Docs · Other tools

Stress test

Takes a code, yours or one it makes, and damages it step by step until it stops scanning, so you know where the limits are before printing.

Checked against the live tool on 27 October 2026 by decoding the codes it made.

All docs

At a glance

Input
Text or a link it turns into a code, or an image of an existing code
Error correction
L, M, Q or H for codes it makes, default M
Decoder
jsQR, in your browser
Shows
The last step that decoded and the first that did not, side by side

The four tests

TestWhat it does
BlurBlurs the code more and more, measured as a fraction of one module
Viewing angleTilts the code away from square, in degrees
ResolutionShrinks the image, measured in pixels per module
One module off a cornerRemoves one module from a corner finder pattern

How results are judged

TestResultVerdict
Blur0.30 of a module or moreGood. This is the figure that usually decides a real scan.
Blur0.20 to 0.29Workable, but tight. Print larger than you think you need.
BlurBelow 0.20Fragile. Shorten the data or print much larger.
Viewing angle45 degrees or moreNormal. Angle will not be your problem.
Viewing angleBelow 45Lower than expected, which usually means the code is already dense.
ResolutionReads at 2.5 px per module or fewerNormal. On screen this code is hard to make too small.
ResolutionNeeds moreNeeds more pixels than usual per module.
CornerFailsExpected, and the point. Corner markers have no error correction.
CornerStill readsUnusual. The decoder recovered, but never rely on this.

The print size it suggests

The summary gives a printed width at 0.4 mm per module, including the quiet zone. For small codes that figure can be under 20 mm (0.8 in). Do not print below 20 mm even so: that floor from Print size and the scan badge still applies.

Found a mistake on this page? Suggest a fix on GitHub