View Full Version : Random Question about creating an alien race.

01-24-2013, 01:07 PM
I was watching Sesame Street with my son and had a kind of weird thought. We as humans see red, yellow and blue as primary colors.

Is this a rule for humans? i.e. the structure of our eye, or is this a rule of optical physics.

Could you design an alien that sees orange, green and purple as primary? And what would the consequences of such a difference?

Freehand 5.5
01-24-2013, 01:32 PM
It is a matter of sensorial bandwidth.
There are Insects that see ultraviolet and snakes that see infrared. They see colors, we don't even have names for.
So yes, color reception and the resulting color theory is a thing of human perception.

Why do you ask?

And it goes further. If you would live in a different spectrum of light for long time, your brain will start to set this
spectrum on default. We adapt what we see. Our visual perception trains and changes constantly.

01-25-2013, 12:49 PM
Actually, our physiological perception of color bears little resemblance to the color theory we teach either artists or children. The primary colors we teach to children aren't really related to any physical properties of light. And in the art world there are actually two sets of primary colors.

Red, green and blue are the additive primaries (adding more light makes the color brighter). Mixing all three equally gives you white light, and any color of visible light can be created using these three colors. Cyan, magenta and yellow are the subtractive primaries (adding more pigment makes the color darker). Theoretically, mixing all three of these in equal measure will give you black. I say theoretically because that would require perfect pigments, which is something we cannot actually make. The reality is that mixing the colors typically gives you a very dark brown.

Furthermore, the additive primaries are the subtractive secondaries and vice versa. That is, red and cyan are complementary colors. If you subtract red from an image in Photoshop, you'll increase its cyan. Conveniently, when we talk about RGB and CMYK colors, they're arranged in a particular order that makes those relationships easy to remember. Just write them down in two lines:


Red is complementary to cyan, green to magenta, and blue to yellow.

Color perception is also a cultural artifact. There is at least one society in the south Pacific that has only one word to describe both blue and green. As a result, they cannot even tell the difference between the two colors. They can still describe the color with qualifiers, much like we can say something is "pale blue" or "a slightly purply blue," but no matter how green something is, they'll still associate it with blue. There is another culture in southeast Asia that doesn't have a way of describing hue; they can only talk about light and dark. Alas, I don't remember the names of either of those peoples; it's been far too long since my anthropology classes. It will be in my notes somewhere, but I'm not sure where those ended up.

Anyway, the consequences of a race that perceives color differently than we do might be text that is difficult to read because their rules of contrast are different. Maybe light blue text on yellow is very contrasty to them, but we just get a headache looking at it. Their notions of color harmony could be very different, resulting in art that we find ugly being thought very beautiful by them. Although, again, that's largely a cultural thing; our own Western ideas of color harmony are quite different today than they were even 50 years ago.

01-27-2013, 01:26 AM
Actually, our physiological perception of color bears little resemblance to the color theory we teach either artists or children.No, the very reason there's such a thing as "color theory" is directly related to how we perceive color. The purely physical properties of light do not lend the kind of richness to our perceptions of color that our physiology does.

There are basically three kinds of color sensitive cells in our retinas; they are each sensitive to a different wavelength of light. "Red" cells are most sensitive to lower frequencies (and thus we perceive low-frequency light waves as red), "blue" cells are most sensitive to high-frequency light (which we perceive as blue), and "green" cells are most sensitive to light of intermediate frequencies (which we see as green). These three kinds of retinal cells are the reason our color theory has three primary colors.

The colors we see in the world are just mixes of any number of frequencies of light. Any given mix will create a different reaction in the three kinds of cells: the red cells will react most strongly to any low-frequency light included, the blue cells to any high-frequency light, the green cells to mid-frequency light. The trick is that each cell basically gives off a single value; a given mix of light might cause red cells to give off a value of 100, blue cells a value of 50, green cells a value of 20 (veeery loosely, and on an arbitrary scale.). Because each cell only gives off a single value, though, any mix of light that produces the same response (100 red, 50 blue, 20 green) will be perceived as the same color. (Or, depending on your definition, is the same color.) This is why we can just mix pure red, green, and blue to simulate (almost) any color, like in a computer screen: the red light excites the red photocells to the required level while leaving the green and blue unaffected, and so on.

