Converting between RGB and CMYK
Tom Arah explores the RGB and CMYK colour spaces - and how best to get from one to the other.
The two colour modes might look similar.
Computer-based design for print looks like it should be simplicity itself - after all, all it really boils down to is moving the colours you see onscreen onto paper. But, as anyone who's tried producing commercial print knows, controlling its colour is anything but easy and disappointment is the rule rather than the exception. The obvious questions are: why is it so hard, and what can be done about it?
To understand this we first need to think about just what colour is. Effectively our perception of any colour depends on the way the human eye's photo-sensitive receptor cells, called "cones", interpret the different wavelengths of light. We have three different types of cones receptive to different wavelengths within the visible spectrum between infra red and ultraviolet. Different wavelengths produce different responses in each of the cones and it's the particular combination of all three responses that produces the sensation of seeing a particular colour.
Each cone's responsivity overlaps so it's not true to say that the different cones correspond to our perception of red (R), green (G) and blue (B) light but it does mean that with different combinations of these three "primaries" we can reproduce the largest possible proportion of the visible spectrum. Within this light-based RGB colour mixing model the absence of light produces the perception of black while it's the combination of all wavelengths that produces the effect of white (as Newton showed with his prism). Because of the way that the primaries combine together to produce colour, the RGB system is called "additive".
Because the computer monitor deals with light the RGB mixing model is tailor-made for it (the same is true of desktop scanners and digital cameras). By varying the brightness level of red, green and blue phosphors the monitor can produce a huge range or "gamut" of colours. Because of the convenience of storing each primary's contribution as a single byte of information this leads to 256 (2 to the power 8) possible levels for each of the three Red, Green and Blue components which, when combined together, mean that over 16 million possible colours (256 x 256 x 256) can be specified ranging from black (R0, G0, B0) through to white (R255, G255, B255). It's this beautifully efficient 24-bit RGB system that we're all familiar with from the computer's ubiquitous RGB colour mixer.
So far so good. Now all we need to do is get these RGB values onto paper - but this is where we hit the fundamental problem. Colours on the printed page aren't emitted as light as they are from the computer screen - in fact just the opposite. Inks actually work by bsorbing light so that a red pigment produces its effect by absorbing all the wavelengths in white light apart from the red wavelengths which are reflected. This means that the ink-based colour mixing in print is "subtractive" so that mixed colours lead to black and the absence of colours produces the effect of white - it's the effect we all first came across when scribbling with different coloured felt-tip pens.
However we can still produce an equivalent ink-based three-colour mixing system with the widest possible range by finding the "complements" of the RGB primaries. The inks that fill these roles are Cyan (C) which absorbs red light and reflects blue and green, Magenta (M) which absorbs green light and reflects blue and red and Yellow (Y) which absorbs blue light and reflects red and green. The inks are translucent which means that by combining two subtractive primaries you produce the equivalent of an additive primary so that for example overlaid cyan and yellow inks together produce a green reflected light. Working in this way different combinations of the three subtractive primaries, CMY, can be used to produce a printed equivalent of the RGB spectrum.
.but RGB is additive and CMY subtractive.
Unfortunately inks are subject to impurities so that combining maximum levels of each of the inks doesn't produce an absolutely pure black but rather a muddy brown. The solution is to include a separate dedicated blacK ink (indicated as K rather than B to prevent confusion with Blue). This offers a number of advantages but, most importantly, it allows the output of high-resolution unscreened black type with maximum contrast so ensuring maximum readability.
The resulting CMYK colour mixing model or "four-colour process print" is the basis of full-colour commercial print. The system's big advantage over the use of one-off specially concocted "spot colour" inks, such as the Pantone range, is that it can be used to specify a wide spectrum within the same layout. Most importantly, by breaking an image into halftone spots containing varying amounts of cyan, magenta, yellow and black ink (see RW79), the eye can be fooled into seeing continuous tones - essential for the reproduction of photographs. In other words, by breaking down the colours in your pages into just four "separations" for each of the four CMYK inks you can produce any of the full-colour, photograph-filled layouts that you see on the newsstand.
