How QR codes work

QR codes encode data through data analysis, error correction, structure generation, and module placement.

QR codes use a specific encoding process to convert data into a visual pattern of black and white squares, called “modules.” It happens in four broad stages.

1. Data analysis

• The input data is analyzed and converted to binary.
• It is classified into one of four modes: numeric, alphanumeric, byte/binary, or kanji.
• The most efficient encoding mode is selected automatically.

2. Error correction

• Reed–Solomon error-correction codes are added.
• Four levels are available: L (7%), M (15%), Q (25%), and H (30%).
• Higher levels make the code denser but more resilient — it can still be read when partially damaged or obscured.

This tool uses level H (30%), the highest level. That is what lets a logo sit in the center without breaking the code.

3. Structure generation

• Data is arranged into a pattern alongside fixed function patterns.
• Three large squares in the corners (position-detection patterns) let scanners orient the code.
• Alignment and timing patterns help scanners read the code at any angle.
• Version and format information is embedded.

4. Module placement

• Data becomes black and white modules following specific placement rules.
• A quiet zone (margin) surrounds the finished code.
• Masking patterns keep the black/white balance optimal for scanning.

Learn more

• ISO/IEC 18004:2015 — the official QR code specification
• Thonky QR Code Tutorial — a thorough technical walkthrough
• Reed–Solomon codes for coders — the mathematical foundation