Yellow Plus Blue Makes Green

[HydrogenBond]

This idea came to me the other day and I thought it would make good topic.

 

Here is the experiment. I have two spot lights, one light gives off yellow light at 570 nm, and that other light gives off blue light at 475 nm. If we blend equal amounts of the yellow and blue light, we will get green light which is at about 510nm. Are the two light energy streams, averaging energy, to get the intermediate green energy level?

 

The analogy is we have two bottles of water, one is hot and is one cold and we mix them. Before we mix, we take an IR image of each. After we mix, the IR signal from the blended mixture will be in the middle since the temperature averages. We averaged the energy.

 

The next provocative question is a logical extrapolation. Say we blended visible light with the cosmic microwave background radiation. We blend higher and lower energy. Do we get a red shift. The analogy would be a dense drop of visual energy, dropped into a very dilute multi-light year pool of microwaves to get an average blend which cools the light.

[JMJones0424]

I'm afraid you are confusing additive colors like light with subtractive colors like paint. One would expect yellow light combined with blue light to produce a whitish light. I think this is solely an artifact of our trichromatic color perception, though others here would be better able to answer that.

[Qfwfq]

Interaction between photons is indirect and quite negligible. It is the retina that produces an intermediate sensation when combined wavelengths are incident.

 

JM is right in saying that particular combination, in mixing of light, gives white, but in mixing of pigments I reckon it gives black (or grey).

 

The effect of mixing light can be (at least somewhat) acheived by finely accosting pigments instead of mixing them. It can also be obtained (very well) on the PC screen by using RGB notation.

 

O + O = O

 

O + O = O<--white

[JMJones0424]

...that particular combination [yellow and blue], in mixing of light, gives white, but in mixing of pigments I reckon it gives black (or grey).

No need to reckon at all, the mixing of yellow and blue pigments most certainly gives green, not black or grey, as anyone who has ever mixed paint could tell you.

 

Why it is that additive and subtractive colors are not "opposite" as you assumed is something I have often wondered. Perhaps if we were bi-chromatic it would be?

[Ken]

This idea came to me the other day and I thought it would make good topic.

 

Here is the experiment. I have two spot lights, one light gives off yellow light at 570 nm, and that other light gives off blue light at 475 nm. If we blend equal amounts of the yellow and blue light, we will get green light which is at about 510nm. Are the two light energy streams, averaging energy, to get the intermediate green energy level?

 

The analogy is we have two bottles of water, one is hot and is one cold and we mix them. Before we mix, we take an IR image of each. After we mix, the IR signal from the blended mixture will be in the middle since the temperature averages. We averaged the energy.

 

The next provocative question is a logical extrapolation. Say we blended visible light with the cosmic microwave background radiation. We blend higher and lower energy. Do we get a red shift. The analogy would be a dense drop of visual energy, dropped into a very dilute multi-light year pool of microwaves to get an average blend which cools the light.

 

I'm back. :D

 

The two primary (pun) issues are the spectral purity of the light sources and the absorption spectra of the Cone species.

 

The experiment requires that you use a narrow bandwidth monochrome source of light. Your unfiltered spot lights will be very impure. You would need, at a minimum, selected interference filters to get the monochromatic light that you would get from a spectrophotometer. In fact I've used the cannabalized output heads of spectrophotometers in research when I wanted a large range of wavelengths without using all of the filters in the lab.

 

On the Cone side the chemicals (pigments), that are analogical to Rhodopsin in the Rods, are generally called Iodopsin and more specifically (from Rushton) Erythrolabe, Chlorolabe, and Cyanolabe - freely translated as Red-catching, Green-catching, and Blue-catching. These are not monochromatic, but have relatively narrow absorption curves centered on a narrow range of wavelengths.

 

 

The image uses the unimaginative labels S (short wavelength), M (Med.), and L (Long) in place of Rushton's more elegant terminology. B)

 

Our perception of color basically comes from a comparison of signal strength from each of the light absorbing pigments.

 

Extend a line upward from any wavelength and you will see the proportion of signals from the pigmented cones to give the perception of a particular color. There are points that give the same proportionality at different wavelengths and these provide indistinguishable color perceptions, but dissimilar brightness perceptions. Those common proportionality points would only occur where less than three types of cones are stimulated.

