Measure actual printer speed by timing a representative document workload and dividing completed printed pages by elapsed minutes. Record first-page latency separately from sustained output so the result describes the work you actually need to schedule.
Open the Printer Speed Calculator.
How to Measure Actual Printer Speed for Mixed Documents
A useful speed result needs both a numerator and a timing boundary. Write down whether pages means printed page images, physical sheets, or completed document sets. Then define precisely when the clock starts and stops.
For a mixed-document workload, use the files, paper, colour mode, duplex setting, and quality setting that the job requires. A plain-text draft-mode test does not establish the time needed for a presentation containing photographs, charts, transparency effects, and heavy PDF processing.
The calculation is straightforward:
Observed pages per minute = completed printed pages ÷ elapsed minutes
The calculator uses that observed count and elapsed time directly. It does not need another colour, duplex, or quality multiplier after the test: those conditions have already affected the measured time. Report the conditions alongside the result so someone else can understand and repeat the measurement.
Select a Workload That Represents the Queue
Choose a small but representative group of files. Include the kinds of pages that occupy meaningful portions of the real queue, rather than selecting only the fastest or most complicated example. Keep the sample unchanged when comparing printers or settings.
For example, an office might test a 12-page text report, an 8-page spreadsheet export, and a 10-page PDF presentation. Together these contain 30 page images per set. Five sets produce 150 printed page images, even if duplex printing uses fewer physical sheets.
Record how those sets are submitted. One print command requesting five copies can behave differently from five separate submissions, particularly when file transfer and rendering happen once for a repeated job. Neither workflow is inherently wrong; the relevant one is the workflow you intend to schedule.
Keep paper and media settings consistent. Heavy stock, specialty labels, small sheets, and supported high-quality modes can use different transport or drying behaviour. If the production job uses more than one media type, measure those parts separately or reproduce the actual sequence.
Exclude confidential material from a test file when a neutral representative document is sufficient. The useful characteristic is the workload’s complexity and output mode, not the real customer names or financial figures it happens to contain.
Define Three Timing Boundaries
Different boundaries answer different questions. End-to-end timing helps answer when a user will receive the completed job. First-page latency explains the delay before output begins. Sustained output helps plan a longer run once the printer is producing pages.
| Measurement | Start | Stop | Useful interpretation |
|---|---|---|---|
| First-page latency | Submit the job | First completed output arrives | Initial waiting time |
| End-to-end job time | Submit the job | Last required output arrives | Total user-visible completion time |
| Sustained interval | After an initial defined portion | Final page in the timed portion arrives | Ongoing output rate |
| Finished-job time | Begin production preparation | Accepted sets are ready | Includes handling and finishing |
Be consistent about completion. A duplex sheet is complete only after its second side has printed. Stopping when a sheet first enters the duplex path understates the time needed to deliver it.
HP’s explanation of printing-speed measurements distinguishes test modes and timing boundaries, including first-print timing and measures that exclude an initial document set. This is why a manufacturer’s headline speed and a locally timed short job can differ without either calculation being arithmetically wrong.
Do not start a stopwatch at the first sheet and divide the entire job count by that shortened interval. If the first part of the job occurred before timing began, those pages belong outside the timed numerator too.
Calculate a Mixed-Document Test
Suppose the 150-page-image workload finishes 7 minutes 30 seconds after submission. Convert the time to minutes: 7 + 30 ÷ 60 = 7.5 minutes. End-to-end throughput is 150 ÷ 7.5 = 20 printed pages per minute.
Assume the first 30-page set is complete at 1 minute 50 seconds, and all five sets finish at 7 minutes 30 seconds. The remaining four sets contain 120 page images and take 5 minutes 40 seconds, or 5.6667 minutes. Sustained output over that defined interval is 120 ÷ 5.6667 = 21.18 pages per minute.
These are different measurements of the same run. The 20 PPM figure includes the initial submission and first-set processing. The 21.18 PPM figure describes the later four sets. Preserve both labels instead of presenting the faster number as the total-job rate.
If the first completed output arrives at 24 seconds, first-page latency is 24 seconds. Do not automatically subtract all 24 seconds from every future estimate. Some processing and transport overlap, and a different file or printer state may have a different startup delay.
