Short answer
Usually because the blue on screen was outside what four inks can print, and the conversion replaced it with the nearest printable colour of the same measured hue — which, for vivid blues, looks violet. A build with too much magenta, magenta running heavy on press or warm viewing light can each push it further. The cure is to choose the printed blue yourself, from a soft proof or a printed swatch, with less magenta than an automatic conversion gives.
1. The blue is outside the printable range and the conversion landed on a violet
Test: Put the HEX value into the CMYK gamut check. If it is flagged outside and the nearest printable colour shown already looks purple on screen, the conversion explains the print.
Conversions keep the hue angle of CIELAB while cutting chroma, and CIELAB's blue hue lines bend: a colour with the 'same hue' as a vivid screen blue but far less chroma reads as violet.
2. The ink build has too much magenta for the cyan
Test: Read the CMYK values of the blue in the file. If magenta is equal or close to cyan, the build is the full blue overprint, which is itself a violet-leaning blue on any paper.
There is no blue ink in four-colour printing. Blue is cyan with magenta on top, and the more magenta there is relative to cyan, the further the result swings towards purple.
3. Magenta ran heavy, or cyan light, on press
Test: Look at the rest of the sheet. If neutral greys lean pink and reds look heavy as well, magenta printed stronger than intended; compare with a proof under the same light if there is one.
A dark overprint of two inks is sensitive to the balance between them, so a small drift in either ink's strength shows first in blues and in greys.
4. Warm lighting is turning the printed blue towards violet
Test: Carry the print to a window in daylight, out of direct sun. If it looks bluer there and more violet under a warm lamp at home, the light is contributing.
Warm lamps give out little blue light, so a blue ink reflects less of what makes it blue and looks darker and redder.
5. The screen showed a cleaner blue than the file contains
Test: Open the file in a colour-managed viewer on another screen set to its standard or sRGB mode. If the blue already looks less electric there, the first screen was exaggerating it.
Wide-gamut screens and vivid picture modes stretch saturated blues, which raises an expectation no print can meet.
Causes are ordered by how often Colourwise expects each to be the explanation — an editorial judgement, so test them all.
A blue that only went duller is the ordinary loss of saturation in print and belongs with muddy prints. A blue that changed family, from blue to violet, is a hue problem, and it has to be located before it can be fixed. Soft-proof the file with the printer's profile, or read the CMYK values of the blue. If the preview is already purple, or magenta is close to cyan, the cause is in the file and no press adjustment will cure it. If the preview is blue and the sheet is purple, look at the rest of the sheet: pinkish greys and heavy reds point to magenta printing strong. If the sheet looks blue by a window and violet under the lamps at home, the light is doing it. Each of those has a different remedy, and the first is by far the commonest.
Colour conversions bring an unprintable colour into range by reducing its chroma while holding its hue angle in CIELAB, the space colour profiles work in. For most colours that is harmless. In the blue region it is not, because CIELAB's lines of constant hue angle are not lines of constant perceived hue: follow one inwards from a vivid blue and the colour drifts towards violet. The table on this page measures the drift by re-reading the hue in OKLCH, a newer space built to keep hue steady. Screen blue sits at about 264°. Its nearest printable colour at the same CIELAB hue sits at about 291° on coated paper and 300° on uncoated — violet, not blue. Royal blue and cobalt drift by only 4° to 6° on coated paper, because they start closer to what the inks can print, but by 11° to 18° on uncoated.
For comparison, the solid overprint of cyan and magenta on coated paper sits at about 278° on the same scale, and process cyan at about 240°. Printable blues that read as blue live between the two.
Printed blue is cyan with magenta over it, and the ratio between them sets the hue. Equal amounts give the full overprint, which already leans towards violet; as magenta falls relative to cyan the colour moves through blue towards cyan. There is no universal recipe, because the balance that reads as blue changes with the ink set and the paper, so pick it from evidence: a printed swatch book for that kind of paper, or a soft proof with the printer's profile on a calibrated screen. Then fix the values in the file as CMYK, so that no later conversion rebuilds them. Starting from a slightly greener blue on screen achieves the same thing by another route, since the conversion then lands on the blue side of violet.