 

Color "blindness", or color defective vision, depends on the absence of one or more of the three pigment types.

 

The second part of your question deals again with absorption spectra. The visual pigments are effectively transparent to Cosmic radiation.

[Ken]

No need to reckon at all, the mixing of yellow and blue pigments most certainly gives green, not black or grey, as anyone who has ever mixed paint could tell you.

 

Why it is that additive and subtractive colors are not "opposite" as you assumed is something I have often wondered. Perhaps if we were bi-chromatic it would be?

 

 

Mixing paint pigments affects the reflectivity of the light. Reflected light is colored because the pigment absorbs most of the off-wavelengths and only reflects the unabsorbed wavelengths, that's why its called subtractive mixing.

 

Additive mixing combines wavelengths from different sources, or of different relative spectra, to provide mix which stimulates different Cone receptors in the manner I described above.

 

My Master's thesis involved mixing monochromatic Red light with Monochromatic Green light to produce a stimulus that subjects would match to a spectral Yellow by adjusting the relative flux of the two test stimuli. At match the subjects could not distinguish between the adjustment-based color they produced and the spectral color we provided. This was part of a more involved question that dealt with explaining a relatively rare form of color defect called Anomalous Trichromatism. But that story is a horse of a different color. :rolleyes:

[Qfwfq]

No need to reckon at all, the mixing of yellow and blue pigments most certainly gives green, not black or grey, as anyone who has ever mixed paint could tell you.

Perhaps they were mixing yellow and cyan?

 

Why it is that additive and subtractive colors are not "opposite" as you assumed is something I have often wondered. Perhaps if we were bi-chromatic it would be?

In mixing pigment, the primaries are yellow, cyan and magenta. Yellow pigment mostly absorbs blue light, cyan pigment mostly absorbs red light, so neither absorbs green light and voilà... that's how it works.

 

 

Mixing paint pigments affects the reflectivity of the light. Reflected light is colored because the pigment absorbs most of the off-wavelengths and only reflects the unabsorbed wavelengths, that's why its called subtractive mixing.

That's what I was reckoning on, with the fact that blue and yellow are complementary to each other. Do you draw the same conclusion?

 

One note: Strictly, the light not absorbed isn't reflected by a pigment, it is scattered. You see the material's colour best when it is matt; the more it is mirror-like polished, the less it alters the colour of what you see reflected in it (but of course the best mirrors use metals with a flat spectrum).

Edited November 24, 2010 by Qfwfq
blithering dumb error

[Ken]

Perhaps they were mixing yellow and cyan?

 

In mixing pigment, the primaries are yellow, cyan and magenta. Yellow pigment mostly absorbs blue light, cyan pigment mostly absorbs red light, so neither absorbs green light and voilà... that's how it works.

 

 

That's what I was reckoning on, with the fact that blue and yellow are complementary to each other. Do you draw the same conclusion?

 

A good explanation but perhaps a better Physics explanation is obtained by over-laying the spectral distribution curves of the mixed colors. The only light reflected/scattered is in the wavelengths not absorbed. Complementary colors are more on the subjective side, related to a subjective arrangement of color sensations around a color wheel or Chromoticity diagram.

 

One note: Strictly, the light not absorbed isn't reflected by a pigment, it is scattered. You see the material's colour best when it is matt; the more it is mirror-like polished, the less it alters the colour of what you see reflected in it (but of course the best mirrors use metals with a flat spectrum).

 

Point well-taken. Absolutely correct.

[JMJones0424]

Perhaps they were mixing yellow and cyan?

 

No. I said yellow and blue, like the OP. Yellow pigment combined with blue pigment does not produce black or grey, it produces green.

[Ken]

No. I said yellow and blue, like the OP. Yellow pigment combined with blue pigment does not produce black or grey, it produces green.

 

 

A lot of the confusion comes from color names and issues of Saturation of color. Monochromatic light is highest saturation, i.e., deepest shade of the wavelength. When the bandwidth is enlarged or there are other bandwidths included the color is said to be de-saturated.

 

It's probably best explained by the CIE Chromaticity diagram. Personally I understand this at a conceptual level but my math/geometric skills (lack) don't let me really comprehend at any deep level.

 

Here's a pretty standard definition of the diagram:

 

Chromaticity is commonly defined as the quality of color characterized by its dominant or complementary wavelength and purity taken together. This chromaticity chart plots the human tristimulus color space according to:

 

x = X/(X + Y + Z),

y = Y/(X + Y + Z).

 

Colors along the periphery of the "horseshoe" plot (spectral locus) are pure, monochromatic, saturated colors. Less saturated colors approach the center becomes (i.e. it becomes whitish or grayish). The straight line across the bottom is called the "non-spectral line of purples."

 

Here's what it looks like:

 

 

As you can see, blue is an area of that diagram. Cyan would be a specific location in the blue space.

 

Drawing a line between almost any two points on the periphery will transect the diagram, and the depth into the space would indicate less saturation than the periphery, and the specific point within the diagram for a given hue would be a function of the proportional energies.

 

Finally, since pigment mixing is a function of absorption of light, the green produced by any combination of a color named yellow and pigment color named blue/cyan would reflect/scatter less energy than in the original energy from the light source. The resulting green would be "less" green, or greenish-gray. The more subtractive mixing, the more blackish-green the color would appear to the eye.

 

I'd call it a draw. :D

 

Second edit:

 

I learned the practical side of this taking watercolor lessons. The richest black is about 2/3 green and 1/3 purple (actually non-spectral mix of red and blue).

 

For a non-Math type like me it's always an almost epiphanic moment when a mathematically-derived prediction is verified materially. :rolleyes:

[Qfwfq]

No. I said yellow and blue, like the OP. Yellow pigment combined with blue pigment does not produce black or grey, it produces green.

Call the colours what you like but here is CYAN and here is BLUE, try mixing pigments of each with YELLOW and tell us how green it gets in the case of yellow with actual, true blue.

 

A good explanation but perhaps a better Physics explanation is obtained by over-laying the spectral distribution curves of the mixed colors. The only light reflected/scattered is in the wavelengths not absorbed.

That's pretty much how my own explanation worked.

 

Complementary colors are more on the subjective side,

Why subjective? The complementery of blue is a mix of reed and green, which is yellow. It isn't the same for other species but most people (i. e. without impairments) have the same RGB receptors so I wouldn't really call it subjective. At that point, if you take any colour in the three byte notation, such as 1AE973 and subtract it from FFFFFF (white) you get its complementary, which is E5168C.

 

I made a big bitmap in which single pixels alternate, chessboard style, between these two colours. Since accostment doesn't actually add, it makes the average of the two and can't reach the effect of full white; the effect is like halfway grey but enlarging it in a typical image viewer can make the pixels distinguishable. The actual visual effect depends on the rendering used though and with LCD screens you may need to view from right angles. For comparison I made one in which each pixel is halfway grey 7F7F7F and the two look most alike when the viewer app is set to smooth pixels.

[Ken]

Call the colours what you like but here is CYAN and here is BLUE, try mixing pigments of each with YELLOW and tell us how green it gets in the case of yellow with actual, true blue.

 

That's pretty much how my own explanation worked.

 

Why subjective? The complementery of blue is a mix of reed and green, which is yellow. It isn't the same for other species but most people (i. e. without impairments) have the same RGB receptors so I wouldn't really call it subjective. At that point, if you take any colour in the three byte notation, such as 1AE973 and subtract it from FFFFFF (white) you get its complementary, which is E5168C.

 

I made a big bitmap in which single pixels alternate, chessboard style, between these two colours. Since accostment doesn't actually add, it makes the average of the two and can't reach the effect of full white; the effect is like halfway grey but enlarging it in a typical image viewer can make the pixels distinguishable. The actual visual effect depends on the rendering used though and with LCD screens you may need to view from right angles. For comparison I made one in which each pixel is halfway grey 7F7F7F and the two look most alike when the viewer app is set to smooth pixels.

 

Great examples. Thanks.

 

My point about the somewhat subjective nature of the perception of complemenataries goes back to the source data for the CIE diagram. It defines the human tristimulus map. Its derived from subjective color match data.

 

But I suspect that you have a much better handle on the mathematical aspects of the diagram than I ever did. I think :) that means that we are pretty much in agreement on all of the vital issues but view them from very different perspectives. Thanks for sharing your perspective, it is very helpful for me.

[HydrogenBond]

In the example of mixing cold water with hot water to get warm water, the IR signals will average and not retain their input characteristics. This is due to matter mediating the energy addition by averaging vibrational/rotational energy states.

[CraigD]

Photons, spectra, and reminiscences of my Things of Science spectroscope

 

Here is the experiment. I have two spot lights, one light gives off yellow light at 570 nm, and that other light gives off blue light at 475 nm. If we blend equal amounts of the yellow and blue light, we will get green light which is at about 510nm. Are the two light energy streams, averaging energy, to get the intermediate green energy level?

JM, Q, and Ken have made an excellent answer and explanation, but perhaps too detailed and not addressing your speculation directly enough.

 

The light of a equal blend of light from a monochrome 570 nm wavelength yellow light source and a monochrome 475 nm blue one consists of roughtly equal numbers of about 2.18 eV and 2.61 eV photons, not 2.43 eV (510 nm wavelength) photons. Photons are gauge bosons, and obeying their statistic, don’t interact directly with one another (though they can interfere with one another to produce interference patterns), and thus can’t change their wavelength/frequency/energy.

 

You can see this experimentally with a simple, cheap spectrometer. If photons “mixed” in the way your describe, HBond, a simple spectrometer would show a single narrow spectral band (at 510 nm wavelength, though with the simplest spectrometer, it’s difficult to measure precise spectra wavelengths). However, what you will actually observe are two bands (at 570 and 475 nm).

 

Note that, to simplify the discussion, I’ve rather sneakily accepted your specification of “spot lights” that emit monochrome light. Most everyday spotlights – the sort that we illuminate stages from the little one in my basement to huge stadium venues – are far from monochrome, consisting of a tungsten filament lamp which produces a wide, continuous spectrum of light, covered with a gel that absorbs some, but far from all, of the wavelengths of light other than that of their intended output color. Rather than a spectrum consisting of a single sharp band, theirs have fat, rounded “humps” spanning many wavelengths. The preceding wikipedia link has an example of a standard gel absorption spectrum, and its output for a sunlight light source – a more detailed discussion can be found at various gel manufacturers’ sites, such as this Rosco guide manual (PDF).

 

The analogy is we have two bottles of water, one is hot and is one cold and we mix them. Before we mix, we take an IR image of each. After we mix, the IR signal from the blended mixture will be in the middle since the temperature averages. We averaged the energy.

In the example of mixing cold water with hot water to get warm water, the IR signals will average and not retain their input characteristics. This is due to matter mediating the energy addition by averaging vibrational/rotational energy states.

This is true, because water consists of atoms consisting of fermions, which follow statistics that allow them to interact with bosons, and each other, which photons alone – light – cannot do. The spectrum of the light the mixed, warm water emit is, thus, as you say, different than that of the cold and hot water before they are mixed.

 

Spectra are very interesting, and, in the visible range, easy to investigate at home. One of my earliest “science-y” memories is of using a home-made (from a ca. 1970 Things of Science kit) cardboard tube slit spectrometer to look at the spectra of lightbulbs, florescent bulbs, and after some cajoling to get my parents to drive me around town at night, sodium and mercury vapor streetlamps and neon signs.

 

:) Things of Science was one of the monthly delights of my childhood. I was sad to discover that this subscription program ended in 1989, a few years too early for me to get them for my own kids. :(

[Ken]

Photons, spectra, and reminiscences of my Things of Science spectroscope

 

 

JM, Q, and Ken have made an excellent answer and explanation, but perhaps too detailed and not addressing your speculation directly enough.

 

The light of a equal blend of light from a monochrome 570 nm wavelength yellow light source and a monochrome 475 nm blue one consists of roughtly equal numbers of about 2.18 eV and 2.61 eV photons, not 2.43 eV (510 nm wavelength) photons. Photons are gauge bosons, and obeying their statistic, don’t interact directly with one another (though they can interfere with one another to produce interference patterns), and thus can’t change their wavelength/frequency/energy.

 

You can see this experimentally with a simple, cheap spectrometer. If photons “mixed” in the way your describe, HBond, a simple spectrometer would show a single narrow spectral band (at 510 nm wavelength, though with the simplest spectrometer, it’s difficult to measure precise spectra wavelengths). However, what you will actually observe are two bands (at 570 and 475 nm).

 

Note that, to simplify the discussion, I’ve rather sneakily accepted your specification of “spot lights” that emit monochrome light. Most everyday spotlights – the sort that we illuminate stages from the little one in my basement to huge stadium venues – are far from monochrome, consisting of a tungsten filament lamp which produces a wide, continuous spectrum of light, covered with a gel that absorbs some, but far from all, of the wavelengths of light other than that of their intended output color. Rather than a spectrum consisting of a single sharp band, theirs have fat, rounded “humps” spanning many wavelengths. The preceding wikipedia link has an example of a standard gel absorption spectrum, and its output for a sunlight light source – a more detailed discussion can be found at various gel manufacturers’ sites, such as this Rosco guide manual (PDF).

 

 

 

This is true, because water consists of atoms consisting of fermions, which follow statistics that allow them to interact with bosons, and each other, which photons alone – light – cannot do. The spectrum of the light the mixed, warm water emit is, thus, as you say, different than that of the cold and hot water before they are mixed.

 

Spectra are very interesting, and, in the visible range, easy to investigate at home. One of my earliest “science-y” memories is of using a home-made (from a ca. 1970 Things of Science kit) cardboard tube slit spectrometer to look at the spectra of lightbulbs, florescent bulbs, and after some cajoling to get my parents to drive me around town at night, sodium and mercury vapor streetlamps and neon signs.

 

:) Things of Science was one of the monthly delights of my childhood. I was sad to discover that this subscription program ended in 1989, a few years too early for me to get them for my own kids. :(

 

This is great addition to the thread. I find it really stimulating to see how several different disciplines contribute to understanding of a problem of this nature. Reading these explanations that come from different perspectives has broadened my knowledge of a field I spent quite a bit of energy learning about.

[granpa]

green is between yellow and blue as you go around the colors of the rainbow

 

FF0000 red

FFFF00 yellow

00FF00 green

00FFFF green-blue

0000FF blue

FF00FF purple

 

http://en.wikipedia.org/wiki/Web_colors#HTML_color_names

[Ken]

green is between yellow and blue as you go around the colors of the rainbow

 

FF0000 red

FFFF00 yellow

00FF00 green

00FFFF green-blue

0000FF blue

FF00FF purple

 

http://en.wikipedia.org/wiki/Web_colors#HTML_color_names

 

 

:D

 

To paraphrase Shakespeare: There is more to visual perception than a computer screen.

 

But even with a computer screen, you don't go "around" colors of the spectrum, you go across them from short wavelengths to longer wavelengths - approx 430 nm to about 750 nm.

 

Everything you've written has been perfectly sound for computer screens/TV screens that provide light stimuli. But human and animal photic perception is based on another sort of "engineering", different sensors, different sensor distributions, different signal "multiplexing" (or coding), different kinds of analyses, and a function of both long-past and immediate-past data streams.

 

You could do a schematic or functional algorithm for the various components of image detection, combination, or isolation and come up with a block diagram that could lead to effective code for a computer device but it wouldn't exactly replicate the animal experience for a variety of reasons. Primarily because the coding into bio-messages is fundamentally not the same as in strictly electrical messages.

 

Translating one set of symbols (ROYGBIV) into Hex symbols really doesn't provide any more useful information. You can "count" the pieces of data in the flow with any base numerical system. It simply makes certain kinds of descriptions more or less convenient for specific purposes.

 

But all of that does not reduce my respect and admiration of the technical skills that have led to the remarkable level of success in forming image displays (stimuli) on monitor screens. As a Basketball fan I can still be impressed by an awesome Tennis serve. :)