It's a formidably flexible system and handling CMYK is the secret of successful commercial print. That's why in professional DTP applications and print-oriented vector drawing programs, such as InDesign and Illustrator, you're encouraged to specify your colours for drawn objects and type as CMYK rather than as RGB (incidentally as commercial print predates the computer, CMYK colours are usually specified as percentages so that white is 0C 0M 0Y 0K and a solid green would be indicated as 100C 0M 100Y 0K). But we have a problem. The main advantage of CMYK print is the ability to reproduce the continuous tones necessary for colour photographs but, as we've seen, photographic colours are actually stored as RGB values.
Clearly we need to translate between the two colour models. And as each colour space is closely linked, with black and white inverted and each pair of subtractive primaries creating an additive primary and vice versa, mathematically the mapping should be fairly straightforward. In practical terms, it's certainly simple. In Photoshop all you need to do is change the Image> Mode command from RGB to CMYK and all the colours in your image are automatically translated from 24-bit RGB values to 32-bit CMYK values. You then need to save your image to a format that supports CMYK with TIFF being the most common standard for import into your DTP application. It looks as if it could hardly be easier.
Commercial print is built on CMYK plates and inks.
But hang on. Depending on the image you selected you will probably notice that some of the colours in your new CMYK version look different. And sometimes they will look very different - convert an RGB image based on Photoshop's rainbow gradient, for example, and the CMYK version is almost unrecognizable with all the bright saturated colours rendered dull and matt. This is an extreme case of the disappointment every designer will have experienced when seeing the colours in their four-colour process print-run turn out to be a dull travesty of the original onscreen layout.
The usual response to this colour shift is to blame the printers but the effect is unavoidable. The reason is very simple: the CMYK colour space is smaller than the RGB colour space. This isn't too surprising when you remember that CMYK has to produce its gamut with reflected light and impure inks. This also explains why the RGB colours that are unprintable or "out of gamut" are those that are purest and most saturated such as those in our rainbow gradient. Fortunately these saturated colours tend to be comparatively rare in nature but you can see just how much of the RGB gamut is unprintable with Photoshop's RGB colour mixer which indicates unprintable colours with an exclamation mark - move the three red, green and blue sliders and you'll soon find whole swathes of the RGB colour space that are simply out of bounds for process printing!
We have another major problem. So far we've been acting as if there is one fixed CMYK gamut, as if the same percentages of ink will always produce exactly the same colour, but that's simply not the case. The CMYK gamut varies depending on the press and printing conditions and the translation from RGB needs to take this into account. What this primarily boils down to is the actual inks and paper used. Different regions of the world and different press setups tend to use slightly different inks and combine them slightly differently to produce different gamuts. Just as important for a medium dependent on reflected light is the underlying paper and how this interacts with the inks. Clearly the same inks printed on uncoated newsprint and on the sort of coated bright white paper you might find in a coffee table book will produce two very different gamuts. And as the perceived strength of each primary is determined by the size of its halftone dot, the absorbency of the paper must also be taken into account.
There's another complicating factor - the black plate. As we've seen, this is really an added extra to the main CMY primaries which, in an ideal world, would combine together to produce black. One of the advantages of this is that equal percentages of cyan, magenta and yellow can be shifted onto the black plate which helps cut down on the possibility of over-inking while giving extra control over the shadows and tone in an image. There are two main options here the first called "undercolour removal" (UCR) where black ink is used to replace equal amounts of CMY only in neutral gray areas and "gray component replacement" (GCR) where they are replaced even in coloured areas. Another advantage of the dedicated black plate is that you can produce richer blacks say translating R0 G0 B0 to produce a "warm black" with added magenta and yellow or, say for a polar image, a "cool black" with extra cyan.
