Color separation
for screen printing

One layer, one screen, one ink

Separation is the step where a design becomes a set of printable screens. In spot colour work the logic is direct: every ink you plan to print gets its own layer, and that layer becomes a film, which becomes a screen.

A two-colour design needs two screens. If your design contains a tint of one of those inks — say a 40% version of the red — that tint does not need a third screen. It is printed with the same red ink, using halftone dots that cover 40% of the area. This is the single most useful idea in the whole subject: lighter shades of an ink are not separate inks, they are halftones of it.

The three approaches, and when each applies

Process work needs per-channel dot gain curves, careful angle management to avoid moiré and rosettes, and a calibrated printing setup. Index separation avoids the angle problem by using a random dither, which is why many textile printers prefer it over CMYK for photographic art. Spot colour avoids both problems entirely, which is why it accounts for most of what small shops print.

Deciding how many screens you need

Every extra screen costs emulsion, a frame, setup time and registration risk. So the useful question is not "how many colours are in this file" but "how few inks can carry this design".

1. Count the colours that must be exact — brand colours, a specific red. Those are inks.
2. Look for shades of those colours. A pale version of an ink you are already printing is a halftone, not a screen.
3. Look for colours you can merge. Two near-identical greys almost always become one.
4. Check what black can absorb. Outlines, shadows and texture often live happily on the black screen.

Automatic separation gives you a starting point, not an answer. That is why PrintFilmReady lets you merge two proposed inks into one screen, duplicate one into two, and rename them — the software proposes, the printer decides.

Why automatic separation sometimes gets it wrong

Clustering software works on colour distance. Left to itself in plain RGB terms, it treats light grey and dark grey as being as different as grey and red — so a greyscale-heavy image with a few small red accents can end up spending all its inks on greys and never separating the red at all.

The fix is to cluster the way a printer sees: hue and saturation decide which ink a pixel belongs to, while lightness decides the tint level within that ink. PrintFilmReady weights its analysis this way, which is what allows a small but distinct red patch to claim a screen of its own even when it covers a fraction of the image.

Photographs are a different animal

A photograph contains thousands of blended tones, and no set of six inks reproduces it exactly. What good separation can do is produce a convincing interpretation: a poster-like reduction that reads correctly from a metre away. If you need genuine photorealism on a light garment, that is process work, and it belongs in a calibrated RIP rather than any browser tool.

Common questions

We are new, and reviews, testers and ideas make this better. Write a review · Become a tester · Send an idea · Show us your work