Static vs Dynamic QR Codes
What the redirect actually buys you, measured on real encoded codes, and which payloads are worse off for it.
16 min readRounded modules or separate dots, with the three corner squares left square, because those are what the scanner finds the code with.
The page opens on rounded modules at High error correction. That pairing is not decoration: rounding spends margin, and High is what puts it back.
✓ Scan-tested: this code reads back correctly.
Contrast 21:1, Strong
Version 3 · 37 modules · print at least 14.8 mm wide
A module is a square of ink, and a scanner decides whether it is dark by sampling inside it. Round the corners and you remove ink from that square. Separate the modules into dots and you remove more, and you open gaps between neighbours that the scanner can sample into by mistake.
That makes shape part of the same budget as contrast, logo coverage and print size. Here is each shape at each error-correction level, encoded and read back at build time.
Square
Full module area. Nothing is given up.
Rounded
Corners clipped, horizontal runs stay joined.
Dots
Every module separated. The most inked area lost.
| Shape | Level L | Level M | Level Q | Level H |
|---|---|---|---|---|
| Square | reads back | reads back | reads back | reads back |
| Rounded | reads back | reads back | reads back | reads back |
| Dots | reads back | reads back | reads back | reads back |
The big squares in three corners are the finder patterns, and they are how a scanner locates the symbol and works out its orientation and scale. A scanner sweeps a line across the image looking for a specific run of dark and light in the proportion 1:1:3:1:1. That measurement depends on where the edges are.
Round those edges and you move the thing being measured, which degrades detection before it degrades decoding, so the code gets harder to find rather than obviously broken. This tool keeps the finder patterns square in every shape, which is why a dotted code here still has three crisp corners. That is not a design choice you can override, and generators that let you round the corners too are selling you a code that is harder to find.
Take the SVG rather than the PNG for a shaped code. Rounded corners and circles are exactly the detail a raster export softens first, and softened edges are what a scanner misreads. If the design also needs no background box, the transparent tool tests each surface separately, which matters more once the modules are already giving up area. For the full identity, the branded code tool combines shape, colour and a logo with every check applied at once.
Yes, when the rest of the code has margin to spare. Rounding removes inked area from every data module, so it spends the same budget that low contrast and small print sizes spend. The grid on this page shows each shape decoded at each error-correction level.
Because they are what the scanner finds the code with. The finder patterns have a fixed 1:1:3:1:1 ratio measured across a scan line, and rounding moves the edges that measurement depends on. This tool keeps them square in every shape, deliberately.
Rounded. Horizontally adjacent modules stay merged into one shape, so a run of five reads as a bar with soft ends. Dots separate every module, which loses the most inked area and is the shape most likely to fail in print.
High. Dots spend margin, so give the code the largest recovery budget available. If you also want a logo or a brand colour, reconsider dots: those spend from the same budget, and three at once is where codes stop working.
Print at the width shown under the preview or larger, and test the actual print. Ink spread on paper closes small gaps, which sometimes helps a dotted code and sometimes blurs it. This is the one design choice worth proofing on the real printer.
Colour, a logo in the middle, artwork behind it, each one measured against what still scans.
What the redirect actually buys you, measured on real encoded codes, and which payloads are worse off for it.
16 min readThe usable square inside a round sticker is 70% of its diameter. Measured minimum sizes for real links, against every common sticker size.
8 min readWhat each platform already ships, the measured cost of each payload type, and how to check an offline claim yourself.
26 min read