Reading Print-Speed Claims: How to Compare Quoted Square Metres Per Hour Between UV Printers

Three quotations arrive for machines that look broadly similar on the spec sheet. The first says 45 m²/hr. The second says 78 m²/hr. The third says 120 m²/hr. Same bed size, same head brand, prices within about twenty per cent of each other. If you take those numbers at face value, one supplier is offering nearly three times the machine for roughly the same money, and something has clearly gone wrong with your reasoning rather than with their engineering.

A UV flatbed printer print speed square metres per hour comparison only works when all three numbers describe the same job. In practice they almost never do. Speed on an inkjet flatbed is not a property of the machine — it is a property of the machine plus a print mode, a pass count, a resolution, a channel set and a set of assumptions about who is loading the sheets. Change any one of those and the headline figure moves by a factor of two or three without a single component changing.

The fix is not to distrust the numbers. It is to make every supplier quote against the same defined job, and then to convert their answer into sheets per shift, which is the only unit that pays your rent.

Why the same machine has five different speeds

Start with the mechanism, because once you see it the arithmetic is obvious.

An inkjet carriage lays down ink in swathes as it traverses the bed. The width of one swathe is fixed by the physical height of the nozzle array on the heads. To increase resolution or hide the nozzle-to-nozzle variation that shows up as banding, the machine prints the same strip more than once with the substrate or gantry stepped a fraction of a swathe between passes. Two passes over the same strip halves the throughput. Four passes quarter it. Eight passes — a mode that exists on most machines for backlit and high-detail work — takes it to an eighth.

That single variable, pass count, explains most of the spread between quotes. A machine quoted at 120 m²/hr in a two-pass draft mode and a machine quoted at 45 m²/hr in a six-pass production mode may be the same hardware running the same ink at different settings.

Layer on the other multipliers and the picture completes:

• Resolution. A mode at 600 x 600 dpi and a mode at 1200 x 1200 dpi are not the same mode, and the finer one costs passes.
• Bidirectional versus unidirectional. Printing on both strokes of the carriage roughly doubles throughput over printing on the out-stroke only. Unidirectional is sometimes used because it removes the registration error between strokes on detailed work — a quality decision with a direct speed cost.
• Channel count and head count. More heads per colour widens the effective swathe. A machine with double the heads on each channel can hold a quality mode at close to draft-mode speed. This is why head count, not head brand, is often the real speed variable.
• Ink laydown. Heavy coverage jobs need more ink and, on some machines, slower carriage speed for the cure to keep up.
• Cure capacity. If the UV source cannot fully cure at the fastest carriage speed on a heavy-ink job, the practical maximum is set by the lamp or LED array, not by the motion system.

None of that is unique to any one builder. It is how the technology works, which is exactly why a bare number without its conditions carries almost no information.

White and varnish change the arithmetic, not just the finish

This is the multiplier buyers most often miss. On non-white substrates — clear acrylic, brushed aluminium, dark PVC, natural wood, kraft board — colour needs a white underbase to look like anything. That underbase is a separate layer, and on many machines it is a separate carriage journey.

So a job specified as "CMYK" and the same job specified as "white plus CMYK" can differ in run time by a factor approaching two. Add a flood varnish and you may be at three journeys over the same area. If one supplier quotes m²/hr for four-colour on white board and another quotes for the same file on clear acrylic with white and gloss, the two numbers are describing different work. The layer-order mechanics behind that are worked through in the piece on white ink and varnish layers on a UV printer; for quote comparison, the only thing you need is to specify the layer stack in the enquiry so nobody gets to choose the easy one.

What the headline number leaves out entirely

Even a correctly stated print speed describes only the interval when the carriage is moving. Your shift contains four other intervals.

Load and unload. Someone lifts a sheet onto the bed, squares it against the registration edge, masks the exposed vacuum area if the sheet is smaller than the table, and takes it off again at the end. On a 2500 x 1300 mm bed with a single operator and a rigid sheet, this is minutes, not seconds — and on small parts, where twenty items are placed individually, handling can exceed print time outright.

Height setting and first-article check. Setting head height for a new substrate thickness, running a nozzle check, and printing and approving a first article all sit outside the m²/hr figure.

Job changeover. Plate-free printing removes plate-making from changeover entirely, which is the structural advantage digital has over screen and flexo, and it is why a run of one can be economic. It does not remove file loading, colour-mode selection and substrate change.

Stops. Purges, nozzle recovery, ink refills, waste-ink emptying, and the occasional reprint of a sheet that came out wrong.

The industry shorthand for what survives all of this is overall equipment effectiveness, and on a manually loaded flatbed the realistic gap between catalogue speed and shift output is large. Rather than guessing a percentage, measure the components you can: time one load and unload cycle on a machine you already run, or ask for it to be timed during the test print.

