17

We know that CMYK is subtractive color model and RGB is additive color model.If we make 3D( R: x-axis, G: y-axis, B: z-axis) cube for RGB it starts with darkness (black) and we gradually add light; an CMY(K) image requires to be illuminated by white light.

When a printer performs printing operation, how do CMYK and RGB work together to print on white paper? And why does CMYK technology need both, white paper and white color simultaneously?

4

6 Answers 6

26

Both photographic film and photographic digital imaging chips are sensitive to the three light primary colors. In other words, both record components of the image i.e. red, green and blue.

Photographic color film displays images using dyes that are the complement of the three light primaries. In other words, the finished developed image using cyan (complement of red), magenta (complement of green), and yellow (complement of blue). Stated differently, the color image displayed on both slide film and negative film is comprised of cyan, magenta and yellow dye. Complement means opposite.

When color images are displayed on a computer or TV screen or projected on a screen, the color image is fractured into its three light primary colors. The image we see is comprised on glowing dots called pixels (picture elements) of red, green and blue. The image is comprised by controlling how much red, green, and blue reaches our eyes by adjusting the intensity of each glowing dot of light.

When it comes to displaying color pictures on paper (prints), the image is fractured into tiny dots of cyan, magenta, and yellow dye / pigment. Again the scheme is to adjust the amount of red, green and blue light that reaches our eyes using dyes that act as filters. Using red, green, and blue tiny dots for prints yields substandard results. This is because the laydown of the colored dots on paper are not individual -- they overlap.

Allow me to explain – The color print is generally viewed via white light illumination from a lamp that is external from the print. Light from this lamp must traverse the dye / pigment which is transparent. The light then hits the white reflective paper and is reflected back towards the viewer. This light again traverses the dye / pigment. In other words, the viewing light makes two transits through the dye / pigment on its way to your eyes.

Now transparent dyes / pigments act as light filters, A light filter passes its name and absorbs the other colors. A cyan filter is a red blocker of light, passing green, and blue. The magenta filter is a green blocker of light passing red, and blue. A yellow filter is blue blocker, passing red, and green. In other words, we view color pictures on paper via the fact that colored dye / pigment stops some colors while passing others. The CMY method for prints on paper works because the dye / pigment stops one color and passes two colors. The results are vivid color pictures on paper.

When dye / pigment overlap, what happens? Magenta + yellow overlapped yields red. Magenta + cyan yields blue. Yellow + cyan yields green. This scheme works for reflective copy like prints on paper. The TV and computer screen uses red, green and blue due to the fact that the individual pixels are not overlapped.

We can make wonderful yellow dye / pigment, fair magenta dye / pigment but poor cyan dye / pigment. The deficiency in our ability to make cyan reduces the contrast of the print on paper. It is necessary to add a black dye dot to prints on paper to bolster contrast. This black dot colorant is called a “Kicker” (kicks up the contrast). Thus, we use the CMYK scheme for prints on paper. We can’t use red, green, blue dye/pigment because if any of the light primary colors are overlayed, the result is black (total absorption) whereas if two of the subtractive primaries CMY are overlayed the result is RGB.

The TV and computer screen works using three sub-pixels red, green, blue, the triad makes one pixel, and they are individual i.e., not overlayed.

Nobody said this stuff is easy!

P.S. Suppose white light is playing on paper coated with a transparent red dye atop a green transparent dye. The top red dye passes red and blocks green and blue. Now red light plays on the green under dye. This light is void of green and blue. The green dye can pass green light, but none is present, it has been absorbed by the red dye above. The result is, no light passes the second dye coat. Together the two dyes have absorbed all thee of the light primary colors. Black is the result if any two primary colors are overlayed. If the dye is a subtractive primary, these pass two colors and stop only one color. A cyan dye stops only red. A magenta dye stops only green. A yellow dye stops only blue.

Overlay Magenta with Yellow. The magenta dye stops green, the yellow dye stops blue. The red light is not stopped, we see red.

