Online Ruler

How Accurate Is an Online Ruler?

Two things decide whether a screen ruler tells the truth: the pixel density your browser assumes, and how closely you matched the calibration card. Here is how much each one costs you, and how to check your own screen.

The short answer

An online ruler has two separate errors, and only one of them can be calibrated away.

The first comes from the browser. Web browsers treat 96 pixels as one inch by convention, which has nothing to do with how large the pixels on your screen actually are. That error is fixed and computable from your screen's real pixel density.

The second comes from the calibration itself. Once you match the on-screen card outline to a real card, the scale is only as good as that match: the outline spans 85.60 mm, so every millimetre you are off in matching it is about 1.2% of the whole scale.

After a careful card calibration at 100% browser zoom, what is left is the matching error plus the physical effects listed below. Before calibration, the error is whatever the 96-pixel convention happens to give on your screen, and on a laptop or a phone that is often 30% or more.

Why 96 pixels per inch is usually wrong

A CSS pixel is a unit of layout, not a physical length. On a 24-inch monitor with 1920 x 1080 pixels there are about 92 pixels per inch, so 96 of them are close to an inch. On a 13.3-inch laptop with the same 1920 x 1080 pixels there are about 166 pixels per inch, so 96 of them draw an inch at about 0.58 inch of glass. The same convention, two very different rulers.

The table works the same numbers out for common screens. “Uncalibrated 1 cm” is how long a centimetre is drawn before calibration; error is the difference from a real centimetre.

Screen Pixels per inch Uncalibrated 1 cm reads Error
24-inch monitor, 1920 x 1080 about 92 10.5 mm about +5%
27-inch monitor, 2560 x 1440 about 109 8.8 mm about -12%
15.6-inch laptop, 1920 x 1080 about 141 6.8 mm about -32%
13.3-inch laptop, 1920 x 1080 about 166 5.8 mm about -42%
6.1-inch phone, 390 px viewport about 139 6.9 mm about -31%

Those figures assume browser zoom at 100% and a device pixel ratio that does not rescale the page. Zoom in to 125% and every length on screen grows by a quarter, which is why calibration has to be done at the zoom level you will measure at.

What calibration removes, and what it cannot

Calibration works because the reference object has a size fixed by a standard: an ID-1 card is 85.60 mm by 53.98 mm, a US quarter is 24.26 mm across, a 1 euro coin is 23.25 mm across. Matching the outline to one of those tells the page how many screen pixels make a millimetre on your display.

What is left depends on the match:

How far the outline is from the real card Error in the whole scale
0.5 mm over 85.60 mm about 0.6%
1 mm over 85.60 mm about 1.2%
2 mm over 85.60 mm about 2.3%

On a 10 cm measurement, 1.2% is about 1.2 mm. That is the honest ceiling of the method: a screen ruler calibrated against a card is a millimetre-level instrument, not a tenth-of-a-millimetre one.

Some errors calibration cannot touch at all:

  • Parallax. A thick object rests on the glass at one edge and stands away at the other, so its top edge does not sit where its base does. Doubling the distance between object and glass roughly doubles the apparent shift.
  • A raised screen protector or case lip. The card sits on the raised edge rather than the glass, so the match is made a little too small.
  • Rounded screen corners and curved edges. On many phones the glass curves at the bezel, and any graduation drawn there is not flat against the object.
  • Viewing angle. Reading the far end of the ruler from the side adds the same kind of offset as parallax.
  • Soft materials. Fabric and paper stretch, so the number describes the stretched piece rather than the one you started with.

There is also the screen's own state: a fingerprint-covered or wet screen scatters the graduations, and a display with a non-square pixel grid is rare but not unknown.

How to check your own screen in a minute

After calibrating, place one of these on the ruler and compare the reading with its published size:

Object Size Standard
Bank card, long edge 85.60 mm ISO/IEC 7810 ID-1
Bank card, short edge 53.98 mm ISO/IEC 7810 ID-1
A4 paper, short edge 210 mm ISO 216
US quarter, diameter 24.26 mm United States Mint
US penny, diameter 19.05 mm United States Mint
1 euro coin, diameter 23.25 mm European Central Bank

Read all of them, not just one. If every object reads long by the same proportion, the scale is off and needs recalibrating. If only one is off, that object is probably not the size you assumed: worn coins, trimmed paper and non-standard cards all exist.

