Talk:Additive color

I deleted the image I had created and posted many years ago. Now that the definition has changed/evolved, the image no longer adds enough value. Also it is incorrect: the text refers to the "tartan ribbon"-the image is of a girl. Not good. Also, a person could not sit still long enough to have 3 photos taken and be in register. Thus, poof, it should be gone. If someone can get the rights to post a copy of the real tartan ribbon image (which IS out there) that would be better.--Dkrolls 12:00, 15 May 2006 (UTC)

Additive vs. subtractive primaries

The article doesn't answer the one question that I had...why are the additive primary colors RGB whereas the subtractive primary colors are RYB. --grr 14:11, 5 December 2006 (UTC)

“Primary” colors are completely arbitrary. Red, green, and blue are used because they have several desired properties: they allow a gamut which roughly approximates the shape of the gamut of human vision, and mixing them in the proper proportions (in, e.g., a computer or television display) yields neutral colors. With ink, cyan, magenta, and yellow are used because they are reasonably well spaced around a perceptually-uniform hue circle, and they define a fairly large gamut. But many printing processes (or painters’ palettes) use more than 3 “primaries” for increased gamut, for instance Pantone’s Hexachrome. --jacobolus (t) 01:53, 6 May 2007 (UTC)

It says the the way humans perceive a mix of blue and green is yellow, but is unrelated to the yellow wavelength that is understandable but then it says that the photographer guy used color filters so somehow the yellow penetrated both the one which only allows green to go through and the one which only allows blue to go through this has nothing to do with human perception but light physics can someone explain?

A filter that we percieve as 'green' would be one that absorbs most of the reds and blues, but passess most of the greens. So it would also pass most of the yellow tones. Similar with the blue filter: it would also pass yellows. This is all further confused if the object you perceive as 'yellow' is actually reflecting that green plus blue, and could be absorbing yellows. You wouldn't know unless you used a chromatograph. 149.135.98.159 (talk) 11:56, 20 January 2008 (UTC)

The article is just wrong. Even the sentences itself lack logic. Saying:"It should be noted that additive color is a result of the way the eye detects color, and is not a property of light. There is a vast difference between yellow light, with a wavelength of approximately 580nm, and a mixture of red and green light. However, both stimulate our eyes in a similar manner, so we do not detect the difference." is saying the same as "it is a result of how we see but we can't see the difference". Better should be: "we can't see the difference if a color is a result of a subtractive or an additive process". In fact alot in the article is wrong.

No, this sentence that you find lacking in logic is in fact perfectly reasonable, and a decent description of how additive color works. --jacobolus (t) 20:26, 21 August 2007 (UTC)

This is very difficult to me: 1. you admit that we can't see the difference yet it is the result of how we see. In my logical thinking: if we can't see the difference is saying it is unrelated 2.You seem to agree with my definition of additive / subtractive below (point3). Yet insist that it's something different too. Maybe I'm wrong. Could you explain in full that "additive color is the result of how we see"? I mean, if we would see differently (maybe animals, bees aso) then additive would be something else? IMO additive is independent of how we perceive the light (=color)--BartYgor 15:33, 22 August 2007 (UTC)

Original author here. I added that section because I have heard many people make statements that indicate a misunderstanding about additive colour. Additive colour is a trick. It fools the eye and brain into thinking they are seeing yellow, when they are not. What they are seeing is a mixture of red and green lights. Yellow light would weekly trigger both the red and green sensors in the eye roughly equally, because each sensor detects a broad, overlapping band of wavelengths. The red and green lights also do that, so we register "yellow".

There is no right or wrong yellow. If people see yellow, it's yellow, nothing else. And in your definition only the colors of one pure wavelength are the true colors. Any other color is the result of combining other wavelengths (those true colors). So in you definition any color we see (be it the result of an additive -mixing light- or subtractive -filters- process would be named an additive color. That would not be correct--BartYgor 01:58, 23 August 2007 (UTC).

Yes, animals' eyes have different colour sensors, and detect light differently. So, for many animals, things like colour television sets would simply look weird - the colours would be all wrong. For an animal, you would have to use a different set of 'primary colours' - maybe 4 or more, ranging into the infrared or ultraviolet. Indeed, humans have widely differing colour eyesight, and some persons dislike colour on television because their eyes do not match the Red, Green and Blue wavelengths used.

This is IMO about how animals see light diferently than we do. It's about color but has nothing to do with additive color. --BartYgor 01:58, 23 August 2007 (UTC)

I consider the links to the other articles important here. Those articles show how the eye detects colour. I would consider a lengthy discussion on how the eye works to be out of place in this article. Robbak 00:31, 23 August 2007 (UTC)

I agree, because IMO additive color production has no more to do with how we see than any other color article.

1. First if we can't see the difference there is no such thing as additive or subtractive color.

My meaning is that the eye cannot detect the difference between monochromatic yellow light and a mixture of red and green lights. This sort of yellow light is produced by yellow LEDs and lasers. Robbak 00:55, 23 August 2007 (UTC)

Yes so if you can't tell by the color if it's additive of subtractive (you can't even tell from the wavelength spectrum) then there is no such thing. If all sticks are equally long there are no short or long sticks (I mean ther's no reason to define such). You can talk about additive color mixing or additive color production as to say producing a color by adding lightsources.--BartYgor 01:58, 23 August 2007 (UTC)

They describe different types of media: one with colored light sources (or transparencies, etc.), and another with light reflecting on a paper. There is quite certainly a difference, which is quite substantial. But you're right that in either process at the end the eye perceives a spectral distribution, and has no way of independently discerning whether that light was reflected or not. --jacobolus (t) 20:26, 21 August 2007 (UTC)

because of all the confusion I would make this clear in the article.--BartYgor 15:33, 22 August 2007 (UTC)

2. You can get green from an additive process by simply mixing yellow and cyan.

Can you? If you get a yellow light and a cyan light and shine them on a wall, what you would get would be more like white, with possibly a green cast. We are not talking about mixing paints here, or pigments on paper. That is subtractive colour. Robbak 00:55, 23 August 2007 (UTC)

It would indeed not be a fully saturated green. But more than a green cast. (See the CIExy chromaticity diagram). Mixing cyan and yellow subtractively would also not get you fully saturated green (yet more saturated than additive that's true). I mean ther's no real difference here while the article implies there is. It adds to the confusion IMO.--BartYgor 01:58, 23 August 2007 (UTC).

Yes, this is true. --jacobolus (t) 20:26, 21 August 2007 (UTC)

Then why not change it in the article?--BartYgor 15:33, 22 August 2007 (UTC)

3. Additive has little to do with how we see color except the fact that we see color iso wavelengths and therefore lack some information. Even the basic definition of additive is not mentioned: additive means that overal luminance risen with mixing (you add light, energy, luminance); subtractive means you subtract (light, energy,...).

Additive means that you are adding colours of light. Subtractive means that you are absorbing them. I would consider the increasing luminance to be a side effect of adding that extra light. Additive colour works because of the way we see. Additive colour only works because the human eye has 3 detectors which we can stimulate with three different colours of light. Robbak 00:55, 23 August 2007 (UTC).

Additive color would also work if we had 2 or 4 receptors. You probably need 2 or 4 primaries, but that doesn't change the concept of additive color. Therefore the increase in luminance is the prime property of additive color mixing not the fact that it has 3 primaries (as I explain below - that is the result).--BartYgor 01:58, 23 August 2007 (UTC)

I'm not sure what you mean by “little to do with” it. --jacobolus (t) 20:26, 21 August 2007 (UTC)

As it is stated: "color except the fact that we see color iso wavelengths and therefore lack some information". THe fact that we can't see additive color is the only relation it has with our eyes.--BartYgor 15:33, 22 August 2007 (UTC)

4. Primary colors in no way are abitrary in a physics way. For us humans we can call any color primary. But when we mix light or paint; it's no longer just in our mind because physics will influence the result. So if we define additive primary colors (it would be better to speak about the color of the additive primary lights) as the colors of the lights which can't be the result of mixing two other colors of light then it is easy to see that light of pure wavelength 700nm (red) can't be the mix of a lower and a higher wavelength as a higher wavelength is in the infrared region so we can't see it. The same holds for pure wavelength 380 (blue). Green is not that easy: therefore you would need to look at the CIE chromaticity diagram (http://en.wikipedia.org/wiki/CIE_1931_color_space). It has the property that the you can predict the chromaticity (meaning the hue and saturation; but not the brightness - it may be somewhat strange but in color theory we name color as the combination of hue, saturation and brightness while in everyday speaking with color one talks about the chromaticity and adds dark or light to indicate the brightness) of the result of mixing two wavelenghts as the (weighted) average on the figure. For example if you would mix yellow (570nm) with an equal amount (same luminance) of cyan (490nm) you will find the chromaticity of the mix as the centre of the line connecting those two points on the diagram: in this case some unsaturated green. Now if you look at the figure you can see that also a green of about 520nm can't be mixed by two other wavelength because the figure makes a sharp turn there. (You can also see why mixing green of 550 and red of 600nm would give you yellow or in other words the mix 550+660 would appear to you as having the same hue as light of 570nm). If you would define additive primary colors as colors where you can make the most other colors of in an additive process the same primaries would be the result. As connecting the lines on the chromaticity diagram would give you the largest possible triangle or gamut. But there are no additive primaries that can give you all colors.

I think you have the point that I was trying to make. Additive colour has nothing to do with the physics of light. It has everything to do with the human eye. If you mix red and green lights they do not magically change into yellow light. It is still a mix of red and green, as passing it through a prism would show you.

As I said there's no such thing as true and false yellow. --BartYgor 01:58, 23 August 2007 (UTC)

You also seem to misunderstand subtractive colour as well. We see 'Cyan' because a pigment absorbs the red, leaving the Green and Blue. We see 'Yellow' because pigments are absorbing the blue, leaving that mixture of Red and green. Mix a 'Cyan' pigment with a 'Yellow' pigment, and that mixture will absorb both the blue and the red: You will be left with green.

Did I say something else?--BartYgor 01:58, 23 August 2007 (UTC)

Going to your descripotion, if you mixed yellow and cyan, you would be triggering all three of the sensors in the eye: Yellow would flag the red and green, and cyan would trigger the Green and Blue. Yes, you would see it as greenish, but the overall impression (given that I have not performed the experiment) should be of white. Robbak 00:55, 23 August 2007 (UTC)

If I speak of green I mean the hue. Even if it is very unsaturated (grey) it is still green. In your definition when would something scientifically stop being green? --BartYgor 01:58, 23 August 2007 (UTC)

“Primary” colors are indeed quite arbitrary, “physics” having little to do with it. You can't actually get any spectral color by mixing any two other spectral colors, but if you pick any three colors as the corners of a triangular gamut that surrounds your white point, then you can make some color of every hue. So for example, you can't make the most colorful possible cyan by mixing green and blue. But you can make some cyan color. Red, green, and blue (additive) primaries just happen to form a rather large gamut, compared to other possible choices of primaries. --jacobolus (t) 20:26, 21 August 2007 (UTC)

IMO you're wrong their: the color matching functions on http://en.wikipedia.org/wiki/CIE_1931_color_space explain just that. IN the CIE experiment they tried to mimic the spectral color by RGB. For the region of 420nm to 550nm they had to use a trick ('negativ' red light). That means you're partly right that from 380nm to 550nm all spectral colors could be the primaries according to my first definition (you can't get them by mixing other colors). I picked green of 520nm because it's is the most difficult get approximate. the color of light of about 490nm can be more easily approximated (yet you're right you can't get it 100 %right). I would therefore call primaries: the colors of the lights which can't be the result of mixing two other colors of light and with which you can make the most other colors of in an additive process (combining bothe definitions).--BartYgor 15:33, 22 August 2007 (UTC)

That's a terrible definition. The primary colors chosen in an additive process depend on many trade offs, including cost, viewing conditions, available technology, etc., and are nearly never pure spectral wavelengths. --jacobolus (t) 02:52, 23 August 2007 (UTC)