In color printing we what to control the intensities of the red, green and blue light that hit the paper and are then reflected to our eye. Cyan dye controls the amount of red we see. Magenta controls the amount of green we see. Yellow controls the amount of blue we see.

When we print with dye on paper, the best way to control how much light is reflected back to our eyes is to use cyan, magenta, and yellow. The subtractive light primaries.

Cyan passes green and blue and blocks red. Magenta passes red and blue and blocks green. Yellow passes red and green and blocks blue. Color film uses CMY – Color prints use CMYK – Digital images projected or on a TV or computer monitor use RGB. Prints on paper made from digital images use CMYK K is the Kicker because cyan dye is nearly impossible to make right.

There are exceptions but most exceptions yield substandard results. Color theory is what you need to study – why not Google the subject?

32
  • "We can’t use red, green, blue dye/pigment because if any of the light primary colors are overlayed"--- don't understand this, if suppose red and blue is overlayed then then it produce black?
    – Alok Maity
    Feb 13 at 17:26
  • 5
    A red object reflects red light thus it absorbs green and blue. A blue object reflects blue light and absorbs green and red. Overlay blue and red - blue absorbs green and red - red absorbs green and blue. Note when overlayed the two together absorb all three-light primary colors - the result is black if any two of the three light-primaries are overlayed. In TV and computer, the three are juxtaposed, not overlayed. That makes the difference. Feb 13 at 17:37
  • 5
    A perfect cyan dye stops red and passes green and blue. We can make cyan dye but to get it to stop the red it must be too dark. So we make it lighter. This imperfect cyan dye also unwaning stops some green and some blue. Maybe you can help and make a better cyan dye, the photo community would greatly benefit. Feb 13 at 18:05
  • 1
    Viewing a print on paper using light other than white adds a color bias that is usually a deterrent. Same is true if the paper is off-white. Feb 13 at 18:12
  • 1
    Contrast is the starkness of lack of - black to white scale. Cyan, magenta, and yellow overlap should yield a rich black, it dosn.t due to impurity of cyan dye. Add a kicker (black dye) adds needed zip (contrast). Feb 13 at 19:19
20

Let me debunk one thing: "RGB and CMY are two different color models."

They are not, they both form part of one RGB-CMY model. Or probably an RYGCBM model.

Ok. That is not entirely true but in essence yes. On one model 3 are primary and the other 3 secondary and vice-versa.

Look at any modern color wheel all 6 colors are present. Some models are dimmer and less saturated because they are simulating the capabilities of the final output, but they are all there.

Some color wheels do not have black or darker colors, because a color wheel is in reality the top view of a color solid. I use the word solid because they can be many shapes. Cubes, cylinders, a double cone, spheroids...

CMY(K) image requires to be illuminated by white light.

That is the full point. Where does the light come from? Do I start with the full spectrum, do I need to add more and more light, different wavelengths? or I need to remove them.

Historically it is easier to remove the light. That is simpler. Any colored material does that.

When a printer performs printing operation, how do CMYK and RGB work together to print on white paper?

They work together because they are all part of the same color solid. But it needs to have a conversion because one is based on some primary RGB and the other need to use CMY as primary colors.

Allow me to spam you with an old webpage I wrote some time ago. I am sorry that it is not in English, just take a look at the image.

A basic conversion is just taking one channel and using it as the complementary color. Take the R channel and use it to print C ink.

But why? Imagine you have on your Red channel some black zones that have no red, that is why it is dark on the red channel. That means that you should print a lot of Cyan because Cyan has no red.

Let me simulate this.

On an RGB image, I am breaking the RGB channels into a grayscale image.

We can identify each one compared to some clear color zones, Red ballon translates into a bright zone on the Red channel. The same with G and B.

enter image description here

If I use those channels as the complementary color directly I could still have a decent color reproduction. The image on the bottom left is a reference image.

enter image description here

Why is CMYK more efficient/beneficial than RGB for performing printing on paper?

I think the next examples are not what you asked, but let me explore.

Ok. Let's try not using CMY colors, but RGB. These are red, green, and blue inks, using the exact same channels as the RGB file.

