QR code anatomy: every region of the symbol

The map

Every QR code, from version 1 (21 × 21 modules) to version 40 (177 × 177), is built from the same regions. Some are function patterns, fixed geometry the decoder uses to find and sample the symbol. The rest carry your data.

Region	Where	Size	Protected by Reed–Solomon?
Finder patterns	Three corners	7 × 7 each	No
Separators	Around each finder	1 module wide	No
Timing patterns	Row 6 and column 6	1 module wide	No
Alignment patterns	Spread across the grid (v2+)	5 × 5 each	No
Format information	Beside the finders	15 bits, stored twice	Own BCH code
Version information	Near two finders (v7+)	18 bits, stored twice	Own BCH code
Dark module	Fixed cell beside bottom-left finder	1 module	No
Data + error correction	Everything else	Varies by version	Yes
Quiet zone	Outside the symbol	≥ 4 modules per side	No

Function patterns

The three finder patterns are what a scanner locates first: concentric squares whose scan-line signature is the same from any angle. Each is wrapped in a one-module light separator so it never merges with data.

The timing patterns (alternating dark and light modules along row 6 and column 6) give the decoder a ruler for module coordinates. From version 2 up, alignment patterns (5 × 5) anchor the sampling grid against lens distortion and print curvature; version 40 has 46 of them.

Metadata regions

Format information is 15 bits (error-correction level, mask pattern, and its own BCH error correction) written twice, because nothing else can be read without it. From version 7, version information adds an 18-bit, BCH-protected declaration of the symbol size, also written twice. One module, the dark module, sits permanently dark beside the bottom-left finder at row 4 × version + 9, column 8.

The data region

Everything not claimed above holds data and Reed–Solomon codewords, placed in two-module-wide columns zig-zagging up and down from the bottom-right corner. This is the only region the error-correction level protects, which is why a logo belongs in the middle, never on a corner. See how logos work without breaking a code.

The quiet zone

The quiet zone is a border of background colour, at least 4 modules on every side, that the standard counts as part of the symbol. Crowd it and the finder patterns stop being findable, the most common self-inflicted failure in the not-scanning checklist.

How much of a QR code is overhead

For version 1, the fixed regions add up precisely:

Region	Modules
3 finder patterns	147
3 separators	45
2 timing patterns	10
Format information (2 copies)	30
Dark module	1
Total overhead	233 of 441 (53%)

That leaves 208 modules (exactly 26 codewords), for data and error correction. As versions grow the overhead share falls, which is one reason longer payloads are proportionally more efficient.

Knowing the map tells you where styling is safe: the data region tolerates damage up to the error-correction budget; the function patterns and quiet zone tolerate none. When in doubt, verify the rendered code decodes.

FAQ

What are the three big squares in a QR code?

Finder patterns. They are 7 × 7 concentric squares in three corners that let a scanner locate the code and work out its orientation from any rotation. Covering any one of them usually makes the code unscannable, regardless of error-correction level.

What is the small square inside a QR code?

An alignment pattern, a 5 × 5 square that anchors the decoder's sampling grid against distortion. Version 1 has none, version 2 has one, and the count grows to 46 by version 40.

Which parts of a QR code can be damaged safely?

Only the data region, up to the error-correction budget of roughly 7–30% depending on level. Finder patterns, timing patterns, format information and the quiet zone are not protected by Reed–Solomon error correction.

What is the border around a QR code called?

The quiet zone, a margin of background colour at least 4 modules wide on every side. It is part of the specification, not decoration, and removing it is a common reason codes fail to scan.