Blackbody colour of metals versus stars

[Nathi ORea]

Hello. I am new to this forum and joined because I am at home nerding out trying to work something out.

Why do white hot metals seem to be much cooler than white hot stars. The attached picture is from Wikipedia relating temperature of a hot metal to its temperature.

For example a red giant or red dwarf has a temperature of many thousands of degrees but clearly shines red. Whereas metals only need to be a little over 1,3oo C to glow white.

My understanding was the colour at which something glowed related only to temperature and not to the material it is made from?

Thanks.

Naithi

 

[DrStupid]

I think it's just color balance. Both spectra are actually not white but a section from a black body spectra. If you could compare a red dwarf with white glowing metal it would appear blue (or the metal would appear red, depending on what your eyes decide to be white).

 

[Nathi ORea]

That is a really interesting perspective. Thanks so much for your reply.

I guess our eyes do see colour in relative terms rather than absolute terms sometimes. Perhaps a bit like perfect pitch and relative pitch in music.

 

[sophiecentaur]

Our colour vision has evolved in order to assign a 'colour' to an object under illumination from the Sun and not to cope with hot luminous bodies (illuminants). There's no reason to expect us to be able to deal with stars, the Sun and glowing metals in any reliable way.

The apparent colour of astronomical bodies is a particularly hard one as (apart from the Sun and Moon) they are only seen under low light conditions. We see low light scenes mainly with our monochrome Rod Cells. Keen astronomers make a big thing of the colours of stars. They get very excited about it and have big arguments about appropriate names to give star colours. OTOH, once you use a long exposure or a high gain photographic array objective lens for an astrophotograph, you can see a very impressive range of colours (modified of course, by the analysis curves of the sensors and filters used.. So much so that the images can look really 'unnatural'.
Edit: A spectrometer will give as accurate a spectral measurement as the money you care to spend on it. (That goes for hot metals too)

 

[Nathi ORea]

Thank you so much for your reply.

All that absolutely makes so much sense.

I did realize that astro photography gives a totally ‘unnatural’ view of what objects would look like with your own eyes, like nebulae and galaxies and stuff. This actually reminds me of another thing I would like to put to the forum I have been wondering for while.

I think it is probably interesting to note too that only the brighter red stars we see appear red with the naked eye. The fainter ones look white.

I really appreciate the answer. This was my first PhysicsForums question and I think you guys hit it out of the ball park.

Interesting note on the spectrometer. I am a teacher and we have the cheap black plastic box spectroscopes. Probably wouldn’t quite cut it.. lol

 

[sophiecentaur]

I am a teacher and we have the cheap black plastic box spectroscopes.

You should try them on your binoculars. (one eye). Use a bright star that's out on its own and look at what you get. I bought a cheap spectroscope which is sold for gemmology. It's about as big as a pen top but it's amazing what it shows. I can see absorption bands in sunlight (compared with filament bulb that has none). Stars are more difficult but a teacher has access to clamp stands etc and a crude mount will give you a useable setup. All you need is a dark, clear night.

 

[Nathi ORea]

I actually have a 8 inch Dobsonium telescope. I'll have to try it over the holidays.

Thanks so much!

 

[sophiecentaur]

I actually have a 8 inch Dobsonium telescope. I'll have to try it over the holidays.

Thanks so much!

As they say "It. would be rude not to!"
If you have that light bucket then you can fit the spectroscope inside the focuser with rolled up cardboard.

Good luck.

 

[Drakkith]

Look at an incandescent light bulb from a distance in the dark. I'm betting it won't quite be pure white. It is possible for the light to overwhelm all three types of cone cells in your eye if the source is too bright or you're too close, which would make it look white no matter what the 'true' color is.

 

[sophiecentaur]

Look at an incandescent light bulb from a distance in the dark. I'm betting it won't quite be pure white. It is possible for the light to overwhelm all three types of cone cells in your eye if the source is too bright or you're too close, which would make it look white no matter what the 'true' color is.

