Recovery codewords add resilience
Reed-Solomon recovery can reconstruct some damaged codewords. It does not preserve a second visible copy of every missing square.
Episode V002 · Source record
QR Code adds recovery codewords to the encoded data. In one fixed Version 3 test, a Level H symbol returned its exact harmless text after an 11 by 11 center block was removed. Move the same-size block over a finder region and both selected decoders reported no detection.
The sourced answer
Reed-Solomon recovery can reconstruct some damaged codewords. It does not preserve a second visible copy of every missing square.
A reader must locate, orient, and sample the grid before correction can recover the payload. Damage to a finder region can prevent that earlier stage.
The familiar Level H figure describes approximate codeword restoration, not permission to remove thirty percent of any visible area.
Original QR damage model
The model generated every symbol locally with the plain-text payload HIDDEN LOGIC V002 TEST PAYLOAD. It then applied module-aligned white squares and decoded every saved case.
Deterministic digital test. Not a universal scanner rating.
All clean controls returned the exact payload. zxing-cpp and Apple Vision agreed on the selected center-pass, finder-fail, and larger-center-fail outcomes.
Equal-area location test
The Level H symbol with an 11 by 11 center block returned the exact fixed plain-text payload in both selected decoders.
Move the same block over the top-left finder region and neither selected decoder detected a symbol.
Damage location can matter because detection, format reading, sampling, and codeword recovery are different jobs. The model does not rank every possible location.
Direct source library
Each note states what its reference supports and where the evidence stops. External sources open in a new tab.
Standard and mechanism
These public records support the QR Code scope, selectable correction levels, Reed-Solomon recovery, symbol versions, and reading conditions.
International Organization for Standardization
The public record identifies the fourth edition of the QR Code symbology specification and its broad scope. The paid normative text is not quoted or reproduced.
DENSO WAVE
The page lists L, M, Q, and H as selectable correction levels with approximate codeword restoration figures. Those figures are not guarantees for the same percentage of visible square area.
DENSO WAVE
The page describes Reed-Solomon recovery code added to the original data, byte-level correction, and the capacity tradeoff created by stronger correction.
DENSO WAVE
This overview describes codewords, the three corner position patterns, and the fact that restoration can still fail depending on damage conditions.
DENSO WAVE
The page establishes the Version 1 through Version 40 size progression. The original V002 test uses only Version 3 at 29 by 29 modules.
DENSO WAVE
The page describes the four-module quiet zone around a standard QR Code. One clean digital test is not production guidance for cropped or printed symbols.
DENSO WAVE
The page explains why usable physical size depends on printing and reader resolution. V002 does not claim a universal phone-camera result.
DENSO WAVE
The FAQ supplies the trademark and editorial-use boundary. This page does not reproduce publisher figures, logos, interfaces, or trade dress.
Implementation evidence
The selected zxing-cpp 3.1.1 files show one implementation path from detection through parsing, block construction, and Reed-Solomon decoding. They are not the source code inside every reader.
zxing-cpp contributors
The detector source supports the bounded explanation that a reader first has to locate and sample a plausible symbol before payload recovery can succeed.
zxing-cpp contributors
The parser source provides implementation evidence for reading format and mask information and extracting symbol codewords.
zxing-cpp contributors
The decoder source supplies implementation evidence for Reed-Solomon correction and subsequent payload interpretation.
zxing-cpp contributors
The data-block source supports the explanation that data and correction codewords are distributed through blocks rather than mapped as one duplicate visible picture.
zxing-cpp contributors
The encoder source shows one implementation adding recovery bytes, interleaving codewords, and selecting a mask. It is not claimed as every encoder's implementation.
Safety boundary
A technically successful scan can still return a deceptive or harmful destination. V002 uses harmless plain text and performs no network action.
U.S. Federal Trade Commission
The FTC warns that a QR Code can lead to a spoofed site or malware and recommends checking the destination. This safety note does not alter the engineering result.