QR tools

These tools answer questions a generator cannot

A generator turns text into a square. It cannot tell whether that square will survive a 45 mm label on a curved bottle, whether the payload you pasted is a calendar entry or an unstructured note, or why a photographed code from last year's flyer refuses to open. The six pages in this section handle those separate questions, each with its own inputs and its own kind of answer.

Choose by the question you are holding

When the decision is which payload type to encode, the format selector maps a task to a scheme and names what that scheme cannot carry. When the artwork is already decided and the label size is not, the physical size calculator turns module count, quiet zone and viewing distance into millimetres. When a printed sample scans on one phone and fails on another, the contrast and quiet zone checker measures the image you upload instead of guessing from a screenshot.

Reading and inspection are separate jobs

Reading a single code belongs on the decode page. Reading forty images from an event folder is a different job: the batch reader walks a selected set and reports what each one contained, so a mismatched or expired destination is visible as a list rather than as forty manual attempts. The content parser goes one level deeper on a single payload: it shows the segments, the declared mode and the exact bytes, which is what settles an argument about a stray character or an unexpected space.

Artwork that is measured, not eyeballed

The SVG label generator produces a vector template at a stated outside width, with the quiet zone included in that measurement rather than trimmed off to make the dark pattern look larger. Vector output matters at press: a raster export that looked sharp on screen can gain soft module edges after scaling, and soft edges are exactly what a camera struggles to resolve at distance.

What runs where

Every tool on this page computes in the browser. Uploaded images, pasted payloads and generated files stay in the tab; nothing is sent to a server and nothing is stored between sessions. That also sets a limit worth stating plainly: none of these pages can verify that a destination is safe, that a contact record is current, or that a printer reproduced the file it was given. They check the code and the geometry, and they leave the decision to you.

Choose a payload format before generating pixels

A QR code is a container, not a promise that every scanner will interpret its contents in the same way. The format selector is useful when the decision is between a web address, plain text, a phone number, Wi-Fi credentials, a vCard or an event. For example, a café menu needs one HTTPS URL that also makes sense when typed into a browser. A staff-only Wi-Fi code contains a network name and authentication fields, so it should not be photographed and treated as public signage. The tool can show a suitable structure; it cannot decide whether the information is appropriate to publish.

Worked sizing case: a 60 mm menu tent

Take a 60 mm outside panel for a code that has a 37 by 37 module matrix. QR Code needs a four-module quiet zone on each side, so the complete panel is 45 modules. The module pitch is 60 divided by 45, or 1.33 mm. That is the number a printer and designer can discuss. If a longer destination raises the matrix to 49 modules, the same panel becomes 57 modules including its border and the pitch falls to 1.05 mm. The size calculator makes this trade-off visible before a production proof discovers it.

Use the contrast check on the finished image

Colour names such as navy, cream and gold are not a scan result. Export the actual PNG, JPEG or rendered PDF page and load that file into the contrast and quiet-zone checker. A design may begin with a dark foreground and pale background, then acquire compression artefacts, a translucent logo panel or a photograph that enters the margin. The checker gives evidence about that image; it cannot inspect a matte laminate, reflection from a metallic carton or a physical crop. Treat a passing image result as the point to print a proof, not the point to skip one.

Batch reading answers an inventory question

Opening one image in the reader answers “what is in this code?” A folder of event badges, shelf labels or draft posters raises another question: which files contain the intended values and which are stale? The batch reader lets a person select a set of images and returns a list that can be saved as CSV. Compare that list with the approved source values. A file that decodes cleanly to last month’s address is still a production error. Conversely, a file that fails locally should be held for image and print diagnosis rather than silently omitted from the approval list.

Content parsing is not destination validation

The content parser exposes segments, modes and bytes for one payload. It is useful when a copied string contains an invisible space, an unexpected non-ASCII character, or a structured field has been assembled incorrectly. It does not visit a URL, test a telephone number, authenticate a Wi-Fi network or decide whether a calendar invitation is trustworthy. Keep that boundary clear in an approval record: parser output establishes what the symbol stores; a separate destination check establishes whether the stored value is the intended, current one.

An SVG is editable artwork, not a print verdict

The label generator produces a vector at a stated outside width so that a layout application does not have to enlarge a screenshot. Preserve the code's square aspect ratio and its quiet border when placing it. A 50 mm SVG can be reduced only if the resulting module pitch and scan distance still work; it must never be stretched into a 50 by 42 mm rectangle to match a box panel. GS1's 2D guidance calls the module dimension the X-dimension and explains why both module size and quiet zone affect scanning. The last test belongs on the real substrate.

A compact local workflow

For a new public link, choose URL in the format selector, generate the code, inspect the downloaded output, calculate its complete panel size, place the untouched asset in artwork, and scan a physical proof. For a suspected damaged label, begin with the reader and contrast checker, compare the source file with the printed object, then use the parser only if the recovered text looks wrong. These are different paths through the same tools. They deliberately avoid a generic “generate and hope” workflow because the useful next action depends on whether the uncertainty is format, geometry, pixels or content.