Inch ruler · fractions, not decimals

An inch ruler on screen, marked in fractions — down to a thirty-second where the screen can carry it

Eight inches of imperial scale: whole inches numbered, then halves, quarters, eighths, sixteenths and thirty-seconds told apart by the height of the mark, exactly as they are on a steel rule. Nothing on the face is a decimal, because nobody buys a 0.3125-inch drill bit. Until you match a bank card the scale comes from the CSS specification's fixed inch of 96 pixels rather than from your display, and the readout says so in those words.

Calibrationuncalibrated Referencenone matched px/mm3.780 assumed Error bandnone — not calibrated Unitin Length shown8.00 in assumed
012345678
Scale

This page is set up for inches and the face follows the control above. Millimetres and centimetres are a separate instrument, with a legibility floor of their own.

UNCALIBRATED — assumed 96 CSS ppi

This is not actual size yet. Match a bank card to make it so.

Nothing is being withheld on this display: every graduation in the current scale is drawn.

Match the outline to a bank card

There is no imperial reference object worth using. A US Letter sheet is 8.5 by 11 inches and would be a fine straight edge, but its trim tolerance has not been through a two-source verification here, so it ships carried and refused rather than quietly calibrating from a guess. The object that has been verified is the ID-1 card in your wallet — 85.60 by 53.98 mm under ISO/IEC 7810, which is 3.370 by 2.125 inches — and it calibrates an imperial face exactly as well as a metric one, because what it produces is a pixels-per-millimetre ratio and the rest is arithmetic.

Hold the card flat against the outline below and drag until the edges agree. The long edge earns ±1.2% and the short edge ±1.7%; on a phone in portrait the long edge will not fit across the glass, so stand the card upright and match it down the screen instead, which keeps the full 85.60 mm reference and the tighter band. What the objects are, where their dimensions come from and which ones were checked and withheld is on the calibration page.

324 px wide · steps of 1 px · 129–1369 px for this edge
Edge you are matching
How the card is held

One match covers all five calibrated rulers on this site; the screen ruler needs none, because it measures what the browser laid out rather than what you hold against it. If you would rather check it against a second object before trusting it, the sheet-of-A4 cross-check lives on the actual-size ruler.

What each mark is worth, and what it costs to draw

Every level of an imperial scale is half the one above it, so the density each one needs doubles the whole way down — and because a mark needs three device pixels of pitch to survive as a mark, the thresholds come out as a clean run of powers of two. The last of them is the interesting one: a thirty-second of an inch needs exactly 96 pixels per inch, which is the same 96 the CSS specification hands out to every display in the world whether or not it has them. That coincidence is why an uncalibrated inch ruler always looks like it can resolve thirty-seconds, and why so many of them keep drawing the marks after a calibration says otherwise.

MarkStepDevice px per mm neededScreen density neededOn a 24-inch 1080p panel (3.61)
Inch marks, numbered25.4 mm0.1183 ppiyes
Half-inch12.7 mm0.2366 ppiyes
Quarter-inch6.35 mm0.47212 ppiyes
Eighth3.175 mm0.94524 ppiyes
Sixteenth1.5875 mm1.89048 ppiyes
Thirty-second0.79375 mm3.78096 ppino

A 27-inch 1440p panel sits at 4.28 device pixels per millimetre and does keep its thirty-seconds; a 24-inch 1080p at 3.61 does not. Neither of them is a bad monitor and neither number is a criticism — it is simply what the raster can hold. When a level goes, the note under the ruler names it and quotes the density that would have kept it, which is a different thing from a ruler that draws thirty-seconds on a television.

Fractions of an inch, in decimals and millimetres

Exact values, not rounded ones. The inch has been defined as exactly 25.4 mm since the international yard and pound agreement of 1959, so every row here is arithmetic with no uncertainty in it at all. That is worth saying plainly on a site about measurement error: the conversion is exact and the reading is not, and any error you end up with came from the ruler and the eye rather than from this table.