Accurate colour perception just HAS to be in the linear region of the receptors and you need to bear in mind that there is no "true" colour of anything and definitely not a "pure white". White is what you get when the reflected components excite three colour sensors equally. That only happens for one particular illuminant but you can fudge it to deal with alternative lighting. Controls on a camera have settings of Auto, different colour temperatures, 'flouorescent' etc. and they attempt to re-centre, the colorimetry on an 'agreed' illuminant source spectrum (a sort of Daylight). You eyes do the same thing fairly successfully except when the context of a scene fools you. As it happens, I was on FaceTime with my Granddaughter yesterday and she was playing with a red LED torch. When she pointed it at her face, the colour was red but, when she pointed it at the camera, it had a white (overload) region at its lens and red nearby. The sensors are pretty linear and then just limit on '255'. Light from almost any star will be near the bottom end of the three response curves so, as well as the luminance dominating what you see, there will be imbalance in your assessment of the ratios of the three colour sensations. We just never evolved to be astronomers, lol.

 

[Eric Bretschneider]

Accurate colour perception just HAS to be in the linear region of the receptors and you need to bear in mind that there is no "true" colour of anything and definitely not a "pure white". White is what you get when the reflected components excite three colour sensors equally. That only happens for one particular illuminant but you can fudge it to deal with alternative lighting. Controls on a camera have settings of Auto, different colour temperatures, 'flouorescent' etc. and they attempt to re-centre, the colorimetry on an 'agreed' illuminant source spectrum (a sort of Daylight). You eyes do the same thing fairly successfully except when the context of a scene fools you. As it happens, I was on FaceTime with my Granddaughter yesterday and she was playing with a red LED torch. When she pointed it at her face, the colour was red but, when she pointed it at the camera, it had a white (overload) region at its lens and red nearby. The sensors are pretty linear and then just limit on '255'. Light from almost any star will be near the bottom end of the three response curves so, as well as the luminance dominating what you see, there will be imbalance in your assessment of the ratios of the three colour sensations. We just never evolved to be astronomers, lol.

It is important to keep in mind that our visual system attempts a sort of "white balance" - that is to say that our eyes adapt to the lighting conditions and attempt to render things with "normal" colors. The overall level of illumination is very important for this to happen.

Daylight illumination (direct sun + sky) is equivalent to a black body temperature of about 6500 K. It will look white to us. However, at night, the exact sample spectrum will look decidedly blue (especially if it is providing low level illuminance).

I work with semiconductor growth and can attest that a large area (disc 300mm diameter) of material at ~1400 K looks rather white, even under normal room illumination. This is despite the fact that a dimmed incandescent lamp (~2200 K) looks distinctly yellow. Due to the large area of the disc, it provides a great deal of illumination and our eyes make an effort to color balance.

A bit of handwaving for the explanation, but visual adaption is an important issue related to the OP. Starlight is by default results in very low levels of illumination, hence we see colors. On a planet orbiting a star, everything would appear more or less white.

 

[jim mcnamara]

This is the response curve of pigments in the rods and cones of eye - rods are black line, for night vision, turned off at higher light levels. This is what the other posters are mentioning. This is the "raw data" for color vision -- the three types of cones.

While this graph helps to explain things, like color-blindness, someone who is color blind will have trouble using the graph. Sort of paradoxical.

AFAIK there no complete understanding how this mish-mash gets turned into what we perceive. If you know computers - think "black box" post-processing is as close as we get at moment

Credit unm.edu: https://www.unm.edu/~toolson/human_cone_response.htm

Which you may want to read

 

[Eric Bretschneider]

The "black-box processing" you are referring to is known as chromatic adaptation. It has been extensively studied in color science for over 50 years. Suffice it to say that the phenomenon is relatively well characterized.

Color science is a rather understudied field. Most people working in technology have no reason to delve into the topic, so color science can indeed be perceived as a black art (pun intended).

 

[sophiecentaur]

a large area (disc 300mm diameter) of material at ~1400 K looks rather white, even under normal room illumination.

