If so, what is its name?
You may call it… Tim?
(sorry!)
"…Why does Adobe use a color wheel that has yellow and blue as complementary colors…"
Because the complement of a color is the color which must be added to it to produce white. Yellow (which is really red and green) when added to blue yields white. And cyan (blue and green) is the complement of red because cyan + red yields white. And magenta (blue and red) is the complement of green because magenta + green yields white.
"…I know that RGB is the subtractive color wheel and CMY is the additive one…"
John, you have that backwards. RGB is the additive system, CMY the subtractive one.
I’ll show my ignorance (and perhaps learn something). Why is the RGB used for video screens and CMY for printers?
RGB are the additive primaries.
The additive color system deals with light (what monitors display). Adding R+G+B adds up to white by increasing the wavelengths of light perceived, as you add more colors of light.
CMY are the subtractive primaries.
The subtractive color system deals with pigments (what printers apply to paper). Adding C+M+Y subtracts down to black by decreasing the wavelengths of light perceived, as you add more pigments to a paint/ink/whatever.
Thanks John,
That is what I thought but I needed conformation. As an author, I am writing an article about using a computer and PhotoShop to aid in the identification of postal stamp color shades. Not a big deal to most but I like to be accurate.
Again, many thanks. I was going to say the RGB was used if it was projected at you (with light) and CMY was used if you were looking at it with other lighting.
Re: I was going to say the RGB was used if it was projected at you (with light) and CMY was used if you were looking at it with other lighting.
This is valid. RGB is projected light, CMY is reflected light. Hence they are opposites.
I was going to say the RGB was used if it was projected at you (with light) and CMY was used if you were looking at it with other lighting.
Emitted vs. reflected.
The subtraction can be explained like this:
A cyan dot on the paper is a thin filter (thickness about 1 micrometer). White light passes the filter and the red spectral part of the light is lost (subtracted). The remaining light is reflected by the surface of the paper and passes the filter again (or besides). This doesn´t change the color further.
Similar for magenta (absorbs green) and yellow (absorbs blue) .
The total amount of reflected light adds – here we have also an additive process because it´s now emissive light.
A sketch of these filters shows the surprising result that the three colors can be printed overlapping, half overlapping or separated – the hue is theoretically the same. At least this allows some uncertainties in registration.
Best regards –Gernot Hoffmann
Listen to Gernot, fellas.
He’s our resident Dr. Color.
Go, Gernot, Go!!!
Have you ever seem the RYB color wheel that is taught in high schools? Those are the primaries, and OGP (P is for purple) are the secondaries.
That RG and OB are complementaries can be proven by a simple test. Take a blue sheet and an orange sheet. Cut the orange sheet in half, discard one half and use the other half to cover the center of the blue sheet, with the long edges pointing in the same direction. Hold this assemblage against a white wall and stare fixedly at it for a long minute. Then yank it away and after-images that reverse the colors should appear. Our teacher told us that this was the test for primaries.
JOhn,
WHY are you posting this again?
It was fully discussed in your last posting on this topic:
john schutkeker "Photoshop Has a Serious Flaw in it’s Color Calculations" 11/11/03 6:59am </cgi-bin/webx?13>
If you failed to understand the answers you then received you need to make your question much more specific to those answers. It would also be far better to keep all your posts about a single subject in the original posting and not create another new topic
Right, but if you read the previous replies, you’ll see that RGB and CMY are the two primary systems, both additive and subtractive. How does RYB fit into this triad and why are there three distinct color wheels?
John,
The color "wheel" is aptly named. The question you ask is an eternal one—it goes around and around and leaves your head spinning. I have yet to see a satisfactory answer.
Joining the disparate ends of the linear color spectrum to form a circle gives rise to a gross discontinuity in wavelength at the interface where the ends artificially butt. Furthermore, in an RGB wheel, there is no provision for violet which, spectrally, lies beyond blue.
George
I guess that must be the answer. My original theory was that violet, or indigo as I knew it, was a fiction, since indigo is really just really dark blue. If you throw that away, you can wrap the spectrum neatly into the RYB wheel, which makes purple the complement of yellow.
But this disagrees with why RGB and CMY work. So I guess what I’m saying is that Photoshop needs to make the RYB color wheel the working color wheel, since that’s the one that’s intuitive to people. Then they would want to add the mathematics to transparently convert from RYB to RGB and CMY.
