N-Up, Step and Repeat or Cut & Stack for business cards?

Two hundred staff cards, twenty-four names, imposed twenty-one-up and guillotined. The stacks come off the cutter and every pile is a mixture — Ahmed, then Bell, then Ahmed again, then Costa. Someone now has to sort two hundred cards by hand into twenty-four named bundles, and that someone is you, at six o'clock.

The imposition was not wrong. It was the wrong mode. The whole decision comes down to one question asked before anything else.

The question that decides it

Is every card identical?

If yes — one design, one name, five hundred of them — you want Step and Repeat. It takes one artwork and tiles it across the sheet. Every card that comes off the guillotine is the same card, so the pile order is irrelevant by construction.

If no — different names, sequential numbers, a set of varied designs — you want Cut & Stack. It distributes pages across sheets so that when you guillotine the whole printed pile at once and stack the resulting blocks, the cards come out in the order they were in the source PDF.

And if you have a short run of a few unrelated designs and you intend to sort them by hand anyway, plain N-Up is fine and simpler than either.

Step and Repeat: one design, many copies

This is the straightforward case and it is most of business-card work. You give it one page, tell it how many rows and columns, and it fills the sheet.

For an 85×55 mm card on A3 (297×420 mm), the arithmetic with the default 1 mm X gap and 2 mm Y gap: three columns takes 3 × 85 + 2 × 1 = 257 mm of the 297 mm width, and seven rows takes 7 × 55 + 6 × 2 = 397 mm of the 420 mm height. So 21-up on A3, with 40 mm spare across and 23 mm up the sheet.

That vertical 23 mm is where the job gets tight. A 3 mm bleed plus the default 3 mm mark distance plus a 5 mm mark length is 11 mm outside the trim on each edge, so 22 mm of the 23 mm is spoken for. It fits, with a millimetre to spare and nothing for the gripper. If you need real margin, that is the argument for a Custom sheet at SRA3 — 320×450 mm — rather than A3.

Centre-in-sheet is worth turning on here. It splits the leftover 20 mm and 23 mm evenly around the block rather than pushing everything to one corner, which gives the guillotine an even grip and stops the outermost card from sitting a few millimetres from the sheet edge with no room for its marks.

Cut & Stack: when the cards are not identical

Here is the mechanic, because it is the part people find counterintuitive and it is why the mode exists.

With a normal N-Up imposition, sheet 1 carries records 1 to 21 in reading order, sheet 2 carries 22 to 42, and so on. When you cut the whole stack of printed sheets in one pass, the block that comes from position 1 on every sheet contains records 1, 22, 43, 64 — one from each sheet, at intervals of 21. Every block is a scattered sample. That is the six-o'clock sorting problem.

Cut & Stack inverts the distribution. It works out how many records go in each position across the run, and fills that position on every sheet with a consecutive block.

The arithmetic, with real numbers

Take 210 different cards, 21-up, which is 10 sheets. Divide the total by the number of sheets: 210 ÷ 10 = 21 records per position. So:

• Position 1 on sheet 1 carries record 1. On sheet 2, record 2. Through to sheet 10, record 10.
• Position 2 on sheet 1 carries record 11, sheet 2 record 12, through to record 20.
• Position 3 starts at record 21, and so on to position 21 which starts at record 201.

Cut the printed pile of 10 sheets in one pass. The block from position 1 is records 1 to 10 in order, top to bottom. The block from position 2 is 11 to 20. Set the blocks down in position order and you have 1 to 210, in sequence, having sorted nothing.

The general rule, if you want to check a job by hand:

1. Records per position = total records ÷ number of sheets.
2. First record at position n = (n − 1) × records-per-position + 1.
3. Each subsequent sheet increments every position by 1 — the run advances down the pile, not across the sheet.

Which is worth stating plainly: an imposed Cut & Stack sheet looks scrambled on screen. Sheet 1 reads 1, 11, 21, 31 across the top row. If you proof it by reading the PDF you will think it is broken. Proof it by printing the whole set and cutting it.

When plain N-Up is the right answer

N-Up places consecutive pages consecutively. It is the honest default and it is right in two situations.

First, a mixed short run — five different card designs, forty of each, where you are going to keep the sheets separate and cut them in five small passes anyway. Cut & Stack would be solving a problem you do not have.

Second, proofing. Getting a supplied card artwork onto a sheet to check the trim and the bleed does not need any ordering logic. N-Up it, cut one, look at it.

Side by side

Gutters and bleed: the part that costs you cards

Cards are cut on shared edges, and that changes what the gap between them has to be. There are exactly two sane configurations and a range in between that wastes stock and gains nothing.

Butt cut: gap 0, bleed 0

Set the gap to 0 mm and the bleed allowance to 0 mm. Cards sit edge to edge and one guillotine cut produces the right edge of one card and the left edge of the next. This is the most stock-efficient layout and it is right whenever the artwork does not bleed — a white-bordered card, a card with a keyline well inside the trim.

It gives you nothing to be wrong with, though. A guillotine that drifts half a millimetre now takes half a millimetre off one card and gives it to its neighbour, and on a card with a coloured panel that shows immediately. Only butt-cut artwork that can survive it.

Shared bleed: gap = 2 × bleed

If the artwork bleeds, both cards need bleed into the same gutter, and they need it at the same time. So the gap has to be at least twice the bleed allowance. With a 3 mm bleed, set the gap to 6 mm — 3 mm of card A's bleed and 3 mm of card B's, meeting in the middle where the blade goes.

The FastImpose bleed allowance is one scalar applied all round, so this is simple arithmetic rather than four fields to reconcile. Double it, put it in the gap, done.

Then check the marks. FastImpose puts crop marks 3 mm out from the trim by default. On a 6 mm shared gutter, the marks from the card on the left and the card on the right meet exactly in the centre of the gutter — which is fine, and in fact makes a clean cut line for the operator. On a 0 mm butt cut there is no room for a mark between the cards at all, and you cut to the marks on the outside of the block. More on that in the crop marks post.

If the cards print both sides

Most do. Turn duplex on, and be aware that the backs need to align with the fronts to within the tolerance of your cut — a card is small enough that a 1 mm registration error between front and back is visible as a shifted logo. If your backs come out rotated, that is the flip axis and there is a one-sheet test for it in the duplex post.