Of course those wavelengths are ideal. In PC-monitors you would compare the trade-off better gamut - higher cost, and maybe other technical issues. But that would depend on the aim (if the aim was a to do an CIE experiment you would try to get the clossest match or a match without trade-off as they did in the experiment), but still it explains why monitors work with RGB and why the primaries in additive processes will be somewhere in the red, green, blue region. It explains that the choice is not arbitrary (even if there were no technical issues or other trade-offs). It explains that if you were only to chose three colors (lights) which would you choose (or try to approximate). My definition of primaries holds there are numerous references out there (gamut + can't be mixed from others). But even just gamut as definition gets you the same answers. What you are saying is you can't name the optimum of a process the optimum because it can't be reached (or it's too expensive) --BartYgor 11:31, 24 August 2007 (UTC)

Reading this part over: I think we define primaries basically in the same way: largest gamut. The spectral colors RGB are the best fit to that description. If you don't want to add to the definition that "primaries can't be mixed from other colors": I can live with that, because it would no longer hold in subtractive primaries. So the mix definition is outdated, yet very much used, so something should be said about it in the article. And it should be said that the choice is not arbitrary: you can use other basic colors yet with an obvious trade-off. It think the article should answer the question why RGB are so common. And if you insist on the choice being arbitrary than you can't explain while there is one or at least make things confusing. --BartYgor 11:53, 24 August 2007 (UTC)

No, we define them differently. The “ideal” primaries that you seem so interested in don't exist as far as I'm concerned. There are perfectly valid reasons for picking other than RGB “primaries” even if cost and technology weren't a concern. Primaries only can't be mixed from other colors within the same gamut… many of them could be mixed from out-of-gamut, even spectral lights. But if you define your primaries to just be three spectral colors of some “ideal” wavelength, that's an equally bullshit definition. No spectral color is magically better than another, and none can be mixed from some combination of two others. And “largest gamut” is also not particularly well defined: are we talking about in a roughly perceptually-uniform space like CIECAM? Or just on an xy chromaticity diagram? The article should *definitely* explain why RGB are commonly used. But it shouldn't ascribe to them some mystical significance. --jacobolus (t) 00:26, 25 August 2007 (UTC)

Your definition is then: "primaries are just the colors that you start of with to mix";. If that is your definition than any color will of course be a primary as you can freely pick the colors you will start mixing with. Talk about BS though. And why is stated in RYB article "It predates modern scientific color theory", if primaries are so 100% arbitrary: it's completely within modern scientific color theory. "none can be mixed from some combination of two others" is BS too: look at the CIE experiment -> color matching functions: or is that expriment BS too? And sentences like "And “largest gamut” is also not particularly well defined: are we talking about in a roughly perceptually-uniform space like CIECAM" just proove you don't understand what you're talking about. Gamut is the set of colors that can be reproduced, so a largest gamut will be the largest in any colorspace that encompasses all visible colors, perceptually-uniform or not. Using the xy chromaticity diagram could in theory lead you to wrong conclusions (as it is only a section of the CIE xyY space), yet it gives a good first idea and gives in almost every practicle case the right conclusion. Ther is absolutely no reason te define the colorspace when your saying "largest gamut". "No spectral color is magically better than another", in my definition of primary (largest gamut) there is. In yours there isn't.--BartYgor 13:17, 25 August 2007 (UTC).

No, you misunderstand. a) the RYB model *does* predate modern color theory, and the idea that the colors red, yellow, and blue are uniquely special *is* inconsistent with current understanding of how color works. I'm not sure what your point is there. b) “Largest” depends on color space, because in different spaces, the relative volumes of different gamuts will change. So just picking “largest” based on an xy chromaticity chart is somewhat misleading. Overall, BartYgor, you have generated tens of thousands of words of discussion on several articles, and very little benefit for myself or wikipedia has come of it. I'm going to for the most part stop discussing these subjects with you. I mean no disrespect, but it's not worth the time. I do wish you the best of luck in continuing to learn about color science and applications, and please do keep working to improve Wikipedia's color-related articles. --jacobolus (t) 19:03, 25 August 2007 (UTC)

