How do QR codes actually work?

Point a camera at a QR code and the link opens in half a second. Here is what happens in that half-second, minus the dense maths.

1. Finding the code

The three big squares in the corners are finder patterns. Their black-white-black proportions (1:1:3:1:1) are unmistakable to software scanning a camera frame, so the phone locates the code and works out its tilt and orientation instantly — which is why you can scan upside down, at an angle, or off a curved mug.

2. Reading the grid

The scanner maps the grid of modules and reads each as a 1 (dark) or 0 (light). Timing strips between the finder patterns tell it how many rows and columns exist — anywhere from 21×21 for a short code up to 177×177 for the maximum version.

3. Error correction — the clever part

QR codes carry Reed–Solomon error correction: redundant data distributed through the grid that lets the scanner reconstruct the message even when part of the code is dirty, torn, reflecting light — or deliberately covered by a logo. Four levels exist: L recovers 7% damage, M 15%, Q 25%, H 30%. Our generator defaults to Q and switches logo codes to H.

4. Acting on the data

The decoded bytes carry a prefix saying what they are: URLs open the browser, WIFI: strings prompt network join, vCards offer to save a contact, tel: offers to dial. That prefix is why one square can do so many different jobs.

Why black on white?

Scanners need contrast, not colour — our generator lets you use any colours and warns if the pair would fail. A dark pattern on a light background is the rule; inverted codes break many older readers, which is why we flag them. The deep-dive on scanning failures: why your QR code isn't scanning.