Why CMYK Can't Match a RAL Chip Exactly

Hold a RAL fan-deck chip against a press proof built from the exact CMYK a converter handed you, and the two rarely sit flush. The numbers can be right and the colour still off by a shade. The usual assumption is that the conversion was sloppy, or that a sharper printer would close the gap. Neither is the real story. A coating chip and a process print are different objects, and the distance between them is not one problem but three stacked on top of each other, only one of which a converter can even see.

This is the general version of a question every RAL converter raises and none fully answers on the tool page itself. The RAL to CMYK converter hands you a build and calls it derived, not authoritative; the RAL to Pantone tool hands you a match and a distance score. Both are honest about being approximate. Here is what that approximation is actually made of, and why it survives even a perfect-looking number.

What gamut rules out, and what it doesn't

Any way of making colour can only hit a bounded set of colours: its gamut. Four process inks on paper reach one region; the colours the eye can see form a far larger one. A RAL coating, mixed from its own pigments, can sit anywhere in that larger space, including places four inks cannot reach. The sharpest case is a high-chroma signal colour. RAL 2005 Luminous orange and RAL 6038 Luminous green are fluorescent coatings whose brightness lives well outside any CMYK gamut; no mix of cyan, magenta, yellow and black gets there, because the inks have no fluorescence to give. The converter still returns a build (the nearest in-gamut colour it can manage), but the chip is somewhere the inks cannot follow.

Picture the visible colours as a tongue-shaped region with the much smaller CMYK gamut as a triangle inside it; the saturated coatings are the points that fall past the triangle's edge. Gamut is where most explanations stop, and it is the part the RGB vs CMYK guidecovers for screen-to-print work. But it sets a trap, because it implies the rule is simply "inside the gamut, matchable; outside, not." That is false. Most of the RAL Classic set, the muted greys, browns and dull greens that make up its bulk, sits comfortably inside the CMYK gamut and still will not match exactly. Gamut tells you what is flatly impossible. It says nothing about whether the colours that are possible will actually agree. For that, look at what the converter measures, and what it does not.

The "distance" score isn't a Delta E

When the RAL to Pantone converter reports that the nearest spot ink sits at a distance of 34, it is telling you something real but narrow. That number is the straight-line distance between two colours in RGB, the same arithmetic as the gap between two points on a map, run on the red, green and blue values. It is a fast, honest proximity measure, and it is genuinely useful for ranking: a distance of 6 is closer than a distance of 40. What it is not is a Delta E.

Delta E (written ΔE) is the colour industry's measure of how different two colours look to a person. It is computed not in RGB but in CIELAB, a space built so that equal numeric steps correspond, roughly, to equal perceived steps; the CIE defined the first version in 1976. The two metrics answer different questions. RGB distance asks how far apart the numbers are; Delta E asks how different the colours look. Those come apart badly, because RGB is perceptually lopsided: the eye is far more sensitive to a small shift in some hues, such as greens and neutrals, than to the same numeric shift in deep blues. Two pairs of colours can carry identical RGB distances while one pair looks identical and the other clearly mismatched.

This is why the RAL converters here compute RGB distance and label it as distance, never as Delta E: the underlying colour data holds no CIELAB values, so a true Delta E cannot be calculated from it. Treat the distance score as a sorting key, not a ruling on whether a match will pass. The perceptual question is what a measured Delta E from a spectrophotometer answers, and it is the number print standards quote. The thresholds are well worn: a Delta E around 1 is about the smallest difference a careful observer notices side by side, a just-noticeable difference, with one widely cited study putting that nearer 2.3 (Mahy, Van Eycken and Oosterlinck, 1994). Commercial print standards such as ISO 12647 set their pass and fail tolerances in Delta E, not in RGB distance. None of those numbers come out of a converter; they come out of a meter.

A coating and a print are built differently

Suppose the colour is inside the gamut and the build is good. The chip and the print are still two different physical things, and they return light to the eye by different routes. A coating is a layer of pigment with body: light enters it, scatters among the pigment particles, and comes back out having travelled through the material. A process print is a thin film of semi-transparent ink over white paper, so much of the colour you see is light that passed through the ink, bounced off the paper, and came back through the ink again. One reflects from within a solid; the other is a glaze over a mirror.

That difference changes the colour even when the recipe is right. It is why a saturated coating usually reads with more depth than its printed version, and why finish matters: a high-gloss RAL panel throws a sharp highlight in one direction and looks richer everywhere else, while a matte print scatters evenly and reads flatter. The strangest consequence is that the same coating measures as two different colours depending on whether the instrument includes or excludes that surface shine, which is why coating colours are specified with a measurement geometry attached, something a printed swatch never needs. A worked version of this for one popular colour is in RAL 5015 in CMYK; the point here is general. No adjustment to the ink percentages changes the fact that you are comparing a solid to a glaze.

The match that breaks when the light changes

Here is the part that catches people who have done everything right. You proof the job, hold it against the chip under the shop's daylight lamps, and it matches, genuinely matches, close enough to sign off. The run ships, the customer installs it beside the coated product under warm store lighting, and now the two are visibly different. Nothing changed but the light.

This is metamerism, and it is the deepest reason a print cannot reliably match a coating. A colour is really a whole curve: how much of each wavelength a surface reflects. A RAL coating produces its curve with one or two pigments. A CMYK print rebuilds what looks like the same colour by stacking cyan, magenta, yellow and black, which gives a completely different curve that merely happens to integrate to the same appearance under one particular light. Two different spectra that look identical under a given light are a metameric pair, and the agreement is conditional: change the light from daylight to incandescent to LED, and the two curves, which were never the same, stop matching.

Metamerism is why a "perfect" proof is only perfect under the light you proofed it in, and why serious sign-off happens in a controlled viewing booth at a standard illuminant, commonly D50, rather than under whatever lamps the shop runs. It is also why no converter can promise a match. The tool starts from a single RGB value per colour, which encodes appearance under one assumed light and discards the spectral curve entirely; the information that would predict a metameric failure is not in the data the conversion begins with.

Getting as close as the process allows

None of this makes the print hopeless. It means you treat the converter's output as the first move, not the last. The workflow that gets you as close as the process allows is short, and the order matters:

1. Soft-proof against the real profile. Run the CMYK build through your actual printer-and-paper ICC profile, not a generic preview, to see on screen how far the in-gamut build sits from the target before spending ink.
2. Print a swatch and compare under the viewing light. Print a patch on the production stock and judge it against the chip under the light the job will live in, not only the shop lamps. Agreement under one light but not another is the metameric warning.
3. Adjust toward the chip and reprint. Trim or add the channel the swatch shows drifting, reprint, and repeat, accepting that a chroma outside the gamut will never fully close.
4. Sign off on a hard proof.Approve a final proof on the real substrate in a standard-illuminant booth. That signed proof, not the converter's numbers, is what the run is matched to.

Start from the RAL to CMYK build, run the CMYK test page first to confirm your own channels are clean (a weak ink skews every match before you begin), and decide early whether CMYK is even the right destination. If the colour has to stay a brand-critical match, a spot ink picked from the nearest Pantone can hold steadier than a process build, a call the CMYK vs RAL vs Pantone guide works through. The converter gets you to the right neighbourhood. Closing the last gap is proofing, lighting and judgement, none of which fit inside a number.