Perception of the reflected colour of a surface will depend not only on the surface characteristic but also on the illuminant and the other colours in the scene. If they all share the same illuminant then your eye does a process which is sometimes referred to as "integrates to grey". That's a rather glib description but it involves the brain assuming that the scene is a 'typical' one and that the colours will go together to produce an average of grey. Cameras often use that technique but some of them do much better than you might expect - smart processing, I guess. Your white panel will easily convince the brain that it is dealing with a uniform reflectance over the whole spectrum - so that's what you 'see'.
The brain is much smarter than that but it can cope well with sunlight at most times of day and also a lot of filament bulb temperatures. It doesn't cope well with fluorescent lights (the mainstay of shop lighting until quite recently) and can result in some bizarre clothes purchases when you got them home.

 

[Eric Bretschneider]

What you are talking about is also known as the visual triangle: source, surface, detector. When you look at a source directly, you take one leg out. Color science does apply to sources viewed directly.

 

[sophiecentaur]

What you are talking about is also known as the visual triangle: source, surface, detector. When you look at a source directly, you take one leg out. Color science does apply to sources viewed directly.

Yes. You can do colourimetric matching between light sources (TV, even) but it must be true to say that our colour vision evolved with your "visual triangle". I was going to say that our early selves only ever had one illuminant available at a time but living under a canopy of leaves would give a whole range of 'green lights' which we needed to cope with when looking for food and enemies (and for social interaction).

 

[sophiecentaur]

Most people working in technology have no reason to delve into the topic,

I just read this. Colour reproduction and display is a pretty vast field of technology and few of us are not affected by it. People are very fussy about the colours that their cameraphone can produce and cameras are everywhere in modern technology. It's a far more fruitful field to study than HiFi sound, I would suggest; you can SEE the results of improvements on your new TV better than you can HEAR what your hundreds / thousands of quid can achieve with the new sound system.

 

[jim mcnamara]

@Eric Bretschneider - you should cite several different papers to support various claims. We are a science forum after all.

Are you talking about von Kries transforms for example? We may be talking past each other. Or I was not clear, which is likely.

Paging @Andy Resnick

 

[Eric Bretschneider]

Yes. You can do colourimetric matching between light sources (TV, even) but it must be true to say that our colour vision evolved with your "visual triangle". I was going to say that our early selves only ever had one illuminant available at a time but living under a canopy of leaves would give a whole range of 'green lights' which we needed to cope with when looking for food and enemies (and for social interaction).

Our visual system evolved to operate under one illuminant (the sun), but that illuminant isn't static as perceived on the surface of the earth. The spectral power distribution changes markedly over the course of a day, even when the sky is clear. Cloudy weather complicate the scenario even more.

As you pointed out, there were ample reasons for our eyes to adapt. Can you image the difficulty in survival if you could only identify a predator under direct sunlight or food only when it was in a tree?

 

[Eric Bretschneider]

@Eric Bretschneider - you should cite several different papers to support various claims. We are a science forum after all.

Are you talking about von Kries transforms for example? We may be talking past each other. Or I was not clear, which is likely.

Apologies, but given the age of the research, quoting papers may not be the best approach. The critical work was done long before the information age and even before the digital age. The foundational work was completed when "computer" was a job description, not an electronic device.

I will offer this. It is considered the "Bible" of color science. The first edition was published in 1967. https://www.amazon.com/dp/0471399183/?tag=pfamazon01-20

Researches on Normal and Defective Colour Vision. W. D. Wright and L. C. Martin (1946)

Few people these days realize that Erwin Schrodinger also made major contributions to color science. He published three articles on the topic in Annalen Der Physik in 1920. Going further back you will find that Maxwell, Huygens, Geothe, and Newton all played a role in developing this field.

 

[Vanadium 50]

Going further back you will find that Maxwell, Huygens, Geothe, and Newton all played a role in developing this field.

And Thomas Young.

 

[jim mcnamara]

Yes. Those are valid. Thank you.

