep this tungsten re-cycling working, the inner bulb needs to be rather hot, and so it is kept small, and close to the filament. Normal glass would soften at these temperatures, and so the inner bulb is made of pure silica – silicon dioxide. The outer bulb – made of normal glass (which is silica with impurities such as boron that make it less liable to shatter) is there to prevent anyone touching the inner bulb, and to prevent ultra-violet light from the hotter filament escaping.


The Wiki Page for tungsten halogen light bulbs is excellent. But I also strongly recommend the Lamptech Web Pages. They have been written by someone whose love and knowledge of the subject radiates from every sentence on the page. They also have a movie in which iodine is periodically introduced into a light bulb to clear up the deposited tungsten.

More illusion confusion

April 26, 2011
Circle Colour Illusion

Circle Colour Illusion designed by Akiyoshi Kitaoka. The image shows 12 sets of concentric circles. The 5 inner circles are all identically coloured, but it does not look like it. Read below to see how you can convince yourself. Click for a larger image.

After writing the article on colour illusions yesterday I was dissatisfied with the poor quality of the images in the Scientific American slideshow. Additionally, some of illusions were so amazing that they were hard to believe at all, and having poorly defined images just confused the issue. So I decided to make some of my own.

I used PowerPoint, which has the advantage that if you download the PowerPoint file you will be able to de-construct the images, and play around and convince yourself that your colour perception really is playing tricks with you! Enjoy 🙂

Blue Circle Illusion

Blue Circle Illusion. The grey circles in this image look 'pink-ish', but in fact they are just grey. Click for larger image.

The Vision of Johanna. And Jenny. And you and me

April 25, 2011

The eye on the left appears blue, but in fact it is exactly the same colour as the eye on the right.

Vision and colour are central to our perception of the world. But vision in general, and colour vision in particular, are still in 2011 subtly mysterious. I have been reminded of this several times recently and I just thought I would note three curious things about our vision system which continue to fascinate me.

The first is the variability of colour vision from person to person. At the celebratory drinks after the end of the 14th presentation, Protons for Breakfast graduate and Bob Dylan fan Joanna [pace the title] explained that she had distinctly different colour vision in each eye. The situation was such that at times she simply wasn’t able to say definitely what colour she really thought some things were! My colleague Jenny at NPL has also mentioned this but she seemed less troubled by it. But if I personally know two people with distinctly different colour vision in each eye, then how likely is it that your colour vision is the same as mine? Or that anybody’s colour vision is the same as the ‘standard’ sensitivity curve decided on by the International Commission on Illumination (CIE) in 1931?

  • Test the sensitivity of your colour discrimination here – its tricky!

And then there is the issue of the way we infer colour from context. I took the image at the start of this article from a slide show over at Scientific American. I simply didn’t believe that the two eyes were the same colour. So I downloaded the image, and sure enough in the image on the Scientific American web site, the eyes were slightly different colours. So I edited the image to make them exactly the same – that’s the image at the start of this article – and the illusion is still there. It is very unsettling to realise that our perception of a particular shade as ‘blue’ or ‘grey’ can be quite so context sensitive. In some of the images in the slide show, the author states that the explanation of the illusion is still unknown.

And finally there is the fact that even, after 150 years of careful study, we are still finding out new things about the structure of the eye itself. Scientific American this month reports on some astounding work that has uncovered additional light sensitive cells in the retina of human eye. The work is based on some very simple observations – that blind mice still responded to the daily cycle of light and dark, and still reduced pupil size in a bright light. Following on from this, the author, Ignacio Provencio, and his colleagues, uncovered previously unnoticed cells, comprising around 1% of the normal cone cells which are not part of the normal imaging system. Instead they are linked to the part of the brain which controls our daily cycle – the circadian rhythm – and also controls our pupil size.

So if we understand our visual system so little, how can we be sure that we really understand anything we see? Perhaps, the pattern of tiny pixels on my computer screen do not really indicate that it is now 45 minutes past midnight? Perhaps it is all just an illusion. Somehow, I doubt it. Goodnight.

UPDATED 7:23 a.m. 26th April 2011: Thanks to Nick Day for the comment and the additional links.

%d bloggers like this: