PDA

View Full Version : The color of stars



George
2009-Jun-12, 02:01 PM
There has been much stated on this topic over the years and in many of the threads, but I failed to find a dedicated thread to this subject. So, here we are. :) [This is also in response to Jeff Root's idea to address this.]

There is certainly a subjective element to color as we do not all have exactly the same eyes or possibly not the same color neural processing. Yet there is also the objective observations that tell us red apples are indeed red. Just how much the subjective elements impact one's color determination will vary due to both the spectral source or illuminate and with the spectral sensitivity that varies from person to person.

Some things, however, are far less subjective. Red lasers, for instance, are monochromatic and stimulate the "red" color cones, and to a lesser extent the "green" cones, such that red is the result for probably every one with "normal" color vision. This may be true with nearly all monochrmotaic light sources, even green lasers, which are in the middle of the visible spectrum stimulating the "red" color cones to a lesser extent than green, with very little stimulation of the "blue" color cones.

Most stars M-class and hotter are found with their spectral irradiance close to a black-body (Planck) distribution. This means light from the entire spectrum will be impacting our eyes, so there is a greater tendency to see a whiter version of any one specific color. An M-class "red" giant, for instance, radiates orange heavily, then lesser amounts of yellow, green, and even blue. [Added: It may radiate heavier in the red, but orange is more of the net result due to the other color additions.]

For stellar color discussions, another important factor to consider is the view that the spectral energy distribution (SED) of a star is better represented if it is converted to photon flux distribution. Stars like our Sun have peaks in their SEDs in the blue portion of the spectrum, yet if you convert to their photon flux, then the blue end of the spectrum will be seen to have less flux than the rest of the spectrum, which is very flat.

By converting to photon flux, it becomes more obvious why very hot stars are not as blue as we might expect since the lower wavelengths gain in relative flux as compared to a SED due to E = h v (v ~ frequency, nu). Stars like the Sun have such flat photon flux distributions that they are easily, IMO, classifiable as white stars.

This white appearance for stars is greatly enhanced by what is known as color constancy. Our eyes tend to make the brightest light sources as white light sources. Turn your car's headlights on in the daytime and they will appear much more yellow than when you have them on at night.

Attached is a wonderful image from Stefan Seip (with permission) of the Southern Cross. He used a technique known as progressive defocussing that spreads the star's light over a broad range of the camera's color sensors. [I think David Malin used this technique prior to Seip.] This desaturates the stars and does a better job of revealing their color. Of course, our eyes do not normally smear the star's light, so we won't see so much red or blue as noted in the image.

Assuming we go into space (away from atmospheric effects) and reduce the flux of the stars to a normal photopic range, what color would we see for each class? Here's my prediction:

O & B class - bluish-white
A class - white with a hint of blue
F & G class - white
K class - yellowish-white
M class - orange [perhaps some redish-orange. Carbon stars may be even more red.]
T class - These will vary in color depending on the molecular composition of their atmospheres, which are allowable due to their lower temperatures. [Grant has demonstrated that they can be crimson or maroon in color.]

Ken G
2009-Jun-12, 04:18 PM
Some things, however, are far less subjective. Red lasers, for instance, are monochromatic and stimulate the "red" color cones, and to a lesser extent the "green" cones, such that red is the result for probably every one with "normal" color vision.There is actually still some surprising subjectivity involved here. For example, if we took a normal person and put glasses on them (without telling them, special inserts perhaps) that converted red light to blue light, and vice versa, then after we were done they'd say "hey, apples are blue, not red, what gives?" But if we forced that person to wear those same glasses their entire life, from birth, they would never make that observation--they would just say "sure, apples are red, just like everyone says". What's more, even if we gave them those glasses later in life, they would probably go through a rebellious phase where they said "everyone has all the colors wrong", but this might not last for too long-- like the character in 1984 who eventually agrees that "2+2=5", this person would probably eventually let go of their personal impression that apples were blue, and eventually admit, "all right, apples are red, I don't know what I was thinking". So we have an element of convention in color that is inescapable.

Or another interesting experiment we could do is swap the wiring of someone's red and green cones. This would have a similar effect as the above, just physiologically inside the brain. There might be some temporary cognitive dissonance, like the goggles that turn everything upside down, but since the connections of the "red" cones and the connections of the "green" cones were never fundamentally different, they were just distinguished from each other, and since they are still distinguished from each other after the bizarre surgery I'm describing, I suspect the effect of the surgery would be short-lived, and soon the subjects would return to seeing apples as red. What I don't know is, would they change the color they see to match "red" or would they change their definition of red to match what they see, or, even more importantly, is there really any difference between these possibilities?