In other words there are actually any number of ways of mapping the same RGB values to CMYK percentages and, to produce the best possible results, we should customise the conversion across the full colour spectrum taking into account not just to the individual press, inks and paper but the individual image! Not surprisingly it's Photoshop that steps up to the mark. In Photoshop 7 this is done with the the Edit> Colour Settings> Custom CMYK. command available from the CMYK working space dropdown. This opens up a dialog where you can specify ink colours and set GCR, UCR or custom black generation and so on. As you make changes you can see how the neutral RGB colours from 0 to 255 will be translated to each of the cyan, magenta, yellow and black plates as a graph (notice how each ink's graph is non-linear and different to the others). Once the profile is created and selected, it's this that will determine the final CMYK values when you change image mode.
Creating a custom CMYK profile is a black art.
The power's there, and it's useful to see how the CMYK profile sets up and handles the conversion, but don't panic - no-one expects you to use it. Expecting the designer to be a pre-press expert isn't sensible and fine-tuning each individual image's RGB to CMYK conversion just isn't practical. Thankfully these days Photoshop, along with Illustrator and InDesign, provide recommended Prepress presets for each of the main regions, Europe, the US and Japan, with alternatives available for coated and uncoated paper and in the US's case for web offset and sheetfed presses. Check with your bureau to see which preset they recommend for your job or whether they have produced their own profiles (some bureaux prefer you to convert to a well-defined proofing space which they can set their presses to match).
There's one final factor that you need to know about and want to be in control of - how your out-of-gamut colours are handled. This is managed with the Colour Setting dialog's Rendering Intent setting (you'll need to click on Advanced for this to become available). If you're producing a flat colour logo, for example, and don't mind out-of-gamut colours simply being mapped to the nearest printable colour you can select Absolute Colorimetric (though it would make more sense to make sure that you select colours that work onscreen and on paper). If you're producing a business graphic you might be less concerned with colour accuracy than impact, in which case you want to select the Saturation option. For continuous-tone images such as photographs it can be more important to give up on colour matching and to preserve the full tonal range in which case choose Perceptual. And finally for a cross between colour accuracy and maintaining tonal range you can choose the default Relative Colorimetric option.
Presets make life easier while rendering intents offer flexibility.
By now you should have a reasonable idea of how RGB to CMYK conversion works and how to set it up so that you get good results (and understand why achieving perfect results is impossible, and achieving the best possible results is usually impractical). Knowing what's involved you might well prefer to leave it to your printers who are the press experts after all. If you've got a good relationship that might well be best option, but be careful. Many large-scale printers will simply apply a standard conversion while others will refuse to accept the job with unconverted images as they know that they will be blamed for the inevitable colour shifts. If you're very unlucky, the printers will simply output the pages and photos as delivered and, depending on your DTP package, your RGB images could end up being output as grayscales on the black plate!
If you do take the responsibility for conversion yourself, the next question is when should you do it? Some designers argue that, if you know an image is going to end up being separated and used in commercial print, you might as well work in CMYK as early as possible. This does have advantages: colour correcting in CMYK is much more intuitive than managing RGB and with a dedicated black plate it's easier to control shadow detail, for example, or to apply sharpening just to the black channel. If you have to colour match exactly, say, to add a tinted box that matches your corporate colours, or if your image was professionally scanned as CMYK, then it definitely makes sense to stick with CMYK from the offset.
Generally though, the case for editing in RGB is much more convincing. To begin with, as we've seen, the RGB gamut is much wider than CMYK and it's always better to maintain as much colour information as possible as you work. This is especially the case these days when you might well be producing work for screen/web use as well as for commercial print. You also get more editing power in RGB mode with many features such as filters unavailable in CMYK. And, if you're going to be printing to a non-PostScript colour inkjet, you'll get better results if you stick to RGB (this seems counter-intuitive as inkjets clearly build their colours with ink but they aren't limited to the standard CMYK process versions and are optimized to work with computers so are treated as RGB devices). And, on top of all these other advantages, 24-bit RGB is 25% more efficient than 32-bit CMYK both in terms of storage and processing!