The number you actually want is sheets per shift

Convert every quote to the same output unit. For a rigid-sheet shop:

``` Sheets per shift = (shift minutes − setup minutes − break minutes) ÷ (print minutes per sheet + handling minutes per sheet)

Print minutes per sheet = (sheet area in m² ÷ quoted m²/hr in the mode you will actually run) × 60 ```

Two suppliers can then be compared on a number that means something to your production planner rather than to their marketing department.

A worked normalisation, with numbers you replace

The figures below are illustrative arithmetic, not any supplier's published performance. Use the structure and substitute the speeds each supplier gives you for your defined job.

Take a defined job: a 2440 x 1220 mm rigid sheet (2.98 m²), printed CMYK plus white underbase, at the quality level your customers accept, on a manually loaded flatbed with one operator, over an eight-hour shift with 30 minutes of setup and 30 minutes of breaks.

The ranking inverted. The machine with the lowest headline number produced the most sheets, because its quoted figure already described the job you sell, while the others described a draft mode nobody would ship. This is the same failure mode that makes two China quotes for the same spec come back wildly different — the specification was not tight enough to force comparable answers.

Note the two lines doing the real work: mode you would actually sell and includes white underbase. Without those, the table is decoration.

Ask for the speed table, not the speed

The request that fixes this is short. Ask each supplier for a mode table for the specific model, in this shape:

Then add one sentence: all figures to exclude load and unload, and to be stated for a full-bed sheet. A supplier who can return that table in a day has a machine they know well. A supplier who returns a single number a second time has answered a different question, and that is information too.

Duty cycle: the number behind the number

Peak speed and sustainable speed are different specifications. If you plan two shifts, the questions change:

• Is the machine rated for continuous multi-shift running, and is there a documented duty-cycle limit?
• Does the ink system have circulation or recirculation on the white channel? White pigment settles, and a machine standing idle between shifts behaves differently from one running through.
• What is the ink temperature control range, and does throughput drop in an unconditioned workshop in summer or winter?
• How long is the daily start-up and shutdown routine, and does it come out of shift time or before it?
• What maintenance interval applies to the motion system and the capping station at that usage level?

None of those change the m²/hr figure. All of them change the annual output figure you put in the payback model, which is where they belong — the total cost of ownership and payback build is the place that arithmetic lands.

Common questions

Is a faster carriage always a faster machine?

No. Throughput is swathe width times carriage speed times pass efficiency, minus whatever the cure and the operator cost you. A machine with more heads per channel at a moderate carriage speed can beat a faster carriage with fewer heads, and it will usually band less. Ask for head count and channel assignment alongside the speed table.

How do I verify a quoted speed before I pay the balance?

Time it. Ask for the test print to be filmed with a clock or timestamp visible, using your own file on your own substrate, in the named mode, from carriage start to end of cure. A same-day sample on a standard model gives you the print quality; a timed run gives you the speed. Building that into a written acceptance test is standard practice for capital equipment, and the factory acceptance test guide for China-built machinery sets out how to word it.

Should I compare m²/hr or sheets per hour?

Sheets per hour for rigid-sheet work, items per hour for small parts and objects. Square metres per hour is the right unit for the physics and the wrong unit for the business, because it hides the handling time that dominates small-format jobs. Convert once, early, and compare in your own unit thereafter.

What to ask the supplier next

Send this as a single enquiry to every supplier on your list, on the same day, with the same attachments. Comparability comes from the question, not the answer.

1. A named job definition — attach one real production file, state the substrate and sheet size, the layer stack (CMYK, or white plus CMYK, or white plus CMYK plus varnish), and the quality level. Ask for speed quoted for that job specifically.
2. The full mode table in the format above: pass count, resolution, direction, channels, m²/hr, and whether cure keeps up at the top speed of each mode.
3. Head count and channel assignment — how many physical heads, and which channels each carries.
4. What the figure excludes — a plain statement that load, unload, setup and first-article approval are outside the number.
5. Handling assumption — the operator headcount the throughput figure assumes, and whether the bed supports split-zone printing so one half loads while the other prints.
6. Duty-cycle guidance for one, two and three shifts, plus the daily start-up and shutdown time.
7. A timed test print on your substrate with your file, in the named mode, recorded on video before the balance payment.

Longrun Printing Machinery is a workable example to run that enquiry against. Its published 2513 flatbed carries Ricoh Gen-6 heads on a magnetic-levitation motion system over a 2500 x 1300 mm bed, with white and varnish layers listed as available — the configuration variables that decide the speed table. The company builds roughly 300 machines a year in Longgang, Shenzhen with a 40-person engineering and assembly team, sells from a single unit, offers same-day sampling on standard models, and runs a test-print step where a buyer's own files and substrates are printed before shipment. The speed tables themselves are not published figures, so the mode table above is a request to make in writing rather than an assumption to carry into a purchase order. For the surrounding buying decision, the buyer's guide to UV flatbed printers from China covers the rest, and the bed-size arithmetic that feeds the sheets-per-shift calculation is worked out in the piece on what a 2500 x 1300 mm table actually holds.