"...we see on the ground a number of spots of light, scattered irregularly, some large, some small..."

December 23, 2008 8:26 AM

The Nature of Light and Color in the Open Air"Moreover, this book is written for all those who love Nature; for the young people going out into the wide world and gathering together round the camp-fire; for the painter who admires but does not understand the light and colour of the landscape; for those living in the country; for all who delight in travelling; and also for town-dwellers, for whom, even in the noise and clamour of our dark streets, the manifestations of Nature remain." - Marcel Minnaert

Marcel Minnaert (1893 - 1970) was a Belgian astronomer whose other interests included (among other things) bubbles and their pulsations, the solar spectrum, and poetry. During World War II, he taught astronomy to his fellow prisoners. In 1951, he was awarded the Bruce Medal. A asteroid and crater on the moon have been named for him, as has a building at Utrecht University.

One of his most accessible and abiding legacies is a little book - The Nature of Light & Colour in the Open Air, which is no less than a complete guide to the visual phenomena that surround us. Minnaert explains, in layman's terms, why stars twinkle, illusions of motion, how to build a nigrometer to explore the effects of distance on light and color, what causes the fata morgana, and why, exactly, the sun makes circles when viewed through hair or the twigs of a tree. These and a hundred other visual curiosities are explored and explained at length with helpful diagrams and occasional geometry. The latest edition (linked in the title) has been re-typeset and includes color photography.


Edward Tufte calls it a change-your-life classic.

posted by jquinby (17 comments total)

Rats. I sort of wish that first link in the intro text wasn't there now.
posted by jquinby at 8:28 AM on December 23, 2008

I read about Minnaert while avoiding classwork in college, and was led to Color and Light in Nature. Using the two together really opened my eyes to wonder in nature. Thanks for the reminder - I'll have to get my books back from my brother-in-law!
posted by notsnot at 8:44 AM on December 23, 2008

My library seems to have the 1993 edition! I'll have to check this out, it sounds great. (And after all these years, I still don't understand color. Why does receiving a blue photon and a red photon on my retina produce the same sensation as a purple photon?)
posted by DU at 8:47 AM on December 23, 2008

Oh and I still remember a solar eclipse that, coincidentally, must have been about 1993. Millions of tiny crescents all over the ground in the shadows of trees. Nature's pinhole camera.
posted by DU at 8:49 AM on December 23, 2008

Color is just our perception of different frequencies of light due to our physiology. Purple lights up the "red" cone and the "blue" cone in roughly equal proportions.
posted by notsnot at 8:50 AM on December 23, 2008

Purple lights up the "red" cone and the "blue" cone in roughly equal proportions.

Right, I get that. But my question is the other direction.

I guess the problem is that I'm imagining just a few photons entering at a time. You'd have to get awfully fine timing to have a red and a blue hit close enough in time (and space?) that you'd see it as purple rather than two flashes. But in reality, we usually have trillions of photons streaming in, so it all smears out.
posted by DU at 8:56 AM on December 23, 2008

Well, ignoring luminosity (or using properly attenuated red and blue photons), to your eye, it's the same thing - there is no functional difference between a purple photon, and a red and a blue photon together.

Think of motor oil - assuming viscosity is a linear property (bear with me) there's not functional difference between some 30 weight oil and a mixture of 10 weight and 50 weight.

And as far as "a few photons entering at a time" remember that photons are waves - the classic diffraction experiment shows that a *single photon* acts like a statistically proper proportion of all possible outcomes.
posted by notsnot at 9:17 AM on December 23, 2008

Think of motor oil - assuming viscosity is a linear property (bear with me) there's not functional difference between some 30 weight oil and a mixture of 10 weight and 50 weight.

But the principle of superposition is exactly the opposite of this. A blue wave and red wave do NOT mix to create a purple wave. My eye receives each separately. In principle, I should be able to distinguish "red+blue" from "purple".
posted by DU at 9:19 AM on December 23, 2008

(Maybe I should read the book. I'll stop talking now.)
posted by DU at 9:29 AM on December 23, 2008

Thanks! I cited Minnaert in another thread (the link about the sun and the tree), then was surprised to see that neither he nor his book had been the subject of a FPP yet. I bought a copy of this book at Powell's a few years ago when I was in Portland for OSCON, and buried my nose in it all the way back to the bus stop. I don't think I looked up once.
posted by jquinby at 9:51 AM on December 23, 2008

Nice, thanks. I'm reminded of a little museum I wandered into on the Mall in Washington. Kind of a micro museum in a small house-like building it seemed to be the efforts of one collector, who presumably endowed the museum.