Hum... we have a mess. We have what we asked for. The cyan water is black on the red because we do not have red on the water. But now we have a lot of red because of the black of the ink red channel.

enter image description here

Ok, how about using some other colors, not RGB, nor CMY... how about the colors used traditionally on the RYB color model, before the magenta dyes were invented. The blues were also less cyan-ish before the cyan pigments were invented.

enter image description here

We could live with that. We also could have brighter reds.


The first answer is, we use RGB when adding light, and CMY when subtracting light because adding is the opposite of subtracting then we use the complementary color.


And why does CMYK technology need... white paper

Try starting with black paper and transparent inks... Not good. What do you want to reflect with your print if all the light is already absorbed by the paper?

In the end, the question is, can your object emit light or only can reflect some light that comes from somewhere else.

Returning to the question

Why is CMYK more efficient/beneficial than RGB for performing printing on paper?

If you are referring to a CMYK file instead of an RGB one the answer is just that, in some controlled environments you need to define the exact amount of ink of each channel, therefore you need to save those values on the file itself, then, you need a CMYK file.

Each combination of ink-paper needs different values of ink. That is what a color profile does.

On a home or office environment, and with the humungous amount of color printers, you let the printer driver make the decision on how much ink to inject. That is why is preferable to use an RGB file. Let the driver drive the conversion.

In commercial print, you follow some standards. Gracol, SWOP, European, etc. With standardized inks on standardized papers. So you define one specific conversion and save it on a CMYK file, so the values are fixed for that particular case.

Ballon image.

20
  • 1
    "But now we have a lot of red because of the black of the ink red channel. "--- this line don't understand. Why water is deep red if I use blacky red channel?
    – Alok Maity
    Feb 14 at 9:01
  • On screen, black on one channel means no light of that color. On paper black on one channel means a lot of ink of the color you are using.
    – Rafael
    Feb 14 at 9:06
  • 1
    @AlokMaity: It might help to think of cyan, magenta, and yellow as antired, antigreen, and antiblue, respectively (which they are!). On a computer screen, the water is displayed using lots of green and blue and very little red. On a printed page, the water is displayed using very little antigreen and antiblue, and lots of antired. When you invert the colors to go from RGB to CMY (going from, say, green to antigreen), you also have to invert the amount of the color used (going from, say, lots of green to very little antigreen) in order to keep the image looking right.
    – Vikki
    Feb 15 at 8:06
  • 1
    @AlokMaity: ...OK, now I can't even tell what you're trying to say.
    – Vikki
    Feb 15 at 9:40
  • 2
    This explanation helped me understand things that I had been confused about for many years. Feb 15 at 16:06
11

There are two sort of separate issues here. One is RGB vs. CMY. The other is why you add the fourth color, instead of just using CMY (and sometimes, you use half a dozen colors or so, not just four).

CMY vs RGB

CMY vs. RGB is basically a question of how printing on paper works compared to how (for example) a monitor works.

Something like a monitor uses an additive model. You start with three tiny light sources next to each other, each emitting some amount of a component color. These are then basically added together to produce a result of a single apparent color.

When you're looking at something on paper, however, you start with the ambient light falling on the paper. Thanks to how our eyes adjust to ambient light, we usually see this is pretty much "white" regardless of its color1.

The ink we deposit on the paper then subtracts the colors we don't want. To see something as a particular color, we filter out the appropriate amounts to get from white to that color. To get Red, we have to filter out Green and Blue. To get Blue, we filter out Green and Red. And to get red, we have to filter out green and blue.

Doing that with red, green and blue ink would create a problem though. Since each of those primary colors filters out two other colors, mixing any two of them tries to filter out all the colors, so we end up with something that wants to be black (but probably mostly fails).

So for subtractive color, we start with primaries that only filter out one color. So Cyan is an absence of red, magenta is an absence of green, and yellow is an absence of blue. Since each dye only removes one primary color, we can mix them freely to get the colors we want.

