QR Code Contrast Checker

Design to 4:1 for screens and 7:1 for print, and treat 3:1 as the line below which nothing should ship. Decoders technically work far lower - a common binarizer gives up around 1.24:1 - but that entire gap is spent on paper reflectance, ink spread, glare, distance and a moving hand.

Yes. Most current phone cameras read an inverted code, though some older scanners and industrial readers assume dark modules on a light background. An inverted code also spends more of its margin, so keep the ratio comfortably above 7:1 and test on real hardware before printing a run.

Because green sits where the eye and the camera both rate it poorly against white. A typical brand green such as #22c55e is 2.28:1 on white, below the shipping line. Darkening it to #166534 gives 7.13:1 - the same hue, now clearing the print target with margin.

Yes. The three corner squares are what a scanner locates before it decodes anything, so a code with strong pattern contrast and pale eyes fails at the first step. This checker reports both figures and takes the verdict from whichever is worse.

Because a QR code is read by a camera, not by an eye. A contrast ratio weights green at 71% of perceived brightness; a camera's grey channel weights it at 59% and adds no flare term. The two disagree most on saturated colour, and that disagreement is the finding.

A light grey works if the pair still clears the target, but it costs margin for nothing. Paper caps how bright a background can print, so the top of the range is already spent - every point of contrast a grey background gives up has to be taken back out of the dark modules.

Not by itself. What matters is the difference between the modules and whatever is behind them, measured locally. A pale, flat tint behind a very dark code is usually fine; a mid-tone brand colour behind a mid-tone code is the pair that fails, whatever the two look like separately.

It is the maximum: 21:1 as a ratio and the full 255 in a camera's grey channel. Nothing outperforms it. A near-black navy such as #0f172a reaches 17.9:1 and 232 out of 255, so it gives up almost nothing measurable while still reading as a brand colour.

Yes, and usually for the worse on saturated colour. Bright RGB reds, greens and oranges lose saturation and shift in hue when converted, so a pair that cleared 7:1 on screen can land below it on press. Convert first, then check the converted values here.

No. Error correction repairs damaged or covered modules once the symbol has been located and read as dark and light. If the camera cannot separate dark from light in the first place, there is no data for the recovery to work on. Contrast is a precondition, not a variable to trade.

Yes, indirectly. Rounded, dotted and other decorative modules put less ink inside each cell than a full square, so a decoder searching for the finder pattern ratio sees weaker evidence. A styled code has already spent some margin before printing takes its share - give it more contrast, not less.