March 25, 20263 min read

QR Code Error Correction Explained — L, M, Q, and H Levels

Understand QR code error correction levels L, M, Q, and H. Learn which level to choose for logos, print, outdoor, and digital use.

error correction qr code technical reliability guide

Error correction is what makes QR codes resilient. A scratched, partially covered, or logo-overlaid QR code can still scan perfectly — if you chose the right error correction level. Here's how it works.

The Four Levels

QR codes use Reed-Solomon error correction, the same algorithm used in CDs and DVDs. It adds redundant data so the original content can be reconstructed even when part of the code is unreadable.

Level	Name	Recovery Capacity	Module Overhead
L	Low	7% of data recoverable	Smallest code size
M	Medium	15% of data recoverable	Default in most generators
Q	Quartile	25% of data recoverable	Noticeably larger
H	High	30% of data recoverable	Largest code size

"Recovery capacity" means the percentage of modules that can be damaged, obscured, or missing while the code still scans successfully.

Which Level to Choose

Scenario	Recommended Level	Why
Digital screen display	L	Clean environment, no physical damage risk
Standard print (flyer, card)	M	Handles minor print imperfections
Adding a small logo (under 10%)	M or Q	Logo covers some modules
Adding a larger logo (10-20%)	H	Need maximum recovery to compensate
Outdoor signage	Q or H	Weather, dirt, UV fading
Industrial/warehouse	H	Dust, scratches, harsh conditions
Maximum data density needed	L	Fewest redundant modules = most compact

QRMax defaults to M and automatically switches to H when you add a logo overlay.

The Size Tradeoff

Higher error correction = more redundant modules = a physically larger QR code for the same data. Here's a concrete example with a 40-character URL:

Level	Version	Modules	Relative Size
L	3 (29x29)	841	Baseline
M	4 (33x33)	1,089	+29%
Q	5 (37x37)	1,369	+63%
H	6 (41x41)	1,681	+100%

This is why using dynamic QR codes matters for print — the short redirect URL requires fewer modules, leaving more room for error correction without bloating the code.

How Reed-Solomon Works (Simplified)

The QR encoder treats data as a polynomial and divides it by a generator polynomial. The remainder becomes the error correction codewords. When a scanner reads the code, it performs the same polynomial division. If the remainder doesn't match, the algorithm can locate and correct errors up to the capacity of the chosen level.

You don't need to understand the math to use QR codes — just know that higher levels sacrifice data density for resilience.

Can I change error correction after creating a code?

No. Error correction is built into the QR pattern during generation. Changing the level means generating a new code entirely.

Does error correction affect scan speed?

Marginally. The scanner needs slightly more processing time for higher correction levels, but on modern phones the difference is imperceptible (milliseconds).

What happens if damage exceeds the correction capacity?

The scanner simply fails to read the code. There's no partial reading or garbled output — QR error correction is all-or-nothing.

How to Add a Logo to a QR Code — logo overlay requires high error correction
QR Code Design Best Practices — complete design guide
QR Code Encoding Formats — data modes and capacity