Then the orange range wouldn’t be so annoyingly compressed, and the blue green range so annoyingly expanded. I guess that they’ve left RGB as the reference, because that’s the model used by the monitor, which is what people look at to do their work.
Or if they can’t make RYB the working model, because RGB is mapped to the screen, they need to invent a hack to expand the orange range and compress the blue-green. That distorted color wheel really annoys me, since orange is one of my favorite colors.
RYB doesn’t exist as a color wheel – it’s just a quick way to explain things to first graders. (generally getting close to Magenta, Yellow, Cyan).
Violet is present in an RGB or CMYK color wheel.
What you miss from joining the ends of the visible spectrum is magenta.
John,
In RGB, a mix of blue and red is magenta and magenta shows up midway between blue and red on the RGB wheel.
The RYB scheme extends the spectrum beyond blue to include violet. Violet appears between blue and red on the RYB wheel, and there is no reference to magenta at all.
Magenta and violet are not different names for the same color. They are different hues. It seems that the RYB wheel includes all of the visible spectrum whereas the RGB/CMY wheel truncates the spectrum, omitting the violet range.
Still confused, George
Chris,
"…RYB doesn’t exist as a color wheel …"
Then what is this grapefruit-sized wheel I have in my hot little left hand as I type with my right index finger? Am I hallucinating? It bears the logo of Daler-Rowney (art supplies).
Sure enough, the Daler-Rowney wheel shows red, blue and yellow as primaries, which it says "cannot be mixed from any other colors. It shows orange, green, and violet as secondaries which are obtained from "two primaries mixed together". And it also identifies six tertiaries, which derive from "one primary and one secondary mixed together". It does not ever refer to magenta or cyan.
So, the RYB wheel is out there. Is it, perchance, applicable to artists mixing oil-based paints and not at all applicable to the pigments and dyes used in ink-jet printers and by printing services, nor to the phosphors generating CRT colors?
George
George – it is a very skewed representation of what should be either an RGB or CMY color wheel.
No, it is not applicable to anyone — you cannot get good results with such a wheel because it is not based on visual, light (RGB), or pigment (CMY) mixing.
But I don’t know why such a reputable paint supplier is promoting such bad color practices.
So what about the after-image test? Are you saying that works for the primaries on the other wheels? Does it work for any two colors that aren’t too close together?
George/Chris maybe because of the pigments they use in tempera paintings, it doesn’t relate to those used in press printing…
If the RGB/CMY color wheel truncates the spectrum short of violet, it might well be called the "shrinking violet".
So what about the after-image test? Are you saying that works for the primaries on the other wheels? Does it work for any two colors that aren’t too close together?
Try it yourself. Make an American flag with black stars on a yellow field and cyan and black stripes. Then make an American flag with black stars on a orange field and green and black stripes.
Try the test with both. Which one produces an afterimage that is Red White and Blue?
I could have sworn our teacher’s after-image test was blue/orange.
So I guess that’s it, then. RGB and CMY are the color wheels, and RYB is make believe for kids. Too bad, because I thought Red, Orange, Yellow, Green, Blue, Purple was much more intuitive and effective. I find the way RGB and CMY compress the orange range and expand the blue-green range to be very annoying.
John
"…RGB and CMY are the color wheels…"
You use wheelS (plural). There are not two wheels (one for RGB, one for CMY). One wheel serves both.
It is the human eye that dictates the "primary" colors. The eye’s three color receptors (cones) have peak responses at three distinct wavelengths–R,G, and B. Excite all three receptors equally and the brain "sees" white—a composite of R,G,and B. Filter out any ONE of these components from white light and you get Red/Green (called yellow), Blue/Green (called cyan), or Red/Blue (called magenta). There is nothing physically "prime" about CMY or any triad other than RGB in any scheme, and it is the eye that so distinguishes RGB rather than any inherent property of the radiation itself.