 

[jim mcnamara]

Try Biology, which I was trying unsuccessfully to convey:

J Anat. 2010 Oct; 217(4): 449–468.
Published online 2010 Aug 16. doi: 10.1111/j.1469-7580.2010.01275.x
PMCID: PMC2992420
PMID: 20722872
"Unravelling the development of the visual cortex: implications for plasticity and repair"
James A Bourne

The visual cortex comprises over 50 areas in the human, each with a specified role and distinct physiology, connectivity and cellular morphology. How these individual areas emerge during development still remains something of a mystery and, [snip]

This is what I thought - my position is biology, not necessarily Newton. I was not clear.

 

[Nathi ORea]

As they say "It. would be rude not to!"
If you have that light bucket then you can fit the spectroscope inside the focuser with rolled up cardboard.

Good luck.

Wow! That sounds pretty cool!

 

[Nathi ORea]

Look at an incandescent light bulb from a distance in the dark. I'm betting it won't quite be pure white. It is possible for the light to overwhelm all three types of cone cells in your eye if the source is too bright or you're too close, which would make it look white no matter what the 'true' color is.

I am definitely going to try this tonight! Thank you!

 

[Nathi ORea]

It is important to keep in mind that our visual system attempts a sort of "white balance" - that is to say that our eyes adapt to the lighting conditions and attempt to render things with "normal" colors. The overall level of illumination is very important for this to happen.

Daylight illumination (direct sun + sky) is equivalent to a black body temperature of about 6500 K. It will look white to us. However, at night, the exact sample spectrum will look decidedly blue (especially if it is providing low level illuminance).

I work with semiconductor growth and can attest that a large area (disc 300mm diameter) of material at ~1400 K looks rather white, even under normal room illumination. This is despite the fact that a dimmed incandescent lamp (~2200 K) looks distinctly yellow. Due to the large area of the disc, it provides a great deal of illumination and our eyes make an effort to color balance.

A bit of handwaving for the explanation, but visual adaption is an important issue related to the OP. Starlight is by default results in very low levels of illumination, hence we see colors. On a planet orbiting a star, everything would appear more or less white.

Thank you so much for that. This all very interesting.

Your last paragraph was actually something I was going to ask. It is funny that you said that. All you guys have me thinking and I was wondering if a red giant star would appear white to the eye if you were near it. Say the same angular diameter as the Sun has from Earth. I was thinking it would probably appear white going by what you guys are saying?

 

[Nathi ORea]

This is the response curve of pigments in the rods and cones of eye - rods are black line, for night vision, turned off at higher light levels. This is what the other posters are mentioning. This is the "raw data" for color vision -- the three types of cones.

While this graph helps to explain things, like color-blindness, someone who is color blind will have trouble using the graph. Sort of paradoxical.

AFAIK there no complete understanding how this mish-mash gets turned into what we perceive. If you know computers - think "black box" post-processing is as close as we get at moment

View attachment 273436Credit unm.edu: https://www.unm.edu/~toolson/human_cone_response.htm

Which you may want to read

That graph is really interesting. Thanks. I read your link.

So I guess we can't see longer wavelengths as well as ones around 5OO nm when just our rods are collecting light? Does this mean that red objects appear darker than green/blue objects in low light? So green/blue objects would appear 'brighter'?

 

[sophiecentaur]

So I guess we can't see longer wavelengths as well as ones around 5OO nm when just our rods are collecting light? Does this mean that red objects appear darker than green/blue objects in low light? So green/blue objects would appear 'brighter'?

If you look at the response curves, higher up the thread, you can see that sensitivity to reds peaks higher than the other two colour receptors. Those response curves are the result of a vast number of comparison tests (is A brighter than B? again and again with many different permutations) Ion order to get reliable information, the conditions would not have been under low light conditions because that would be yet another factor. At some low light level then your brain will switch to a low light mode and rather be interested in spotting details and movement etc., rather than colour.
This suggests a possible experiment that you could try on your own - in a cupboard - with various coloured objects. The Purkinje Effect is fairly well documented and it could well apply to viewing astronomical bodies. I'd bet money on finding that it varies more between individuals than colour perception under good lighting.