A dark colour made of two inks is the most sensitive thing on the sheet to their balance. Printing is run to tolerances, and a magenta at the strong end of its range with a cyan at the weak end can turn a borderline blue purple while both inks pass. If copies within one run vary between blue and purple, that is the explanation, and it can be measured against the standard or the proof. Light does the rest: warm lamps emit little blue, so blue ink reflects less of what makes it blue and the magenta component shows more. Judge blues in daylight or under a proper viewing light before deciding that anything is wrong.
| Screen blue | sRGB | OKLCH hue as specified | Nearest printable on coated (FOGRA51) | Its OKLCH hue (change) | Nearest printable on uncoated (FOGRA52) | Its OKLCH hue (change) |
|---|---|---|---|---|---|---|
| Electric blue | #0000ff | 264° | #4c348c | 291° (+27°) | #655583 | 300° (+36°) |
| Medium blue | #0000cd | 264° | #453078 | 293° (+29°) | #605479 | 299° (+35°) |
| Indigo-blue | #3333cc | 272° | #45398c | 285° (+13°) | #5e547e | 295° (+22°) |
| Royal blue | #4169e1 | 266° | #4d5ca7 | 273° (+6°) | #5b6299 | 277° (+11°) |
| Cobalt | #0047ab | 260° | #23479d | 264° (+4°) | #515679 | 278° (+18°) |
| Deep navy-blue | #1e3a8a | 266° | #2a3b80 | 270° (+4°) | #51526a | 283° (+17°) |
Why: Uncoated paper cannot print a dark saturated blue, so the conversion travelled a long way inwards along a hue line that bends to violet.
Fix: Choose the uncoated build separately from the coated one, with less magenta, or use a spot ink for the letterhead.
Why: The cyan and magenta balance is drifting on press, and a two-ink dark shows it first.
Fix: Raise it with the printer with several copies in hand; ask them to measure the solids against the standard.
Each statement is labelled by kind — established fact, a standard’s requirement, observed market data, a convention, or Colourwise’s own interpretation or analysis — with the strength of the evidence behind it.
Colourwise analysisModerate evidence
Computed from published data, the sRGB blue primary has an OKLCH hue of about 264°, while the nearest printable colour at the same CIELAB hue angle has an OKLCH hue of about 291° on premium coated paper (FOGRA51) and about 300° on wood-free uncoated paper (FOGRA52).
Based on: sRGB blue converted to CIELAB (D50, Bradford), clipped towards the gamut cusp at constant CIELAB hue using a cusp-triangle model built from each condition's published solids, converted back to sRGB and re-measured in OKLCH; see the table.
Caveat: The gamut model is simplified, and real profiles bend hue in the blue region to varying degrees to counter exactly this effect.
Source: FOGRA51 characterization data (premium coated offset, ISO 12647-2:2013); FOGRA52 characterization data (wood-free uncoated offset, ISO 12647-2:2013); International Electrotechnical Commission (IEC) webstore and catalogue; A perceptual color space for image processing (Oklab)
FactModerate evidence
The author of OKLab identifies poor hue prediction, particularly for blues, as the main weakness of CIELAB, and the CIEDE2000 colour-difference formula includes a term specifically for the blue region.
Source: A perceptual color space for image processing (Oklab); The development of the CIE 2000 colour-difference formula: CIEDE2000
Colourwise analysisStrong evidence
In the FOGRA51 coated offset data the solid overprint of cyan and magenta is L* 24.74, a* 21.12, b* −47.45, which corresponds to an OKLCH hue of about 278°, against about 240° for the cyan solid.
Based on: The published CIELAB values of the two patches, converted to sRGB and then to OKLCH with Colourwise's colour library.
Source: FOGRA51 characterization data (premium coated offset, ISO 12647-2:2013)
Colourwise interpretationLimited evidence
When a screen blue prints purple and there was no proof, the conversion or the ink build is a far more likely cause than a press fault.
Based on: Colourwise reading of the hue drift computed above, which occurs on a perfectly run press, set against press variation that standards hold within a few ΔE00.
Caveat: Not a measured frequency; magenta-heavy press runs do happen.
Reviewed 1 October 2026. Colourwise summarises its sources in its own words and does not reproduce standards text or proprietary colour data. Spotted an error? Tell us.