An important consequence of this "tri-stimulus" color perception is that the colors we can perceive can essentially be described three-dimensionally. One of these dimensions is brightness, from black through dark colors to light colors to white; if we (or a hypothetical alien race) only had one kind of photocell, then this is all we would be able to perceive. How do we describe different colors of the same brightness? You may be familiar with is division by hue and saturation; hue is the color-wheel color, like red or green or orange, while saturation measures the mixture between the color-wheel color and white or grey; pink and red might have the same hue but red would be more saturated.

A color-blind human has fewer than three kinds of photocells; most kinds of color-blindness lack only one photocell. Their color system is two-dimensional, and there are combinations of light that normal humans can distinguish that the color-blind cannot - they look like the same color. An alien with two photocells similar to human ones would have similar color vision to a color-blind human (though perhaps extended into the ultraviolet or infrared range).

An alien with three kinds of photocells would probably have similar color vision to a human, along with the identification of "hue" and "saturation" as important color principles.

If an alien species has more than three photoreceptors (or if their color-sensitive system works completely differently from ours; some thoughts on this later), then their color system would have more than three dimensions. This means that there would be mixes of light that we see as the same color that the aliens would see as very different colors. How would their color system work? I think hue would still be a useful notion, as most of the color-wheel colors just correspond to pure colors of light. The purples, however, are not pure colors; they are always a mix of high and low frequencies of light. I think these combinations would be similarly important for an alien with more than three photocells; at any rate, let's say for the sake of argument that our idea of "hue" will match up. That leaves two or more dimensions for the aliens where we only have saturation. I can honestly say that I have no idea how these would be perceived; they're probably way too heavily dependent on physiology.

And that's only for a few more than three photocells. What if the aliens perceive light the way we perceive sound? Our ears contain thousands of vibration-sensitive bristles, each sensitive to a different frequency. I think in this case the aliens might be able to perceive color the way we hear the difference between a trumpet and a clarinet and a bell; we call it "tone color", but in reality it's a much more subtle distinction than the differences between color.

One more thing, on seeing in the ultraviolet or infrared. If the aliens come from a planet much like Earth, there's limits on what frequencies they're likely to see at. On a planet with a functioning ozone layer, there is very little ultraviolet light above UV-A, so being able to see in that spectrum would be pretty much useless. Beyond UV are X-rays and gamma rays, which are even less likely to be seen in nature.

On the low-frequency end of the spectrum, all warm objects give off infrared radiation; even the air itself will give off infrared light. The peak frequency given off goes down as the temperature increases (which is why objects go from red-hot to white-hot as we heat them up); an eye that could seebelow a given infrared frequency wouldn't be able to distinguish anything besides a glowing fog. Beyond infrared is microwaves and radio waves. These are more likely than X-rays and gamma rays to be found in nature, but perceiving them wouldn't be very much like vision (their wavelengths are long enough that a detector would have to be pretty big, so they couldn't be put in arrays like our retina).

TL;DR: Brightness and hue correspond to physical properties of light, and are thus most likely to be recognized by aliens. With more than three photocells, perception is dependent on physiology; fewer than three would be like a color-blind human.

01-27-2013, 02:13 AM
Gah. A thousand-word answer and I didn't actually answer the question posed.

I was watching Sesame Street with my son and had a kind of weird thought. We as humans see red, yellow and blue as primary colors.
Actually, we see red, green, and blue. The red, yellow, and blue that we learn in school are (close to) the primary subtractive colors (as Midgardsormr points out). The difference is that paint works by removing colors from white (the color of the paper); removing blue gives yellow, removing green gives magenta (which is close to red), removing red gives cyan (bluish-green).

Is this a rule for humans? i.e. the structure of our eye, or is this a rule of optical physics.
I think I answered this one in the previous answer; it's completely dependent on what frequencies of light our three kinds of photocells are most sensitive to.

Could you design an alien that sees orange, green and purple as primary? And what would the consequences of such a difference?
In general, you could have any different set of photocells with different sensitive frequencies, even more than three of them! (Which would give you more than three primary colors, but would also have big effects on how color is perceived in general.) Let's stick with three, so we don't have to speculate about alien physiology. So, orange, green, and purple. The orange cell is sensitive to light with sorta-low frequency, while the green cell is sensitive to light with mid-range frequency. But the "purple" cell? The problem here is that there's no single frequency of light which is "purple"; purple is what we call light with a mix of low and high frequencies but no mid-range frequencies. But let's say that this is what your "purple" cell is sensitive to: light of high and low frequency, but not mid-range frequency.

