QR Code Error Correction Explained

What Is Error Correction

QR code error correction is based on Reed-Solomon codes — the same error correction algorithm widely used in data storage and transmission (CDs and DVDs use the same technology). By adding redundant data to the QR code, scanners can recover complete data from the redundant information even if part of the code is dirty, obscured, or damaged.

The Four Error Correction Levels Explained

The QR code standard defines four error correction levels:

• L (Low): 7% of codewords can be restored. Lowest density, simplest pattern, suitable for clean controlled environments like digital screens or high-quality printed documents.
• M (Medium): 15% of codewords can be restored. The most common default choice, balancing reliability and code complexity — suitable for most general purposes.
• Q (Quartile): 25% of codewords can be restored. Suitable for scenarios with possible light wear, like product labels or packaging.
• H (High): 30% of codewords can be restored. Most dense pattern but strongest fault tolerance — the required level for outdoor advertising and logo embedding scenarios.

How Error Correction Affects Code Size and Complexity

Higher error correction levels require storing more redundant data, meaning higher code versions and denser patterns for the same data. For example, encoding a simple URL: level L might generate a version-3 (29×29) code, while the same URL at level H might need a version-7 (45×45) code. Dense codes require larger print sizes for reliable scanning, so error correction level selection must be considered alongside use case and print size.

How Reed-Solomon Algorithm Works

Reed-Solomon error correction is based on polynomial interpolation over a finite field (Galois field). Simply put, the algorithm treats original data as coefficients of a polynomial, then computes additional points (redundant data). Even if some points (data blocks) are lost or corrupted, as long as enough points remain, the original data can be recovered through polynomial interpolation. The algorithm is powerful because it can precisely locate and correct error positions, not just detect errors.

How to Choose the Right Error Correction Level

Selection guide:

• Digital screen display: L or M level — simple pattern, best scan performance on high-resolution screens
• General print materials (flyers, business cards): M level — balanced reliability and size requirement
• Product packaging, labels: Q level — handle transport wear and light soiling
• Outdoor advertising (posters, display boards, large prints): H level — handle weather and long-term wear
• Logo embedding: H level required

Misconception: Higher Error Correction Is Always Better

Many people think they should always use the highest error correction level (H) for maximum reliability, but this can sometimes backfire. H level produces a denser pattern that is actually harder to scan at small print sizes because module spacing becomes too small and blurs. Using H level in scenarios not requiring especially high error correction (like digital screens) only increases code complexity with no practical benefit. Choosing the appropriate rather than the highest error correction level is the optimal strategy.

The Importance of Real-World Testing

Any error correction level choice should be verified through real-world testing. Scan tests in the target use environment with phones similar to what your target audience might use (including budget models) are the only way to confirm that the chosen error correction level and size combination is truly reliable. Calculations on paper can never replace real-world scenario testing.