So, with RGB, we basically start from black, and add an appropriate amount of each primary color to get what we want. With CMY, we start with white, and subtract an appropriate amount of each primary to get what we want.

CMY vs. CMYK

CMY vs. CMYK is much less about anything like a theoretical model, and much more about simple practicality and economics.

First of all, at least traditionally, black ink has been much less expensive than colored ink. For example, India Ink is just soot suspended in water.

Second, blank ink tends to be quite "dense"; you don't need a lot of it to get a really black "black".

Colored pigments have often been more difficult to find, so they were usually more expensive. Preparation to turn a raw pigment into usable ink was often more difficult as well.

Mixing colored inks usually doesn't produce a particularly great result either. With most inks, the result tends to be a kind of dull, muddy grey instead of a nice clear black.

CMY uses a lot of expensive ink to produce kind of mediocre results (especially for darker colors). CMYK generally produces better results at lower cost. That usually makes CMYK a pretty obvious, straightforward choice.

Inkjets

Most inkjet printers, however, tell the rest of your computer system that they're RGB devices. It's ultimately ink going on paper, so the final result is subtractive, but all the conversion from RGB to amounts of the inks they use is handled internally by the printer or its device driver.

Many also use more colors. At least usually, they don't use more primaries though--instead they use lighter versions of the same primaries. I believe this is largely to make up for the limited control they have over droplet size. For example, instead of a 500:1 ratio of saturated colors, they can use a 50:1 ratio of a saturated and unsaturated color, so the measurement of the smaller amount isn't quite so critical.


1. But if the light is different enough from white, we can have difficulty adjusting, such as under mercury vapor lamps.
4
  • 1
    For footnote 1, do you mean mercury lamps or sodium lamps?
    – supercat
    Feb 15 at 0:09
  • @supercat: I meant mercury vapor, though sodium vapor cause some of the same problem as well (though mercury vapor is mostly phased out, since mercury vapor is kind of problematic from a health viewpoint). Feb 15 at 1:45
  • Mercury lamps give off primarily ultraviolet and then use phosphors to convert that to visible light, in whatever combination of wavelengths the particular phosphor vendor chose. Sodium light is nearly monochromatic yellow.
    – supercat
    Feb 15 at 15:39
  • @supercat: That depends on the pressure. What you're saying is accurate with respect to low pressure mercury vapor lamps. But high pressure mercury vapor lamps produce light at roughly 405 and 436 nm, which is in the visible range (without any phosphors). Feb 15 at 15:56
5

While RGB and CMY may be similar, adding K changes things. In theory, CMYK values of (0.3,0.3,0.3,0) and (0,0,0,0.3) should appear as identical grays, but in practice there may be differences which may favor one over the other. For example, if one has a field of (0.3,0,0.3) green with some gray text, using the first gray above would yield a uniform field on the cyan and yellow plates, with the text appearing solely on the magenta plate. By contrast, if the gray text were printed using (0,0,0,0.3), then there would be text-shaped holes in the cyan and yellow images, making the image much more sensitive to registration inaccuracies.

On the flip side, if the field had been cyan rather than green, using (0.3,0.3,0.3,0) for the gray would mean that the imperfect registration could cause some false-color fringes in the text which would have been avoided by using (0,0,0,0.3) text.

For most RGB colors, there will often be multiple ways of rendering it as CMYK, and the choice of which is best may depend upon context, artistic judgment, and expectations regarding plate registration accuracy. RGB has no way of recording such judgments.

1
  • Image1, Image2 this two image not understanding, how they are complementing? Could you help me to understand.
    – Alok Maity
    Feb 15 at 12:24
4

The answer is in your question. It is because

TLDR

  • RGB is an additive color model
  • CMYK is a subtractive color model

You are using the terms additive color model and subtractive color model without understanding what they mean.

Physics

Let me ask you a question. Water behaves like an incompressible fluid model and air behaves like a compressible fluid model. Why can we squeeze a container full of air and reduce its size but we cannot reduce the size of a container full of water but only change the shape? Can't we just make the water behave like a compressible fluid model and compress it?