For the calculator, enter 150 completed pages and 7.5 minutes to reproduce the end-to-end rate. For the sustained measurement, enter 120 pages and 5.6667 minutes. Keep the raw minutes-and-seconds observations in the test record so later rounding can be checked.
Count Duplex Pages and Physical Sheets Separately
The three example documents have even page counts: 12, 8, and 10. Duplex output therefore uses 6 + 4 + 5 = 15 sheets per set, or 75 sheets for five sets. Over 7.5 minutes, that is 10 physical sheets per minute alongside 20 printed page images per minute.
A reader who mistakes sheets per minute for page images per minute could infer a twofold difference where none exists. Write the unit in full at least once and maintain the same definition throughout the comparison.
Odd-length documents add another distinction. A 13-page duplex report requires seven sheets but prints only 13 page images. The blank reverse of the final sheet is part of the physical output, not an additional completed printed page.
Adobe’s double-sided printing instructions describe the printer-dependent duplex controls. Confirm the actual output mode with a short test before timing a longer run, especially when a driver can silently retain an earlier simplex setting.
If you measure accepted sets rather than pages, label that rate directly. Five complete sets in 7.5 minutes equal 0.667 sets per minute. That can be a useful scheduling unit when every order uses the same fixed pack, but it is not comparable with a different set’s page rate.
Repeat the Test and Investigate Variation
Run enough repeats to see whether the result is stable. Record the printer’s starting condition: awake and ready, asleep, freshly powered on, or returning from another job. A warm repeat does not represent the first morning job if that job must wait for startup.
Suppose three identical 150-page tests take 7.5, 8.0, and 7.0 minutes. Their individual throughputs are 20.00, 18.75, and 21.43 PPM. The combined-workload rate is total pages divided by total time: 450 ÷ 22.5 = 20 PPM.
Do not casually average rates from tests of very different sizes. A ten-page test and a thousand-page test should not receive equal weight when the question is throughput over their combined production time. Sum the completed pages and the corresponding elapsed times first.
Note interruptions rather than deleting them from the record. A jam, paper reload, network delay, or calibration cycle may be part of normal production availability. You can report an uninterrupted mechanical interval separately, but the schedule still needs an allowance for recurring operational events.
If one run is unusually slow, inspect its job log and the physical sequence. Repeating it under the same conditions can distinguish a recurring limitation from a one-off event. An unexplained outlier is evidence to investigate, not permission to select the fastest result.
Translate the Measurement into a Production Schedule
For a similar long workload, the measured rate provides a starting estimate. At an observed 20 page images per minute, 900 comparable page images require approximately 45 minutes of output time under the same boundary and operating conditions.
Do not add first-page latency again if the 20 PPM value already represents an end-to-end rate for the same type and length of submission. If you instead use a sustained rate, estimate startup separately and explain the combination. Short jobs with repeated submissions can accumulate startup overhead that a long continuous run largely amortizes.
Add the tasks that the stopwatch excluded: loading media, checking proofs, collating, cutting, stapling, packing, and moving completed work. A printer that finishes output in 45 minutes does not establish that the order is ready to collect at minute 45.
Keep print quality and media suitability in the comparison. A faster mode is only useful if it meets the job’s requirements. The HP print media guide explains media characteristics that also need to match the device and production route.
Maintain a small log with file version, page count, copies, sides, media, mode, start condition, raw time, accepted output, and interruptions. Repeating the same defined workload after a major driver or workflow change makes a useful before-and-after comparison without confusing a changed test with a changed printer.
Avoid Decimal-Time Entry Errors
Minutes and seconds are not base-ten decimals. A display reading 3:45 means 3.75 minutes, not 3.45 minutes. Printing 90 page images in that interval gives 90 ÷ 3.75 = 24 PPM; entering 3.45 would incorrectly produce about 26.09 PPM. Convert seconds by dividing by sixty before using the calculator.
For sub-minute tests, the same rule applies. Thirty seconds is 0.5 minutes. Short intervals magnify the effect of reaction time, so use a longer representative run when possible and retain the stopwatch observation alongside the converted value.