Prepress: preparing artwork for production

What prepress consists of

Whatever the process, preparation answers the same questions: which inks will print this, what each one is responsible for, whether there is enough resolution, whether the equipment will hold the finest parts, and what exactly goes into the file.

• Analysis. Size, resolution, transparency, how many colors are actually in the image.
• Color reduction. How many inks are really needed for the image to still be itself.
• Separation into layers. What each ink carries, and the order they go down in.
• Production allowances. Underbase, overlaps between inks, halftoning.
• Validation. Whether the equipment will hold it, before anything is output.
• Output. Files in the formats production accepts, plus a document that explains them.

What can be automated, and what cannot

What has an unambiguous answer can be automated: assigning pixels to inks at a given tolerance, building an underbase, laying overlaps of a given width, screening at a given line count, checking against the numbers in a profile.

What is a judgement cannot: how many inks the order is worth, what to sacrifice in a difficult gradient, whether a thin line is an acceptable risk. The program shows the consequences and records the decision, but the decision is yours.

Repeatability matters more than speed

The expensive thing in prepress is not time, it is divergence: the same artwork prepared twice giving two different runs. So the assignment is deterministic and every parameter lives in a shop profile rather than in an operator’s memory.

A report goes in the archive: print order, angles, line count, profile and every decision taken. Six months later it is what answers "how was this done".

Where the line is today

ColorLayers finishes the job for screen printing. Other processes, such as digital printing, cutting or packaging, need conversions it does not have yet.

We do not publish pages for processes we do not support. If yours is one of them, say so. It changes what gets built next.

What the checks actually look for

Validation is not one pass over the finished separation. Each stage reports what it can see when it runs, which is why a thin line is raised before the halftone and a clash of angles only at the export.

• Features. Strokes and gaps narrower than the profile allows, and detail too small to judge at all at the given resolution.
• Halftone. Dots below the smallest printable line, gaps that close up, and a working resolution too low for the chosen line count.
• Palette. More inks than the press pass allows, a line count the ink will not hold, a print order that contradicts itself.
• Trapping. Overlaps that could not be laid without colliding with a neighbour, and widths quantised to the output grid.
• Separation. Single confident pixels stranded inside another ink, which print as specks.
• Export. Two inks on one angle, an empty plate, a size or resolution that no longer matches the document.

Nothing is corrected on its own. Each issue states what was measured and what the profile required, and the correction is yours to apply and to undo.

What the archive contains

An export is one ZIP with a folder inside, and you choose what goes in it.

• TIFF films. One per ink, named in print order, at the output resolution.
• Combined TIFF. Every plate in one file, for a RIP that expects that.
• Layered PDF and PSD. One layer per ink, for a shop that finishes by hand.
• Preview PNG. What the stack should look like on the garment.
• Palette map PNG. Which ink went where, in flat colour.
• Report. Line count, angles, dot shape, print order, profile and every issue raised.

Re-exporting the same output within a day costs nothing, and ticking more formats does not cost more: the charge is for the separation, not for the number of files.

FAQ