The answer is the real world does not behave that way. Physics only explain the real world using models but the real world behaves the way the real world does.

The same goes for additive and subtractive color model.

RGB

When using light to show color the colored lights behaves like an additive color model. When we add all the colors of light we get white light. That's because color is mixed in our eyes.

Our eyes evolved to sense three colors - red, green and blue and our eyes cannot really see all the frequencies of light (eg. sunlight or white light). Due to this limitation when our brain sees an equal amount of red, green and blue photons it assumes we are being shown white light (or sunlight) and based on this assumption interprets what we see as "white". This is how we trick our eyes to see white on computer screens when there is no real white light being displayed.

CMYK

When inks get painted/drawn/printed on surfaces it behaves like a subtractive color model. This is because inks are pigments and dried chemicals.

The way things are the color they are is because they absorb some light. If something does not absorb any light and reflect all light then we will see it as silver (like a mirror), if the surface is rough enough then that sliver color will become white (like paper) because the reflected light is scattered. Actually, silver and white are the same "color" - that is, both are all the colors of light. The difference is how much the surface scatters the reflected light.

So a banana absorbs blue light. Because bananas absorbs blue light the remaining light reflected off it is yellow. Red ink on the other hand absorbs green light. If you paint a banana with red ink the surface will now absorb green and blue light. The remaining light that reflects off the banana will appear orange.

You can appreciate that if we mix the different chemicals that absorb all the different colors what will be left is no light to reflect. The absence of light is black. So in theory mixing all the paints will give you black instead of the white that you will get when mixing all the lights.

In practice you don't get black. This is because chemicals aren't pure (or pure chemicals aren't pure light absorbers like black holes). There is usually an underlying yellowish dirty background color left over when you mix all pigments. This very dark yellowish reflected light is brown. When mixing inks you end up with brown instead of the theoretical black. Which is why artists and printers use an extra color - black (traditionally called the "key" color) giving you CMYK.

In short it's just how reality works. Inks don't work like lights and lights don't work like inks. The different color models are just physics trying to make sense of it all.

6
  • Image1, Image2 this two image not understanding, how they are complementing? Could you help me to understand.
    – Alok Maity
    Feb 15 at 12:24
  • "how are they complementing?" Very basically, C is the inverse of R. If color values are expressed as a range from 0-1, then if R is .80, then C is .20; same M:G; same with Y:B. When applying a Red filter to an image and taking a black and white photo, the negative obtained is the C channel.
    – Yorik
    Feb 15 at 16:35
  • @Yorik you mean in image1 after using red filter then we take complement of it then we get first image of top-left?
    – Alok Maity
    Feb 15 at 18:01
  • I mean exactly what I stated: take a color image (in the real world), place a red filter over it, take a b&W photo; the negative of that image is used to make a Cyan plate for printing. If you take a copy of the "C" channel (which is a greyscale image, as in your image example), invert it so that light areas are dark and dark areas are light, the result is the R channel. It is complicated by the K channel a little (it affects the math), but if you invert each of the CMY channels in this way and place the results in the respective RGB channels, you will have a decent, washed-out regular image
    – Yorik
    Feb 15 at 19:59
  • this is what is meant as complimentary: for any given brightness of R in a range from 0 to 1, then C is 1-R. For real math, see the comment below the question by "Jason C"
    – Yorik
    Feb 15 at 20:01
1

I've worked in the printing industry, making software that measures the printing efficacy and gives statistics and even directions on how to adjust the printer.

Ink in real life is not mathematically pure. It has thickness, it is imperfect.

Layering more ink to make it darker will use up a lot of ink; the effect is imperfect due to the overlapping dot pattern rather than ink at a single pixel, and the fact that making the ink film thicker also blocks other light than what it's meant to because it is not a perfect filter of the given channel (in particular, Yellow is the dirtiest). Multiple thick applications of ink will be messy and that comes with its own handling issues.

Starting with "Key" (a nice black pigment) and then subtracting that darkness from the other channels gives a much better result.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.