PraedSt
2009-Jun-12, 05:57 PM
Attached is a wonderful image from Stefan Seip (with permission) of the Southern Cross. He used a technique known as progressive defocussing that spreads the star's light over a broad range of the camera's color sensors. [I think David Malin used this technique prior to Seip.] This desaturates the stars and does a better job of revealing their color. Of course, our eyes do not normally smear the star's light, so we won't see so much red or blue as noted in the image.

Assuming we go into space (away from atmospheric effects) and reduce the flux of the stars to a normal photopic range, what color would we see for each class?
You got anything on the net that describes this technique? Sounds interesting, but I'm a bit confused. You mention both 'spreading' and 'reducing' for instance. I can't see how it works. :(

George
2009-Jun-12, 06:10 PM
There is actually still some surprising subjectivity involved here. For example, if we took a normal person and put glasses on them (without telling them, special inserts perhaps) that converted red light to blue light, and vice versa, then after we were done they'd say "hey, apples are blue, not red, what gives?" But if we forced that person to wear those same glasses their entire life, from birth, they would never make that observation--they would just say "sure, apples are red, just like everyone says". Yep. In fact, when I was typing how red apples are red, my first thought was that Ken G might just jump in and show how this can be considered a definitional expression. :)

Your view makes even more sense because there is no such thing as a "red" photon (or any other color photon), it is simply a matter of assignment within our cerebral mapping.

Of course, you have introduced some really fancy glasses that would require batteries or solar cells, which might fail and trip the wearer out big time. :)


Or another interesting experiment we could do is swap the wiring of someone's red and green cones. This would have a similar effect as the above, just physiologically inside the brain. There might be some temporary cognitive dissonance,... You would be wise to do this to someone from the '60's where black lights and fluorescence paint for the psychedelic art were all over their (ok, "our") rooms. ;)


... I suspect the effect of the surgery would be short-lived, and soon the subjects would return to seeing apples as red. Yes, and the red they claim would most likely be the same red a normal person would claim as long as the wavelengths entering the eye were identical. Yet, if they could switch brains quickly, they might be surprised. [Added: assuming we leave some residual color memory to remain before swapping the rest. :)]


What I don't know is, would they change the color they see to match "red" or would they change their definition of red to match what they see, or, even more importantly, is there really any difference between these possibilities? This sounds like another reference frame question, or at least a possible analogy to your GR explanations. The invariance being the wavelength and how it interacts with the retina, though the mapping may vary from individual to individual but may possibly be understood if we could give each some sort of color accelerometer. I think the blue twin would return the younger. ;)

What I think would be very tricky would be how everyone might see all the other millions of colors that come from the spectral mix. The color result from metamers, for instance, probably vary from person to person.

Stars, however, with their Planck distribution, make color determination so much easier, I think. The photon flux is either flat, a little heavy on the blue end, or a little heavy on the red end. There are no spikes that would allow metamers into our color picture, except in the T-class stars. [Quickly added...The interesting thing here is how the M-class stars have the advantage of a heavier red end than the hot stars do for a heavy blue end. This is due to the fact that E = h nu keeps the blue end squashed down more so than the red end.]

George
2009-Jun-12, 06:17 PM
You got anything on the net that describes this technique? Sounds interesting, but I'm a bit confused. You mention both 'spreading' and 'reducing' for instance. I can't see how it works. :( It is simpler than it sounds. If you have a variable rate speed control for your telescope's clock drive, then simply slow-down, or speed-up, the rate so that the star leaves a streak during the exposure. During the exposure, make slight adjustments to the focus so that it becomes more and more out of focus, which spreads the light over a larger area of your camera's sensor.

PraedSt
2009-Jun-12, 06:49 PM
It is simpler than it sounds. If you have a variable rate speed control for your telescope's clock drive, then simply slow-down, or speed-up, the rate so that the star leaves a streak during the exposure. During the exposure, make slight adjustments to the focus so that it becomes more and more out of focus, which spreads the light over a larger area of your camera's sensor.Oh I see. You're right, it is simpler than it sounds. Thanks. :)

EDIT: That's quite clever.

Ken G
2009-Jun-12, 06:56 PM
This sounds like another reference frame question, or at least a possible analogy to your GR explanations.That was my thought also, I suspect that at the core of both is something about how our intelligence makes sense of our environment, and what that expression even means.


