Screen printing runs one screen per colour, so every job starts with separations: one clean file per ink, correctly registered, with nothing from the other colours bleeding in. When the customer sends an Illustrator file with each colour on its own layer, that is a five-minute job. When they send a PDF that opens in Illustrator as a single flat layer, it can turn into an hour of selecting, isolating and cleaning up by hand.
Often the separations are still in the PDF. They were just never shown to you.
Why the PDF might still have them
Most design tools write layers into a PDF as optional content groups, the PDF format’s own kind of layer. A customer who set up their artwork with one layer per ink and exported to PDF has very likely produced a file where those groups are intact. CorelDRAW can write them when layers are kept in the export. Illustrator does it when “Create Acrobat Layers from Top-Level Layers” is ticked. AutoCAD does it when layer information is included in the PDF export.
The catch is that Illustrator does not read those groups back. It opens the PDF, ignores the group tags, and draws everything onto one layer. The separation is in the file; Illustrator simply does not present it.
Check first, in Acrobat
Before separating anything by hand, open the PDF in Acrobat and show the Layers panel: View, then Show/Hide, then Side Panels, then Layers.
If the panel lists groups named after inks, colours or Pantone numbers, toggle each one. A group that isolates one colour when the others are hidden is a ready-made separation. If the panel is empty, the customer flattened the file before sending it, and you are back to manual separation or a request for the source.
This check takes under a minute and tells you which of the two jobs you actually have.
Getting one file per screen
Acrobat can show and hide the groups but cannot save each one as its own document. Illustrator cannot see them at all. What you need is a step that reads the groups directly and writes each one out as a separate PDF, keeping the page size and position identical so they register when stacked.
That is what PDFLayerExtract does. Upload the PDF and the free check lists every group it finds. If the groups match what you saw in Acrobat, the paid package returns one PDF per group and an Illustrator script that opens them all as named layers in a single document.
From there the screen-print workflow is the normal one: one layer per ink, convert each to black for film output or send to the RIP as a spot separation, add registration marks, and you are at the screen room with the separation the customer built in the first place.
Halftones, underbases and what extraction does not do
Extraction gives you the groups as drawn. It does not build an underbase, choke or spread traps, convert process colour into simulated-process separations, or turn a photograph into halftone dots. Those remain the separator’s job, and a good one is still worth every minute.
What extraction removes is the part that has no skill in it: manually pulling apart artwork that was already separated when it left the designer’s machine.
Two things to check on any extracted separation. First, that the group really contains only its ink; some designers leave a white base or a registration mark on the same layer, and it comes across with it. Second, that overlapping colours are meant to overlap. A group exported from Illustrator with overprint intact will show which shapes knock out and which trap. Confirm it before burning screens.
When there are no groups
Not every PDF that looks separated is separated. A file exported with “flatten transparency” or saved from a raster editor has no groups at all, whatever the colours look like on screen. The free check shows an empty result for these, and that is the correct answer: there is nothing to extract, and you have found that out without paying for a package or spending an hour in Illustrator.
Ask the customer for the working file, or separate by hand, knowing that no tool would have done better.