The definition of 'volume'(see that is what most authors seem to do I've met here. THey sway around with terms and don't even bother to define them in the first place so a genuine discussion can take place - looking true wiki talk pages I saw numerous complaints of people that worked hard to get things right and than come back after some time (years) and see somebody else made a mess of things. You may wish me good luck but I have learned one thing: don't trust wiki - trust your own mind) in colorspace can't be anything else than number of colors. This number can not change in whichever colorspace you project the gamut (as long the colorspace is larger than the gamut). The shape might change, but not the volume, not the number of colors as this is the definition of gamut.If i'm wrong: give me an example. Can't be that hard.--BartYgor 20:30, 25 August 2007 (UTC)

And another thing wiki's seem to do is never admit they were wrong (be a man about things!) and you just divert the discussion on other subject. For example I started this point by saying: "Primary colors in no way are abitrary in a physics way." This holds. I say: additive has no more to do with the way we see than color itself. I'm right again. Even the two people who asked some pertinent questions in the beginning feel that something is illogic about everything. But yeah just go on ignoring that....--BartYgor 20:50, 25 August 2007 (UTC)

"The Inter-Society Color Council (ISCC) is the principal professional society in the field of color, encompassing the arts, sciences, and industry. They agree that cyan, magenta and yellow are the primary colors of pigment, which in

combination can form all hues. The most recent draft of the revised California science framework also states this." And so you keep insisting primary colors are arbitrary and blabbing about references. You hold tight on your definition. Best of luck with it. --BartYgor 21:30, 25 August 2007 (UTC)

I might be wrong every now and then; hey I'm glad: I learn. But you dan't even seem to think for yourself. From my definition of primary I deduct complete logic conclusions. So don't talk about BS. Fight me on my definition! And about the definition of primaries: just type that in google and see how rare or even nonexisting you definition is. I don't say it's invalid (you can name anything the way you want) but not on wiki.--BartYgor 15:16, 25 August 2007 (UTC)

THis is more for the primary colors Talk pages IMO -BartYgor 11:53, 24 August 2007 (UTC)

5. Why CMY are the primary subtractive colors read my comments on http://en.wikipedia.org/wiki/Talk:CMYK_color_model#Why_we_use_CMYK_in_printing --BartYgor 15:16, 21 August 2007 (UTC)

That's not really relevant to this article. --jacobolus (t) 20:26, 21 August 2007 (UTC)

Answering the question of Grr.--BartYgor 15:33, 22 August 2007 (UTC)

what? --jacobolus (t) 02:52, 23 August 2007 (UTC)

Insert non-formatted text here

the question that started the discussion.--BartYgor 11:31, 24 August 2007 (UTC)

Stub?

This article is surely above stub class. IMHO, this article contains as much information as it needs: personally, I would consider it complete. Hence, I have rated it B. Feel free to comment, or change it back even, if you disagree. Robbak 02:27, 23 August 2007 (UTC)

No, compare this article to B class articles in this wikiproject, and you will see that this one is far less developed. The article first needs some references, and a few more filled-out sub-sections. At that point we can perhaps consider it "start" class. It's nowhere close to B class, however. --jacobolus (t) 02:54, 23 August 2007 (UTC)

Merging

The topic itself seems to be color composition, or color models. Combining additive and subtractive color composition would allow for an overview and comparison, increasing understanding of the subject matter. (Which is an issue of confusion, as witnessed by the section Additive vs. subtractive primaries, on this page.)

White!?

The addition of red, green and blue light does not yeild white. It may be true to say that the three produce "white light" but "white light" is not white it is in fact transparent. The only time that red, green and blue light produces a white colour is when they are shone on a white background. — Preceding unsigned comment added by 129.11.77.198 (talk • contribs)

A more common example is the emitted light from a CRT, or the transmitted light from an LCD display; the small red, green, and blue areas merge visually to appear white. Dicklyon (talk) 22:06, 28 January 2009 (UTC)

If white is "Every color mixed" then howcome "there's a difference between red and green light mixed and actual yellow light" ?