George
True colour is spectral: a wave length can be assigned to each colour. But the way colour is perceived by humans is different: Cones perceive red , green, and blue (roughly, since the red, green, and blue sensors are really broad curves with overlap). The messages are sent behind the retina to nerve cells that separate colour information into Red-Green and Yellow-Blue opposites, and that signal is sent to the cortex that perceives the colour in part in relationship to other adjacent colours (grey looks slightly bluish if adjacent to a red background) and to the setting of occurance (shapes, etc. Like Oranges are supposed to be orange, etc.). Artists try to make some sense of all this with complimentary and opposite colours. I think this is much more complicated than what might appear in Photoshop: Like artists making shadow areas have bits of opposite colours rather than just mixing in black. Maybe that is what the High School teacher was talking about. Or maybe they were just ignorant.
Typing "Blue and Orange opposite colors" gives over 75,000 matches on google, so it has to be true. 😉
My complaint about Photoshop’s color wheel is that it compresses the orange range and expands the blue & green ranges. This causes a loss of information, because some of the bits used to represent the color are wasted on providing more B/G choices than the user wants. This waste is reflected in the O range, where it’s very hard to find the shade you want.
AFAIK, if you look on the spectrum, this compression doesn’t happen. Plus Photoshop’s magenta that is supposedly an even mix of R/B seems both too light and too blue.
Maybe I’ll be laughed at by the community, but it seems to me that the children’s version of the color wheel, as taught in the high schools, is frequently better than the adult, RGB/CMY wheel. It is more intuitive, and it is neatly divided into six ranges, each of which is an even sixty degrees in extent. RYB are the primaries, and OGP (P=purple) are the secondaries.
What secondary corresponds to orange on the RGB/CMY color wheels? You certainly can’t find orange by typing in round numbers of R,G and B values.
There’s been some talk in other postings about having a Crayola pallette, so maybe someone will appreciate my idea. I think Photoshop should add the option of selecting colors from an RYB color wheel.
Well, you don’t want any Blue at all if looking for a saturated orange. Orange is made from Red and Yellow, right? In RGB Yellow is made from Red and Green, so you need Red + (Red+Green), so 128 Red + (128 Red+128 Green) or 255 Red + 128 Green
John,
I think you may be confusing two separate issues regarding colors and mixing. The RYB model you refer to is taught in reference to mixing colors when painting (e.g., red + yellow = orange, yellow + blue = green, blue + red = purple/magenta). In painting (for example with tempera or oil paints), when you mix complimentary colors, you basically get a brown "mud". Red and green pigments when mixed together will never give you yellow when painting. Likewise, blue and orange. This has to do with the opacity of the pigments and their lack of ‘purity’ (e.g., what is true blue? cobalt blue? ultramarine blue? cerulean blue?) The answer is none of them. The same goes for reds and yellows.
So, the RYB color wheel you have from school is not wrong and neither are the RGB and CMYK color wheels. Yours is for painting. The others have to do with photography, computers, and printing.
Yours is for painting. The others have to do with photography, computers, and printing.
iow, the old emitted light vs reflected light debate.
"the old emitted light vs reflected light debate"
I’m not sure it’s really a debate. There are some similarities, yes, but painting with pigments is distinct from the way we think of color on computers. The problem is that sometimes you can confuse the two and, if you’re not careful, you can get some pretty ugly results! I remember some of my experiments from art school that should never have seen the light of day (or any light for that matter!).:)
Hey, I finally figured out how you guys do the smiley faces!:) 🙂
I’m not sure it’s really a debate.
heh. I fumbled over that word for about a minute. I know it’s not a debate. I was going to say something like:
iow, the old emitted light vs reflected light… um, statement – fact – thing.
but didn’t want to sound like a retard. at least you know what I meant. 🙂
Dave,
You would never sound, um, mentally challenged, to me. I admire you for your knowledge and willingness to help and always learn something from your posts! Keep it up!
Jake
I try to stay out of discussions (hey that would’ve been a better word!) like this, as I don’t know much. But I learn more here than I give out. Thanks for the kind words Jake.
Just as a quick antecdote, my first real life experience with the reflective vs. emittive "discussion" was when I was handed a marketing piece (a prospectus) and told to match the color scheme for a web page we were designing. It just couldn’t be done! We got close, but it was never exact.