This guy's apparent specialty was collecting large paintings with atmospheric and weather themes, foggy gardens, rain on the fields, etc.
posted by StickyCarpet at 10:12 AM on December 23, 2008

This book has been on my amazon wishlist for a while now. Tis the season, so maybe I need to drop some more hints to the family.
posted by zap rowsdower at 10:32 AM on December 23, 2008

this is at my university library and now on my reading list,

thank you
posted by sponge at 10:54 AM on December 23, 2008

I just sold my copy of Light and Color in the Outdoors on Amazon.

I got it because of all the rave reviews; to be honest, I thought it was pretty dry and boring, and definitely not worth $70 unless you're, like, really into optics.
posted by designbot at 12:57 PM on December 23, 2008

Why does receiving a blue photon and a red photon on my retina produce the same sensation as a purple photon?

It doesn't.

No, really.

The first thing to keep in mind is not to confuse color names with colors. What we call "purple" covers a wide range of colors.

Next, take a look at the CIE chromaticity diagram. It relates wavelengths of light, and mixtures thereof, to the colors we actually see.

(Caveat 1: the colors shown in the diagram are only approximations, since computer screens cannot show all possible colors. Caveat 2: this is for a standardized, theoretical reference observer. The real colors may vary from person to person depending on physiology.)

Monochromatic light--that which might be produced by your "single photon"--appears as one of the colors along the upper, curved edge only. Colors in the interior of the diagram, or along the lower straight edge, can be produced only when two or more wavelengths are combined.

The only purple you find along the monochromatic edge is a bit of bluish-purple at the very bottom, around 380nm.

If you combine light of two different frequencies, you can get colors along the line connecting those two frequencies. If you mix, say, blue light at 470nm with red light at 700nm, you can imagine a line connecting those two points, and you can get the color shown at any point along that line, depending on the proportion of the light of the two wavelengths. You can get a number of different shades of purple, some very close to that monochromatic purple at 380nm, but not exactly that shade.

But you might still ask why a given color in the interior of the diagram can be created in more than one way from mixtures of two or more wavelengths (e.g., a specific shade of purple might be made by a mixture of light of 380nm and 580nm, or that same shade might be made by a mixture of light of 490nm and 700nm), or you might ask even if it doesn't work for purple, might it work for yellow since the part of the monochromatic curve from about 550 to 700nm is essentially straight. (Yes.)

The answer is that we (most of us) have only three different types of color receptors in our retinas. So while a given light source might be made up of many many (perhaps infinitely many) different wavelengths of light, at varying intensities, these are essentially reduced to three "numbers" at our retinas. There is necessarily a loss of information--some different mixtures of light wavelengths necessarily look the same. (And the three numbers can be represented as two on the chromaticity diagram because multiplying all three numbers by a constant only makes the color seem brighter or dimmer.)
posted by DevilsAdvocate at 1:05 PM on December 23, 2008

DU, you're right that it's a weird topic. The mechanics of 'how' can be described pretty well, as DevilsAdvocate has done aboce, but the reasons 'why' are mysterious. It's just the way the brain makes color out of signals from the eyes.

A good illustration of why this is weird is to consider sound. Sound wavelengths are very analogous to light wavelengths, but we construct our perceptions very differently. Imagine if I played two or three notes on a piano and we heard not a chord, but a single note, not the same as any of the individual notes played. That's pretty much how color perception works.

It's also one of the reasons why different digital camera sensors and different films can have different feels to their color characteristics. There's no one correct, accepted way of combining wavelengths to produce color, and different mediums have slightly different ways of going about it.

Another fun example is in digital display technologies. You're of course familiar with the red, green and blue dots on a screen or a projected image. A DLP projector shows the red, green and blue parts of the image one at a time. There is no single instant where any two are displayed together. They flicker between all three fast enough that our eyes & brains assemble the information into a single color image.*

* Supposedly about 10% of the population sees an annoying flicker when viewing a DLP projection. Different people have different flicker fusion rates.
posted by echo target at 1:26 PM on December 23, 2008 [2 favorites has favorites]

The nigrometer experiment, by the way, is quite (ahem) eye-opening. Isolating a distant object from its immediate environment can illustrate the effects that the air column has on the its color. A distant window, for example, takes on a bluish tint. As you get closer and closer, the the blueness fades away. And you can throw one together with a paper-towel tube and some aluminum foil.

Kids, by the way, don't seem to be as impressed with this sort of thing as I am.
posted by jquinby at 1:40 PM on December 23, 2008