This is actually kind of interesting. First, could our alien distinguish red light from blue light? The purple cell would respond the same way to red light or blue light. In order to distinguish them, we would require that the sensitive range of the orange or green cells overlaps the purple cells' range at the top or bottom. If there is overlap between orange and red and between green and blue, for instance, then red light would produce a high orange and purple response but no green response, while blue would produce a purple and green response but not an orange. There is very little overlap in humans, though; the reason we have primary colors at all is that green light (for example) strongly stimulates the green cells but barely stimulates the red or blue cells at all. If there was overlap in the aliens between purple cells and both green and orange, then there would be no "purple" primary color! If there was no overlap on one or both ends, on the other hand, the primary color would be either red or blue (or both)!

Orange is still pretty close to red, so let's say that it overlaps purple, except maybe at the very lowest frequencies (which would then be indistinguishable from blue). Then the primary colors, the three kinds of dots in the aliens' TV sets, would be orange (which stimulates only the orange cells), green (which stimulates only the green cells), and blue. Why blue? Because red light will stimulate both purple and orange, while blue light will only stimulate purple. Their TVs would probably look a lot like ours, then, except the red dots would be a little more orange. The colour combinations would look weird, though. A picture of a red apple would use the orange and blue dots; that is, it would be intended to activate the orange and purple photocells, because that would look like what we call "red". To a human, though, the display would look fuchsia (which is how we perceive the combination of orange and blue). In fact, such a display would be incapable of producing anything we'd recognize as "red". To them, our displays would also look strange: the color our TVs display for "orange" wouldn't look at all orange to them, for example.

I'm going on about our TV screens (though similar differences would be found for all color combinations, including paint and ink) because that's where I think the greatest difference would be. They'd probably still have the same color wheel as us, for example, for reasons I explained in the previous post. That is, unless they really can't distinguish red from blue, in which case they wouldn't have anything we would recognize as purple; anything we call "purple" they'd see as the same color as red or blue.

01-27-2013, 09:41 PM
So in a nut shell. The consequences of different eye structures could be:

A human visiting an alien house or ship could be driven "crazy" by the monochromatic nature of the house or ship, when in fact the alien sees a great amount of color. OR vise versa.

01-27-2013, 10:00 PM
Remember that what we call "color" is a concensus thing. Most humans are sensitive to RGB, but the exact frequency and sensitivity of those R, G, and B components varies somewhat from person to person. There even seem to be some women who actually encode _four_ color sensing proteins instead of the usual three. We spend a lot of time learning to describe the various patterns of stimulus in our color sensors in terms of language. In effect, we are self-calibrating to an extent.

If you have a race with only one type of color sensor, then they'll have a completely different experience than we do with our multiple kinds of sensors. Same with only two sensors. More than that will just result in refinements of color (or possible extending into the IR and UV ranges) compared to humans.

Humans have a color response range that's just about the complement of green plants. That is, our maximum sensitivity just happens to be where plants reflect the most. Plants + Humans result in a use of the primary spectral output of our star. If your star is much dimmer and redder, your race is likely to see further in the infrared and less in the blue because that's where their star's output would peak. For a much bluer star, there would likely be enhanced UV response, but likely not a significantly different response in other colors.

If certain colors that the local race uses but that the humans couldn't see were designated as "danger" or "religion" or "sexual", then there is no end of hilarity that might ensue. A similar argument would apply to acoustic or chemical communications media.

EDIT: One thing that I neglected to mention is that although we are sensitive to three(ish) broad groups of frequencies and have panchromatic sensors for low-light conditions, what we describe as "color" is actually pretty variable. Our color sense is optimized for color constancy over a wide range of lighting conditions. If you have someone mix and paint the same "blue" under varying light conditions, they will look quite a bit different if placed side-by-side under bright light.

01-27-2013, 10:15 PM
Hehe, I'm thinking the consequences are those who make "color" tinted sunglasses make a fortune :)