We got close, but it was never exact
I’ll bet it was exact on somebody’s monitor, somewhere <g>
I’ll bet it was exact on somebody’s monitor, somewhere <g>
maybe before the piece was ever printed. but once I got it, there was no way to get the exact shades of a printed piece to show on a monitor. then add in the fact that no one in the office was running a color managed system (developers or processors) and it looked different on everyone’s system. in the end we got as close as we (I) could and got final approval sign off from the dept who wanted the web page, but not before the emitted light vs reflected light discussion. 🙂
I was under the impression that there were only two mixing systems – additive and subtractive. RGB is the additive system, and it is used when mixing with light, like computer monitors and optical filters. CMY is the subtractive system, and is used when mixing pigments, like a computer printer.
I would have assumed that mixing paint pigments on a palette would be the same as mixing ink pigments with a printer, and that painting would also use the CMY system. Why is RYB used rather than CMY in this case? It can’t be that you can pick any three independent colors from the color wheel and call them primaries, can it? What am I missing?
John,
There is a similar discussion here:
Lundberg02 "colour models – tell me I’m stupid" 1/4/04 8:06pm </cgi-bin/webx?13/19>
Maybe that will help you better than we could.
I think the main problem here is that there is not really a RYB model, more of an RYB gamut. The spectrum of colors available in the standard SWOP CMY gamut is much smaller that the gamut available with pigments mixed for a specific purpose (like PANTONE spot inks, or oil paints). That gamut is larger than even the standard RGB gamut, for that matter. If you use colors composed of pigments that are outside of the SWOP CMY gamut to begin with (bright orange and bright green spring to mind) then the CMY gamut will certainly not apply. It’s not that you’re using a different color model, just that you’re using a wider color gamut. For example: Bright orange cannot be accurately reproduced within either the CMY or RGB gamuts, but it can be described using the CMY or RGB models, in a theoretical sense, and reproduced within the RYB gamut.
The RYB model describes a method of mixing pigments that is not applicable to color reproduction applications of CMY (print) and RGB (video) gamuts. It works for paint because the base colors for RYB fall outside the gamut of CMY. (Try getting a pure blue or red out of SWOP CMY inks. Dull, isn’t it?) Here’s the thing: any color created within the RYB gamut can be described with the CMY model theoretically, because the CMY model deals with how various light wavelengths are absorbed and reflected by pigments. But, it may not necessarily be able to be reproduced withing the CMY gamut, because the light wavelengths are not in the reproducible gamut of the CMY model.
My head hurts.
On Mon, 5 Jan 2004 13:59:53 -0800, wrote:
I think the main problem here is that there is not really a RYB model, more of an RYB gamut. The spectrum of colors available in the standard SWOP CMY gamut is much smaller that the gamut available with pigments mixed for a specific purpose (like PANTONE spot inks, or oil paints). That gamut is larger than even the standard RGB gamut, for that matter. If you use colors composed of pigments that are outside of the SWOP CMY gamut to begin with (bright orange and bright green spring to mind) then the CMY gamut will certainly not apply. It’s not that you’re using a different color model, just that you’re using a wider color gamut. For example: Bright orange cannot be accurately reproduced within either the CMY or RGB gamuts, but it can be described using the CMY or RGB models, in a theoretical sense, and reproduced within the RYB gamut.
The RYB model describes a method of mixing pigments that is not applicable to color reproduction applications of CMY (print) and RGB (video) gamuts. It works for paint because the base colors for RYB fall outside the gamut of CMY. (Try getting a pure blue or red out of SWOP CMY inks. Dull, isn’t it?) Here’s the thing: any color created within the RYB gamut can be described with the CMY model theoretically, because the CMY model deals with how various light wavelengths are absorbed and reflected by pigments. But, it may not necessarily be able to be reproduced withing the CMY gamut, because the light wavelengths are not in the reproducible gamut of the CMY model.
My head hurts.
Oh my, I do wish more on this NG would learn to word wrap their post to about 70 or so characters. My head hurts too from trying to read these long lines. I don’t wonder that your head hurts.
joevan
Maybe I’m beginning to understand, because I’d never heard of a "gamut" until I started using Photoshop. It was clearly some kind of color range, but you seem to be saying that the gamut permits only practically possible values, with the tools at our disposal, while the model allows all theoretically possible values.
In that case, is the reason I find RGB inadequate for representing highly saturated shades of orange because the range is compressed, and an eight bit byte, with values from 0 to 255, has insufficient accuracy? This would mean that fractional bits would be required to satisfy the requirements of the model. Is that right?
