Why there is no frequency for the color brown?

[DaTario]

Hi All,

Is it true that the brown color has not a single frequency attached to it?DaTario

 

[Dale]

The same is true of white, or gray, or purple.

 

[Drakkith]

The only colors which have a corresponding frequency(or rather a small, continuous range of frequencies) are spectral colors.

 

[sophiecentaur]

The only colors which have a corresponding frequency(or rather a small, continuous range of frequencies) are spectral colors.

Doesn't it just make you cross that even 'reputable' Science sources confuse wavelength and colour?

 

[Drakkith]

Doesn't it just make you cross that even 'reputable' Science sources confuse wavelength and colour?

Indeed. It makes me 655 nm in the face.

 

[DaTario]

Indeed. It makes me 655 nm in the face.

Is it correct to say that there are colors which are not expressible in terms of a thin an monomodal spectrum?

 

[Dale]

Is it correct to say that there are colors which are not expressible in terms of a thin an monomodal spectrum?

Yes, see my previous reply for a few examples of such non spectral colors.

 

[DaTario]

I saw it. Thank you for sharing, Dale. So it seems that the concept of color has something of the physiology of eye inside, isn´t it?
I have heard of metamerism, i.e., visible radiations with different spectra but recognized by the eye as having equal color. Does it imply that the color concept does not have even an objective association with spectrum?

 

[pixel]

I saw it. Thank you for sharing, Dale. So it seems that the concept of color has something of the physiology of eye inside, isn´t it?
I have heard of metamerism, i.e., visible radiations with different spectra but recognized by the eye as having equal color. Does it imply that the color concept does not have even an objective association with spectrum?

The measurement of color makes use of the reflectance of the object (spectrum), the light source and the three color response curves (color matching functions) for human vision.

 

[Drakkith]

I saw it. Thank you for sharing, Dale. So it seems that the concept of color has something of the physiology of eye inside, isn´t it?
I have heard of metamerism, i.e., visible radiations with different spectra but recognized by the eye as having equal color. Does it imply that the color concept does not have even an objective association with spectrum?

The perception of color is extremely complicated and involves both the eye and the brain. It is also very much subjective. For example, my stepfather is red-green colorblind. To him, a green streetlight at an intersection looks almost exactly like the white streetlights that light up parking lots and such. The green streetlight is just a bit dimmer than the white lights. So for him the spectral color green looks very much like a non-spectral color. He also has difficulty seeing the color yellow. The standard wavelengths given for each spectral color don't apply to him at all.

In addition, even for color-normal people, the brain actively works to make sure that your perception of color changes as little as possible between various environmental light sources. Your perception of the color of your shirt is probably very similar no matter if you're sitting under white fluorescent lights in an office or under a yellowish incandescent at home.

 

[Khashishi]

There are tetrachromatic women who see more colors than most of us. So if you are not so lucky, you see many things as the same color, which they see as different colors.

 

[sophiecentaur]

I saw it. Thank you for sharing, Dale. So it seems that the concept of color has something of the physiology of eye inside, isn´t it?
I have heard of metamerism, i.e., visible radiations with different spectra but recognized by the eye as having equal color. Does it imply that the color concept does not have even an objective association with spectrum?

It's always worth while referring to the https://www.siggraph.org/education/materials/HyperGraph/color/colorcie.htm when discussing colorimetry. It's really worth reading around all the sites that a Google search will throw up. Visible colours (the only colours that exist because colours are in your head) The chart is a totally artificial representation of our (average) perception of colours and mixes of colours and I'm sure there are purists who will cast doubt on such a simplistic representation but the whole mechanism is pretty fuzzy and the accepted human colour analysis curves are only based on 'average' performance.
I'm not sure what you are suggesting here, exactly but it is not actually possible to produce a metemeric match of a spectral colour (i.e. a narrow band of frequencies) by using any combination of primaries because the locus of the spectral colours on the CIE chart is a curve and any straight line joining two primaries will not pass through the required spectral colour. (on the actual curve) You can produce a 'good' yellow (association?) with equal values of R and G signals applied to TV tube primaries but it does not lie on the CIE curve. It will be a 'desaturated' version of spectral yellow and an observer will be able to tell the difference.
There is a difference between how you do colorimetric research and how you make a TV system. The 'best' primaries to use would be monochromatic and spectral but you would never get enough brightness from a narrow band synthesised primary so the ones used are on the vertices of triangles and the spectra of those primaries are far from monochromatic. The gamut of obtainable metameric matches is limited to those colours within the triangle. Note the number of bright clothing colours at sporting events that appear to be the same on the screen. That (imo) must be because they are very saturated and lie outside the obtainable gamut and turn up on the sides of that triangle. If you were at the venue, you would distinguish between all those different bright red kagouls.

 

[LaplacianHarmonic]

You are not studying the correct category. You have to look into the wavelength.

It's a simple matter of looking it up.

 

[sophiecentaur]

You are not studying the correct category. You have to look into the wavelength.

To whom is your remark made?

 

[rumborak]

Color perception is very complex indeed. Just look at that picture that was going around with the yellow (or blue?) dress. The wavelengths and strengths were exactly the same, but to some it looked yellow, to some it looked blue (mostly dependent on where they were looking at the picture).

 

[sophiecentaur]

Color perception is very complex indeed. Just look at that picture that was going around with the yellow (or blue?) dress. The wavelengths and strengths were exactly the same, but to some it looked yellow, to some it looked blue (mostly dependent on where they were looking at the picture).

The Land Retinex theory of colour vision attempts to quantify how we actually perceive colour. See this link and then Google the terms: Land, Retinex Mondrian. It is fascinating and it amazes me that, in the light of his work, colour TV works so well!

