A client sends a file where the footer with contact details is set at 6 points. In the PDF on screen it looks crystal clear. After printing, the letters blend into a grey-black smudge, and some serifs simply disappear. The first reaction is usually “something is wrong with your machine”. No, the machine is fine. The problem lies deeper, in the very nature of offset printing, and it is called halftoning.
What exactly is a halftone screen
A halftone screen is the way a printing press tricks your eye into seeing a smooth shade of grey or colour, even though physically there is only one colour of ink on the paper, dot by dot. Offset, unlike an inkjet printer or some digital technologies, cannot lay down ink at half strength. Ink is either present at a given point or it is not. The whole trick is that the ink dots vary in size and spacing, and the eye, from reading distance, blends them into a uniform tone.
It is exactly the same mechanism as in old newspapers, where a photo up close breaks down into a grid of black dots, and from a distance looks like a photograph. The only difference is how densely those dots are arranged, and that is where LPI comes in.
What exactly does LPI mean
LPI, or lines per inch, is the number of screen lines per inch, indicating how densely the rows of halftone dots are arranged. The higher the LPI, the finer and more tightly packed the dots, and thus more detail and smoother tonal transitions. Standard newspaper offset works around 85 to 100 LPI, commercial offset on a good press usually falls in the range of 150 to 200 LPI, and premium jobs sometimes reach as high as 300 LPI.
People confuse LPI with DPI, and these are two completely different things. DPI is the resolution of a device, that is, how many physical dots an imagesetter or printer can place per inch. LPI is the density of screen lines, that is, how those dots are organised into larger tonal structures. A single screen cell at 150 LPI usually consists of a dozen or so by a dozen or so microscopic DPI dots, so that a smooth gradation from white to full tone can be achieved. High DPI with low LPI gives a sharp but coarsely grainy halftone. High LPI with low DPI simply will not work, because there will not be enough dots to build a gradient.
Where does the link between LPI and text sharpness come from
Text, unlike a photograph, theoretically does not need to be halftoned at all, because a letter printed at 100% black is simply a solid patch of ink, with no need to build a halftone. The problem starts where the text is not pure 100% K black, but for example dark grey, a colour made from a CMYK mix, or a very fine typeface placed over a halftoned image background, or somewhere where the text colour is formed by overlapping screens of several colours.
In that case the letter stops being a uniform patch of ink and becomes a composite of several halftone layers, each at a different angle, each built from separate dots. With large text the eye averages it out anyway. With small text, say below 7 to 8 points, the size of a single halftone dot starts to become comparable to the stroke width of the letters, and then the letter loses its sharp contours, because its edge is built from jagged, separate dots instead of a smooth vector line.
A simple mental test: Imagine drawing the letter “a” not with a pen, but by arranging small square mosaic tiles side by side. The larger the tiles relative to the letter, the more jagged and less legible the shape becomes. Exactly the same thing happens with text halftoned at low LPI or at too small a size.
What this looks like in practice, with a concrete example
Suppose you are designing a label for a jar of jam, and you want to fit the ingredient list and nutritional values in the footer at 6 points, in 70% black, because full black seemed too graphically contrasty to you. At a screen ruling of 150 LPI, typical for commercial offset on labels, a single halftone dot at that shade has a width in the range of tens to just over a hundred micrometres. The stroke of a letter at 6 points has a thickness of a similar order of magnitude. The effect is that the letter's contour starts to “blur” into dots instead of being a smooth line, and with slightly poorer colour registration on the press the letter additionally loses contrast.
The same text printed in 100% black, without halftoning, at the same size looks noticeably sharper, because the press lays down a full, uniform patch of ink instead of a grid of dots. This is one of the reasons why experienced graphic designers stick to a simple rule: fine text always at 100% of a single colour, preferably black, never as a percentage mix or as an overlay of several CMYK colours.
The most common mistakes that ruin the sharpness of fine text
The first and most common mistake is setting black text as so-called registration black, that is, 100% in all four CMYK channels instead of pure 100% K. Such text at small sizes almost always comes out blurred, because it consists of four overlapping halftone screens at different angles, and with the slightest press registration mismatch the letters grow a coloured halo. We have seen files where all the text on a business card, including the six-point GDPR clause in the footer, was in four-colour black, because someone copied the settings from a large headline and forgot to fix it.
The second mistake is designing text over a colourful, halftoned background without sufficient contrast. The letter itself may be sharp, but if the background beneath it is built from a coarse screen with low LPI, the eye loses the text's outline in the background noise. A third, less obvious mistake is using typefaces that are too thin, for example light or thin weights, at sizes below 8 points. The thin stroke of such a font can be narrower than a single halftone cell, so the RIP has to round it off or thicken it just to make it printable at all, and the result can be unpredictable.
What to avoid with fine text: Do not set fine text as CMYK registration black at 100/100/100/100. Do not design microscopic text on a halftoned, multicolour background. Do not use thin, decorative typefaces below 8 points if legibility after printing matters to you.
Does higher LPI always solve the problem
Not always, although it helps. Higher LPI gives a finer screen, so the dots are less visible to the naked eye and text built over a halftoned background looks sharper. But higher LPI has its own technological cost: it requires very good colour registration on the press, good-quality paper with low absorbency (because on absorbent paper, such as uncoated offset stock, the ink spreads and the fine halftone dots merge into a blob regardless of LPI), and an imagesetter capable of reproducing fine structures. That is why the choice of screen frequency is always a compromise between the substrate, the press and the expected quality, not a simple “more means better” switch.
The real solution to the problem of fine text sharpness is not cranking up LPI indefinitely, but avoiding halftoning of text wherever possible. Text at 100% of a single colour is not halftoned as a tint, but printed as a solid patch, so its sharpness depends mainly on the plate imaging resolution and the precision of the vectors in the file, not on LPI at all.
How to prepare a file so that fine text comes out sharp
Rule number one, repeated by every experienced prepress operator: set all black text, especially below 10 points, as 100% K, with no admixture of cyan, magenta or yellow. Rule number two: avoid placing small text directly over complex, multicolour photos or gradients, and if you must, add a solid, simple colour plane or an outline underneath so the letter has clean contrast.
The third matter is the resolution of the file itself, because although it is a different parameter from LPI, the two work together. The production standard for offset and digital printing is 300 dpi at 1:1 scale, and this also applies to graphic elements on which you place text, not just photos. A file prepared at a lower resolution will not give the RIP enough data to build a sharp halftone regardless of the LPI set on the press.
File format also matters. A print-ready PDF is preferred, ideally finalised as PDF/X-1a, or a correctly prepared composite PDF, because such a file retains text as a vector, not as a rasterised image. The RIP still halftones vector text at the plate-imaging stage, but it does so with full contour precision, without the losses that result from the letter having already been saved as a bitmap.
Before you send a file with fine text, check
- Whether all fine black text is at 100% K, and not registration black CMYK
- Whether text below 8 points does not sit on a complex, halftoned background without contrast
- Whether you are using a typeface with a sufficiently thick stroke, not an extremely thin light or hairline weight
- Whether the text is saved as a vector, not as an image flattened to a bitmap
- Whether the file has a resolution of 300 dpi at 1:1 scale for graphic elements
- Whether the file is saved in PDF/X-1a format or a correct composite PDF
What about labels and fine legal print
Labels are a separate story, because there fine text, ingredient lists, warnings, codes, is the norm, not the exception. The most common problem in label files does not actually start with LPI at all, but with an incorrect cut outline, missing safety zones and poorly thought-out contrast in the fine typography, which further worsens the legibility of already small letters. If you are designing a label with text at 5 to 6 points, it is worth assuming from the start that every imperfect choice, mixed colour instead of full black, a thin typeface, lack of contrast with the background, will affect legibility more strongly than on larger formats, because there is no margin for error.
Remember one thing from this whole article, if nothing else: fine text does not need higher LPI, it needs full colour and no halftoning. That is the simplest and most effective fix you can make to a file before it goes to press.
Can this still be improved on the press side
Partly, yes. The choice of LPI, screen angles and halftone type (for example round dot versus stochastic screening) are decisions made at the prepress stage, depending on the substrate and the order specification. We match the technology to the print run and the order specification, because a label on film requires a different approach than a catalogue on matte coated stock, and uncoated paper behaves differently from coated paper in terms of ink spread. But no press setting will fix text that was saved in the source file as four-colour black or as a bitmap with too low a resolution. Prepress fixes more here than production itself.
If you are working on a packaging or label project with a lot of fine legal print, it is a good habit to print a test page at 1:1 scale on an ordinary office printer and read it from the distance at which the product is actually read on a shelf. This is not a fully reliable test for offset, but it will quickly catch extremely problematic cases before the file goes to imaging.
Halftoning and LPI are a topic most print customers steer well clear of, because it sounds technical and needlessly complicated. Which is a shame, because understanding this one relationship, that full colour prints sharp while a colour mix is halftoned into dots, saves quite a lot of nerves when the print run is delivered.
LPI screen ruling halftoning offset prepress fine text DPI