The card is the best single check because it is both long and straight. A 1 mm error over 85.60 mm is easier to see than the same error over a 19 mm coin.

How it compares with a physical ruler and a caliper

A printed plastic or steel ruler carries graduations fixed in metal or plastic, so it does not depend on a browser, a zoom level or a calibration step. Its weak point is the same as this one: you still have to line the zero up with the edge, and you still read it by eye.

A caliper measures between two jaws that grip the object, so it removes the lining-up error entirely. Digital calipers commonly resolve 0.01 mm, two decimal places finer than anything a screen can show. If a measurement needs that, a screen ruler is the wrong instrument.

Between those two, a calibrated screen ruler is useful: it is already in front of you, it has no battery, and it is accurate to about a millimetre over a card's width.

When not to rely on it

Do not use a screen ruler for anything where a wrong number has consequences: legal descriptions, medical dosing or device sizing, engineering fits and tolerances, safety-critical parts, or selling something by measured size where a return depends on it. Ring size is a common trap: fingers are sized in fractions of a millimetre, so a screen ruler and a coin are not enough — use a ring sizer or a jeweller's gauge.

For everything else — is this screw M4 or M5, does this photo fit a 6 x 4 frame, is this cable 3.5 mm or 2.5 mm — a calibrated screen ruler is quicker than finding the toolbox.

How to get the best reading you can

  1. Calibrate at 100% browser zoom, on the screen you will measure on.
  2. Use a card rather than a coin: its long straight edge makes a mis-match visible.
  3. Check the calibration with the card's short edge before trusting it.
  4. Lay the object flat on the glass and put its edge exactly on zero.
  5. Look straight at the graduations, and read the same way twice.
  6. Take two readings and, if they differ, use the middle.

Frequently asked about accuracy

These are the questions that come up most often when people compare a screen ruler with a real one. The pages on calibration and actual size go through the method and the limits in more detail.

Frequently asked questions

Is an online ruler accurate?

Before calibration it depends entirely on your screen: close on a monitor near 92 pixels per inch, about 30% short on many laptops and phones. After calibration with a bank card at 100% zoom, the remaining error is set by how closely you matched the outline: 1 mm over 85.60 mm is about 1.2%.

How accurate is an online ruler after calibration?

It is a millimetre-level instrument. On a 10 cm measurement, a 1 mm matching error is about 1.2 mm. It is not a caliper and not suitable for anything where a wrong measurement has consequences.

Why do browsers assume 96 pixels per inch?

Because a CSS pixel is a layout unit, not a physical length. Browsers assume 96 pixels per inch; real screens range from about 92 pixels per inch on a 24-inch 1080p monitor to about 166 on a 13.3-inch 1080p laptop, and higher still on phones.

Does calibration fix every error?

No. It removes the browser's guess about pixel density. It cannot remove parallax from a thick object, a raised screen protector, the curved edge of a phone screen, or the stretch in a soft material.

How do I know my calibration is right?

Measure the card's short edge. It should read 53.98 mm, and the long edge 85.60 mm. If both read long or short by the same proportion, recalibrate; if only one is off, the card is not a standard ID-1 card.

Can I measure my ring size with it?

No. Ring sizes are given in fractions of a millimetre and depend on the finger, not on a flat object. Use a ring sizer or a jeweller's gauge instead.

Is a phone or a laptop more accurate?

Both need calibration, and both can reach about a millimetre. A laptop gives a longer scale, which makes a mismatch easier to see, so a large screen is usually easier to calibrate accurately than a phone.

Does screen resolution matter?

Yes, it sets the starting error. A 13.3-inch 1080p panel is about 166 pixels per inch, so the browser's uncalibrated centimetre is drawn about 5.8 mm long; a 24-inch 1080p monitor is about 92, where it comes out about 10.5 mm.