The real clinching argument is that, if you're using Photoshop, you can get most of the benefits of working in CMYK while working in RGB. We've already seen that Photoshop marks those RGB colours that are unprintable with an exclamation mark (and clicking on the exclamation automatically converts the colour to the nearest CMYK-friendly colour). Even handier is the Info palette which shows you both RGB values and CMYK percentages and highlights problem colours as you move the cursor over the image. You can even get the benefit of working directly with an image's brightness channel by temporarily converting to LAB mode without the perceptible and irreversible colour shift inherent in converting to CMYK.
Even more useful is Photoshop's soft proofing capability. Using the View> Proof Colors command (shortcut Ctrl+Y) you can get Photoshop to preview how your image would print using the current CMYK profile (using the New Window command first allows you to compare RGB and CMYK versions side-by-side). And because the preview is non-destructive this means that you can swap profiles with the Proof Setup> Custom command to get a good idea of how your image would turn out, say, on coated or uncoated stock.
Photoshop's Info palette and Gamut Warning help you to work to CMYK within RGB.
Best of all is the ability to view all areas of the image that are unprintable with the View> Gamut Warning command (shortcut Shift+Ctrl+Y) which indicates out-of-gamut colours with a gray overlay. Clearly the ideal is to keep these to a minimum, a factor you should always bear in mind when colour correcting images destined for commercial print (you can toggle the gamut warning when the Adjust dialogs are open). You can also use colour correction to bring problem areas back into the fold (usually by lightening or desaturating) and Photoshop conveniently lets you automatically select all out-of-gamut colours from the Select > Colour Range dialog's dropdown.
Working in this way you get the best of both worlds: editing in the RGB gamut but always taking into account and controlling your image's CMYK gamut ready for the final conversion, separation and commercial print. With a colour-managed profile-based workflow you can push the conversion even further downstream - but that's another story and a future article.
Hopefully you've found the information you were looking for. For further information please click here.
For free trials and special offers please click the following recommended links:
For further information on the following design applications and subjects please click on the links below:
[3D], [3ds max], [Adobe], [Acrobat], [Cinema 4D], [Corel], [CorelDRAW], [Creative Suite], [Digital Image], [Dreamweaver], [Director], [Fireworks], [Flash], [FreeHand], [FrameMaker], [FrontPage], [GoLive], [Graphic Design], [HTML/CSS], [Illustrator], [InDesign], [Macromedia], [Macromedia Studio], [Microsoft], [NetObjects Fusion], [PageMaker], [Paint Shop Pro], [Painter], [Photo Editing], [PhotoImpact], [Photoshop], [Photoshop Elements], [Publisher], [QuarkXPress], [Web Design]
To continue your search on the designer-info.com site and beyond please use the Google and Amazon search boxes below:
|designer-info.com: independent, informed, intelligent, incisive, in-depth...|
All the work on the site (over 250 reviews, over 100 articles and tutorials) has been written by me, Tom Arah It's also me who maintains the site, answers your emails etc. The site is very popular and from your feedback I know it's a useful resource - but it takes a lot to keep it up.
You can help keep the site running, independent and free by Bookmarking the site (if you don't you might never find it again), telling others about it and by coming back (new content is added every month). Even better you can make a donation eg $5 the typical cost of just one issue of a print magazine or buy anything via Amazon.com or Amazon.co.uk (now or next time you feel like shopping) using these links or the designer-info.com shop - it's a great way of quickly finding the best buys, it costs you nothing and I gain a small but much-appreciated commission.
Thanks very much, Tom Arah
[DTP/Publishing] [Vector Drawing] [Bitmap/Photo] [Web] [3D]
[Articles/Tutorials] [Reviews/Archive] [Shop] [Home/What's New]