 

[pixel]

Color perception is very complex indeed. Just look at that picture that was going around with the yellow (or blue?) dress. The wavelengths and strengths were exactly the same, but to some it looked yellow, to some it looked blue (mostly dependent on where they were looking at the picture).

I think that was just a matter of whether someone's browser was properly color managed and their monitor calibrated (so as you say, where they were looking at it). It didn't really have to do with color perception per se.

 

[Drakkith]

I think that was just a matter of whether someone's browser was properly color managed and their monitor calibrated (so as you say, where they were looking at it). It didn't really have to do with color perception per se.

Sure it did. People would get into arguments over the color after looking at the picture on the same monitor.

 

[pixel]

Sure it did. People would get into arguments over the color after looking at the picture on the same monitor.

That's news to me. There are variations of vision from person to person, but would people looking at the same monitor disagree about whether the dress were yellow or blue, in such large numbers as to become an internet sensation?

 

[nasu]

Yes, they did. I have seen this at work. The same picture on the same phone. Every body, including visitors, was asked about what he/she sees. Definitely two groups, the white-gold group maybe slightly more numerous.

 

[sophiecentaur]

That's news to me. There are variations of vision from person to person, but would people looking at the same monitor disagree about whether the dress were yellow or blue, in such large numbers as to become an internet sensation?

I think people should read about Land's work in detail before questioning his ideas too much. What he found out was not an "internet sensation'. The work was done decades ago.
Colour TV works because it only deals with a limited range of conditions and the sort of 'illusions' that Land came up with are not noticed in everyday life. TV technology is only interested in 'colour fidelity' and presents a limited range of scenes. What our brains happen to make of a number of scenes that are not commonly encountered is a different matter. The relatively simple tristimulus colour vision system is not a complete model. It has nothing to say about what colours will look like in the presence of other nearby coloured areas - it just gives the same RGB values under all conditions (hopefully). A simple way of putting Land's theory is that the eye 'integrates to grey'. We try to eliminate the effects of the illuminant (colour camera white balance settings etc.) and we also try to reduce the colours any scene to something familiar.

 

[DaTario]

TV technology is only interested in 'colour fidelity' and presents a limited range of scenes.

What do you mean by "a limited range of scenes"?

 

[sophiecentaur]

What do you mean by "a limited range of scenes"?

Firstly, additive colour mixing can only represent colours that lie within the triangle with the phosphors at the vertices. That takes care of enough scenes to satisfy most viewers that the coloured regions of the scene are being portrayed 'well enough'. Secondly, the system will just pass on the sort of illusions that Land was producing in his experiments. An isolated patch of colour in a scene will be reproduced the same, whatever colours happen to surround it. It can't compensate for how the brain will experience it.

 

[Drakkith]

Firstly, additive colour mixing can only represent colours that lie within the triangle with the phosphors at the vertices. That takes care of enough scenes to satisfy most viewers that the coloured regions of the scene are being portrayed 'well enough'.

Indeed. This may shock some, but you absolutely cannot render the color violet on a screen. Purple, yes. But not violet. The two are, in fact, different colors.

 

[sophiecentaur]

Indeed. This may shock some, but you absolutely cannot render the color violet on a screen. Purple, yes. But not violet. The two are, in fact, different colors.

Our brain will 'match' spectral violet (outside the triangle) with a colour on the screen that's produced with a lot of B and a smidgen of R and G (inside the triangle).
But that really isn't a problem in most instances because we just don't come across spectral violet (or any other of the colours that are outside the triangle of primaries) in everyday life. The exceptions are when we look at some laser light and very bright saturated colours which are used for clothing and we are just not fussy about the resulting distortion of the colours in a scene. Colour TV has managed to present us with a very good 'near enough' solution. Take your hats off to the early pioneers of colour TV for squeezing all that information into the bandwidth of a conventional TV signal. (And all with analogue processing, too!)

 

[Eric Bretschneider]

Brown is a relative color. It is a lower intensity amber or orange. It can only be perceived in the context of imagery that includes fields of higher intensity colors.

Gray is similar, in that it can only be perceived relative to higher intensity colors/fields. So while a gray image might have the same relative spectral power distribution as the white paper it is printed on, the radiant flux from the gray regions will be lower than the radiant flux from the white paper.

 

[Eric Bretschneider]

Indeed. This may shock some, but you absolutely cannot render the color violet on a screen. Purple, yes. But not violet. The two are, in fact, different colors.

You can in fact render violet, but not at 100% saturation. Then again it is virtually impossible to render any color at 100% saturation on a display. Doing so would require true monochromatic primaries and then you would only render the three primaries at 100% saturation.

The curved outer boundary of a chromaticity diagram represents the limit of 100% saturated colors. The diagonal line on the lower left boundary is the locus of complimentary colors. They are 100% saturated, but require mixtures of violet and red light.

Brown is a special case of amber/orange. It can in fact be 100% saturated, but to "see" brown it has to be in the context of an image with areas of higher luminosity.

 

[Buzz Bloom]

Yes, they did. I have seen this at work.

Yes. I saw demonstrations of this by Land I think in the 1960s. It was quite amazing. One display was of an American flag. The actual colors used to make the display were two different spectral yellows. About half the audience saw approximately normal (not as intense) red, white, and blue colors, and half saw odd colors. What each person saw was stable for them. No one saw yellow.

 

[sophiecentaur]

They are 100% saturated, but require mixtures of violet and red light.

The sources would have to be spectral, too. You couldn't;t get away with the normal trick of fairly broad band phosphors as in TV.

 

[Lish Lash]

Brown does in fact have a central wavelength. It is a dark, desaturated shade of orange, around 620 nm.