By Jared Ebersole in Video — 02 Feb 2026

How QR Codes Turn a Link Into a Picture

A QR code is not a random patch of black squares. It is a tiny machine-readable map with orientation marks, encoded data, and built-in error correction.

A QR code looks like visual static until your phone understands it. Then, instantly, the square becomes a menu, a payment page, a boarding pass, a map, a Wi-Fi login, or a YouTube link.

That jump feels like magic because the code is doing several jobs at once. It has to tell the camera where it is, how it is rotated, how distorted it is, what kind of data it contains, and how to reconstruct the message even if part of the square is dirty, glossy, bent, or badly printed.

The three large squares in the corners are the first giveaway that the pattern is not random. They are position detection patterns. A scanner uses them to identify the QR code and determine its orientation, which is why the code can still work when it is upside down or tilted. Smaller alignment patterns help with distortion, especially on larger codes or codes scanned at an angle. Timing patterns, format information, margins, and error-correction data all help the scanner separate the useful message from the visual noise.

Only after that does the phone get to the part people think of as "the code": the data modules, the black and white squares that carry the message.

Imagine encoding YOUTUBE.COM. In QR alphanumeric mode, each allowed character has a value. The letter Y is 34. O is 24. U is 30. Numbers, capital letters, spaces, and a limited set of symbols all fit into this compact table.

The real encoding is a little more clever than converting each letter by itself. Alphanumeric mode usually packs pairs of characters together. A pair is turned into one number using the formula 45 x first character value + second character value, then that number is stored as bits. If there is one character left over at the end, that final character gets encoded separately.

The important idea is the same as the one in the video: letters become numbers, numbers become bits, and bits become a black-and-white pattern.

Once the message is converted into bits, the QR code does not simply write them left to right like text on a page. It fills the grid in a defined path, starting near the lower-right side and moving through the available modules in a vertical zigzag pattern. The code has to route around the position markers, alignment patterns, timing lines, and other reserved areas. It is more like packing a suitcase around fixed objects than filling an empty spreadsheet.

Then comes the part that makes QR codes surprisingly durable: error correction. QR codes use Reed-Solomon error correction, which adds redundant information so a scanner can recover the original message even if some modules are missing or unreadable. That is why a QR code can often survive a logo in the middle, a smudge on the corner, or a phone camera that catches it from a bad angle.

When your phone scans the code, it reverses the whole process. The camera finds the corner markers. The software straightens the image. It reads the black and white modules as bits, applies the error correction, decodes the data mode, turns the numbers back into characters, and reconstructs the original text. If that text is a URL, the phone offers to open it.

This is the elegance of the format. A barcode is one-dimensional: useful, but limited by how much can be stored along a line. A QR code is two-dimensional, so it can hold far more data in a small space and be scanned quickly from different orientations. DENSO WAVE introduced QR Code in 1994 for high-speed reading in manufacturing, especially automotive parts tracking. The design escaped the factory because the core idea was too useful to stay there.

The square works because it is both rigid and forgiving. Rigid enough that every scanner knows the rules. Forgiving enough that the real world can be messy.

That is why a tiny printed box can become a door.