I think the blue twin would return the younger. ;)Very funny. :naughty:
What I think would be very tricky would be how everyone might see all the other millions of colors that come from the spectral mix. Yes, and disentangling the physiological from the cultural. If it's true that men recognize about ten colors and women about a hundred, is there some physiological difference there, or purely cultural? (The question is rhetorical, so the "genetics does everything" camp need not turn out to debate the point-- I raise it only to say that even a concept as simple as "color" actually represents a complex interplay between the brain and the environment.)


Stars, however, with their Planck distribution, make color determination so much easier, I think. Yes, it's more like the way I personally relate to color-- a linear scale with just one degree of freedom (average wavelength)! To "Martha Stewart" types, color is more like tensor algebra, it seems to me.

George
2009-Jun-12, 08:38 PM
Oh I see. You're right, it is simpler than it sounds. Thanks. :)

EDIT: That's quite clever.
Yes, ain't it! However, it is not easy to make it come out as nicely as Stefan's.

Attached below was one image Roy Lorenz (tech. with Kitt Peak's AOP) and I did in an attempt to be like Stefan. I had a hard time trying to maintain an even rotation rate on the focus knob. [Naturally, it is superior in character than Stefan's. ;)]

The bluish-white star to the right is 16 Scorpii and is a class A4V. It goes from white to bluish-white to a light blue. But I think the automatic image processing of the Canon Rebel may not have been the greatest. Notice how 18 Sco, the now second best Solar twin, has a hint of blue to it. I think this is wrong, though it would be great if it wasn't. The camera used does not allow manual color temperature control, unlike the more expensive models. Setting the color temperature to around 5900K should produce more accurate color for our vision when imaging stars, I think.

We were a little smoother in the next attachment, which is 18 Sco by itself. I strongly suspect that this color image is accurate in showing it as a white star, from head to tail. Of course, some of the blues are depleted from both images due to atmosphere scattering, but the difference should be neglegible.

Looking again at Seip's Southern Cross, here are the spectral classes of the main 4 stars starting at the top and going clockwise...

Gamma Crucis: M4 III
Delta Crucis: B2 IV
Alpha Crucis: B1 V
Beta Crucis: B0.5 IV

PraedSt
2009-Jun-12, 08:57 PM
Looking again at Seip's Southern Cross, here are the spectral classes of the main 4 stars starting at the top and going clockwise...I was going to ask you about that. Any chance of a bigger picture? It opens up too small, and zooming makes it too blurry. (Your two pics, on the other hand, turn out great).

George
2009-Jun-12, 09:01 PM
If it's true that men recognize about ten colors and women about a hundred, is there some physiological difference there, or purely cultural? What chick flick did that come from? :) I had heard there were differences but could it be that much? Anyone know?


(The question is rhetorical, so the "genetics does everything" camp need not turn out to debate the point-- I raise it only to say that even a concept as simple as "color" actually represents a complex interplay between the brain and the environment.) Yes. Just try to go out and find a spectral color chart that matches anyone elses independent color chart. I finally starting doing a sort of scatter charting of the colors for each 5 or 10 nm range of wavelength to get a better idea of just what color might be assigned to a given wavelength. The variations in these color scales are downright humorous. The advent of LED's, however, and to a lesser extent (since they don't come in a wide range of colors), lasers, has helped a little here. [I've thought about buying a bunch of different LED's with known wavelengths might look to me and to others in comparison to other objects of similar color.]

Ken G
2009-Jun-12, 09:31 PM
What chick flick did that come from? Nah, I'm just speaking from experience, and making up the numbers.

George
2009-Jun-12, 10:14 PM
Nah, I'm just speaking from experience, and making up the numbers. I think I like your numbers over these studies (http://www.colormatters.com/khouw.html).

For instance...

Thomas, Curtis, and Bolton (1978) interviewed 72 Nepalese and asked them to list the names all the colors they could think of. There was a significant difference between men and women. Although, the women consistently listed more color names than men did, the cultural context of this study must be noted since Nepalese women traditionally wear more colorful clothing than men do. [my bold] It seems a little unfair when women are allowed to wear a cheat sheet. ;)

George
2009-Jun-12, 10:17 PM
I was going to ask you about that. Any chance of a bigger picture? It opens up too small, and zooming makes it too blurry. (Your two pics, on the other hand, turn out great).

Sure. This (http://www.photomeeting.de/astromeeting/constellations/040523cruxa1024.htm) is from his amazing website (http://www.photomeeting.de/astromeeting/_index.htm).