What pass count really means, and why the spec sheet speed is not your speed

Every machine is sold on a square-metres-per-hour figure, and almost nobody will get that figure. Not because the number is false, but because it describes a print mode you would rarely choose. Understanding pass count is what lets you read a quotation honestly.

What a pass actually is

The carriage crosses the material, firing ink. The media (or the bed) advances, and the carriage crosses again. In a four-pass mode, the head covers the same strip four times, advancing a quarter of the head’s height each time.

The consequence is that each line of the image is built from different nozzles on different passes. That is the point. If one nozzle is weak, in single-pass printing it draws a continuous white stripe down the whole job. At eight passes it contributes one-eighth of several lines, and the eye cannot find it.

More passes do not just improve quality. They hide the machine’s faults. Roland’s own remedy for banding is to “switch to higher resolution or W-PASS mode” — more passes, papering over whatever caused the band.

Which is why a demo print at high pass count tells you less about a machine than the same file at production speed. See what to inspect at a demo.

The cure system, not the carriage, sets your floor

Here is a sentence from Mimaki’s own specification page for a flagship flatbed, and it is remarkable that they publish it:

“Inks would not be cured when printing speed is faster than that of draft mode.”

The machine can physically move faster than its lamps can cure the ink. The fastest mode a machine can move is not the fastest mode it can print.

Marabu lists what else moves that boundary: “the dryer’s design, age, and performance, as well as the number of lamps and the ink deposit… the colour shade, the substrate, and the printing speed.” Note age. Emitters lose output over their life, so a mode that cured properly last year may not cure now at the same settings — see UV LED vs mercury curing.

What the manufacturers actually publish

Some real, published figures, so you can see the spread across the market:

• Roland MO-180 — 1.90 m²/h
• Roland MO-240 — 2.39 m²/h
• Mimaki UJV100-160 — 23 m²/h
• Canon Arizona 1300 with FLOW — up to 52.8 m²/h
• swissQprint Nyala 5 — around 253 m²/h

An honest admission while we are here: most manufacturers have quietly stopped publishing speed-by-pass-count tables on their public specification pages. We checked Canon, Mimaki, Roland and swissQprint — none carries one. They survive in brochure PDFs. That is not an accident; a table showing production speed next to headline speed is an uncomfortable document.

Our own machine pages publish the speed exactly as the manufacturer states it, including the pass counts where the manufacturer gives them.

Bidirectional and unidirectional

Printing on both strokes of the carriage roughly doubles the number of passes per minute. It also demands that the drops land in the same place travelling left as travelling right, which depends on head height.

Roland states that bidirectional correction “must be performed to match the object to be printed on.” Mimaki says to redo it “after changing media thickness or head height.”

This is why bidirectional errors reappear every time you change substrate thickness. It is not the machine drifting. It is geometry: a taller sheet moves the impact point.

And the single best diagnostic in the shop, which is Epson’s: print the same file unidirectionally. If it cleans up, the problem is bidirectional alignment or head height — not nozzles, not media feed. That one test eliminates half the possibilities in a minute.

How to read a speed claim on a quotation

Five questions. Ask all of them, of every supplier.

1. Which pass mode is that figure? If the answer is a mode name — “Production Plus”, “Quality-Smooth” — ask for the pass count. Mode names are marketing labels, not specifications.
2. Does it include white? A white underbase is an additional layer. On dark or clear material, most of your real work has one.
3. Does it include varnish or primer? Each is another pass.
4. At what coverage? Nobody standardises this, so the honest answer is a demo on your own file.
5. Does the cure keep up at that speed? The Mimaki quotation above exists because sometimes it does not.

And then subtract the time the machine is not printing at all: loading, masking, jigging, unloading. On a flatbed running small rigid items, handling can exceed print time. A machine that prints twice as fast does not make a shop twice as productive.

Where this leaves the buying decision

Speed is the specification most likely to be quoted at its best case and experienced at its worst. It is also the one that determines whether the machine pays — which is why the payback calculator asks for your realistic daily output rather than the machine’s top speed. Those are different numbers, and using the wrong one is how a payback projection ends up wrong by a year.

The honest way to settle it is to bring your own artwork and your own material to a demo, print it at the mode you would actually sell, and time it yourself. Any supplier confident in their machine will let you. Compare machines side by side on the compare page, and see what decides a UV printer price for why head count — the thing that really drives speed — is also what drives the invoice.

Sources

• Mimaki JFX600-2513 specification page — cure speed limitation at faster than draft mode; bidirectional correction after media thickness change
• Roland DG documentation — banding remedy via higher pass mode; bidirectional correction matched to the object
• Epson documentation — unidirectional test as a diagnostic
• Marabu — factors affecting cure, including emitter age
• Published manufacturer specifications: Roland MO-180 / MO-240, Mimaki UJV100-160, Canon Arizona 1300, swissQprint Nyala 5