QR Code Error Correction Explained: Levels L, M, Q, H & Print Sizing Guide
How Reed-Solomon polynomial math recovers damaged barcodes, when to use Level H for custom logos, and the 10:1 distance rule.
- What are the four QR code error correction levels (L, M, Q, H)?
- How does Reed-Solomon error correction recover scratched or torn QR codes?
- Which error correction level is required to place a brand logo in the center?
- What is the formula to calculate minimum QR code print size?
- Why does increasing error correction make the QR code grid denser?
1. How Reed-Solomon Error Correction Works
Invented in 1994 by Masahiro Hara at Denso Wave, Quick Response (QR) codes do not merely store raw alphanumeric strings. They append redundant mathematical polynomials generated by the Reed-Solomon error correction algorithm.
When a smartphone camera scans a QR code that is smudged, ripped, or partially obscured, the Reed-Solomon decoder treats damaged modules as mathematical unknowns in a system of polynomial equations, solving for the original values without requiring external network lookups.
2. The Four QR Code Levels: L, M, Q, and H
The ISO/IEC 18004 international standard establishes four discrete error correction thresholds:
| Error Correction Level | Data Recovery Capacity | Redundancy Overhead | Recommended Application |
|---|---|---|---|
| Level L (Low) | ~7% of codewords restored | Lowest module count | High-density numeric data, clean digital screens |
| Level M (Medium) | ~15% of codewords restored | Balanced module count | Standard marketing default: brochures, restaurant menus |
| Level Q (Quartile) | ~25% of codewords restored | Higher module density | Industrial packaging, curved containers, outdoor labels |
| Level H (High) | ~30% of codewords restored | Highest module density | Custom logo overlays, high-wear factory environments |
3. Why Center Logo Embedding Requires Level H
Placing a brand icon in the center of a QR code works by intentionally "damaging" those central data modules.
Because the scanner cannot read the modules covered by the logo graphic, it relies on Reed-Solomon parity bytes to reconstruct the missing information. With Level H error correction (30% recovery capacity), a central logo covering up to 20% of the total matrix area leaves ample mathematical headroom for immediate optical scanning.
4. The Grid Density vs. Robustness Tradeoff
Adding parity codewords increases the total payload size, which forces the QR code into a higher "Version" (a denser grid with smaller individual module dots).
For a short URL, Version 3 (29x29 modules) at Level M may jump to Version 5 (37x37 modules) at Level H. If printed at small dimensions, denser grids demand higher optical resolution from scanning cameras.
5. The 10:1 Optical Sizing Rule for Print
To ensure swift scanning across consumer smartphones without macro focus hunting, follow the 10:1 scanning ratio: QR Code Print Width = Expected Scanning Distance / 10.
For a handheld restaurant menu scanned from 12 inches away: 12 / 10 = 1.2 inches wide (minimum 1.0 inch). For a subway poster scanned from 5 feet (60 inches) away: 60 / 10 = 6 inches wide.
Frequently Asked Questions
Direct answers to the most common questions people ask online about this topic.
Do static QR codes generated on TTech ever expire?
No. Static QR codes encode text and URLs directly into the 2D matrix pattern. Because they contain no middleman redirect servers, they function indefinitely.
Can you invert QR code colors (light dots on dark background)?
While some modern smartphone camera apps support inverted QR codes, the ISO 18004 standard recommends dark modules on a light background for maximum optical compatibility.
What is the quiet zone around a QR code?
The quiet zone is the solid margin of blank space surrounding the exterior border of the QR code. It must be at least 4 modules wide to prevent adjacent graphics from confusing optical scanner algorithms.