Real SizerCalibration · method, evidence and refusals
Calibrating an on-screen ruler with a bank card
To calibrate the ruler, match an outline on screen to a bank card held just above the display. It takes about twenty seconds. The evidence behind every number, and every object refused, is on the methodology page.
The calibration, step by step
The monospace line under each step is what the ruler shows.
-
1. Open the ruler and see the assumption named
With nothing stored, the ruler assumes the CSS inch of 96 pixels, whatever your panel is, and says so.
UNCALIBRATED: assumed 96 CSS ppi
-
2. Find a clean, undamaged card
Use a flat ID-1 card with smooth edges. Hold it just above the display; never let it touch, press or slide on the glass. A bowed card or a damaged edge gives no straight reference.
Reference: bank card, ID card or driving license
-
3. Align it just above the outline and drag until the edges agree
Look square-on. Match the long edge if you can; on a phone in portrait, hold the card upright and match it down the screen. Use the short edge only when neither fits, because its band is wider.
324 px wide · steps of 1 px
-
4. Press the button and read your error band
The readout turns green and shows your error band: the card's tolerance plus your alignment, added as bounds.
CALIBRATED: bank card, ID card or driving license long edge, ±1.2%
-
5. Come back, and let it re-check itself
The page re-checks the display on every visit. A changed display, a pinch-zoom still in force, or six months without a new match takes the readout out of green and names why.
CALIBRATION MAY BE STALE: your display changed, re-check
The two references offered
A reference needs a tolerance read out of a published standard, two independent sources and a date. The card's is the ISO/IEC 7810:2019 ID-1 row for personalized and returned cards: 85.37 to 85.90 mm across and 53.82 to 54.18 mm tall. A4 serves only as a second opinion. The arithmetic, every source and the fourteen objects withheld are on the methodology page.
| Reference and edge | Nominal | Object tolerance | Alignment slop | Error band | Verified on |
|---|---|---|---|---|---|
| A4 sheet, long edge | 297.00 mm | 0.673% | 0.236% | ±1.0% | 2026-08-09 |
| A4 sheet, short edge | 210.00 mm | 0.952% | 0.333% | ±1.3% | 2026-08-09 |
| bank card, ID card or driving license, long edge | 85.60 mm | 0.351% | 0.818% | ±1.2% | 2026-08-09 |
| bank card, ID card or driving license, short edge | 53.98 mm | 0.371% | 1.297% | ±1.7% | 2026-08-09 |
The cross-check, and what a disagreement means
After a card match, match a sheet of A4 on the ruler. Agreement within 1% earns the cross-checked state. A card at ±1.2% and an A4 long edge at ±1.0% can legitimately disagree by 2.1% with both readings correct, against a gate set at 1%. So a disagreement above 1% means the second object did not confirm the first; it never means the first one was wrong.
What is measured, how it is kept, and what it cannot tell you
What is measured. The band adds the object's permitted spread, from its standard, to a fixed 0.7 mm for lining up two edges by eye.
How it is kept. A match is stored in your browser as device pixels per millimeter, with a snapshot of the display, and compared on each visit.
What it cannot tell you. Whether you picked up the right object, or moved to a different monitor of the same resolution; card bow, a gap, tilt and curved panels are outside the band. The methodology page covers each limit.
Questions
Which card can I use without touching the screen?
Why are only two reference objects offered?
An older-style ID card has a better tolerance. Why is it refused?
Where does the plus or minus 1.2 percent come from?
How often does a calibration need redoing?
Go and measure something
Open the ruler, or start it on the long edge, upright for a phone or on the short edge.