FractionDecimal inchMillimetres
1/320.031250.79375
1/160.06251.5875
1/80.1253.175
3/160.18754.7625
1/40.256.35
5/160.31257.9375
3/80.3759.525
7/160.437511.1125
1/20.512.7
9/160.562514.2875
5/80.62515.875
11/160.687517.4625
3/40.7519.05
13/160.812520.6375
7/80.87522.225
15/160.937523.8125
1125.4

What is measured, how it is kept, and what it cannot tell you

What is measured. Not inches — millimetres. The only thing this page measures is how many pixels of your display a millimetre occupies, taken from an object whose size a published standard fixes, and every inch on the face is that ratio multiplied by 25.4. There is no separate imperial calibration and there could not be a better one, because no imperial reference object in a typical pocket has been through the same verification. Marks are placed at their exact fractional pixel positions and never snapped to the pixel grid: a thirty-second of an inch is three device pixels wide at the threshold, and rounding a mark to the nearest pixel would move it by up to a sixth of its own spacing.

How it is kept. What gets stored is device pixels per millimetre rather than CSS pixels per millimetre, so page zoom divides back out of it on Chrome, Edge and Firefox, and an operating-system display-scaling change does the same on Windows and on GNOME, and a calibration survives those. Safari does not report page zoom through the pixel ratio, and a macOS scaled resolution moves the reported pixel count instead of the ratio; neither is corrected, and both drop the readout out of green and ask for another match. Alongside it the page keeps a note of what the display was reporting at the time — pixel ratio, native pixel count, browser chrome height — and compares them on every visit and on a timer, because a pixel-ratio change can fire no events at all. It is one record in your own browser storage, shared with the millimetre ruler and the actual-size ruler and sent nowhere.

What it cannot tell you. Whether you are looking at the same screen you calibrated on. No page in a browser can detect a swap to a different monitor of the same resolution, so a 1080p profile moved from a 24-inch panel to a 32-inch one is a third wrong with nothing on the page looking any different; the tool re-checks itself whenever the display reports something new and asks again after a long gap. It also cannot tell you that your card is bowed, or held at an angle, or that you have matched the outline to the printed border rather than the edge. The band covers the standard's dimensional spread and ordinary alignment slop, and nothing beyond that is claimed for it.

Questions

Why are the marks in fractions instead of decimals?
Because a tape measure, a drill index and a timber yard are all in fractions, and the conversion done in your head is where the mistake gets made. The numbers on the face are whole inches; every mark between them is a half, a quarter, an eighth, a sixteenth or a thirty-second, and which one it is can be read off how tall it is. The table on this page carries the decimal and the millimetre value of each fraction, so the conversion is on the page rather than in your head.
How fine does this ruler go?
A sixteenth of an inch on anything from a laptop upwards, and a thirty-second on a display of 96 pixels per inch or better. A thirty-second of an inch is 0.79375 mm, and three device pixels across that pitch works out to exactly 96 pixels per inch — one pixel of line and two of clear gap, which is what a mark needs before it is a mark rather than grey. Below that density the level is dropped and the page says which level went.
Why did the thirty-second marks disappear when I calibrated?
Because calibrating replaced an assumption with a measurement of your screen, and your screen came in below the density those marks need. A 24-inch 1080p monitor is 91.8 pixels per inch against the 96 a thirty-second requires. Before calibration the face is drawn at the CSS specification's 96 pixels per inch, which sits exactly on that threshold, so the marks you saw were a property of the assumption rather than of your display.
What is 5/8 of an inch in millimetres?
15.875 mm, exactly. Every row of the conversion table on this page is exact arithmetic, because an inch has been defined as exactly 25.4 mm since 1959 — it is a definition, not a measurement, and it carries no uncertainty of its own. Whatever error a reading has comes from the ruler and the eye, never from the 25.4.
Is an inch on my screen a real inch?
Not until you calibrate it. A browser has no way to ask the display how large it is, so CSS defines an inch as exactly 96 pixels whatever the panel underneath happens to be. On a 15.6-inch 4K laptop at 100 percent scaling that CSS inch renders as about a third of a real inch; on a 24-inch 1080p monitor it comes out about 5 percent too long. Matching a bank card replaces the definition with a measurement of your own display.

The other instruments