Aztec Code vs QR Code: Why Your Boarding Pass Uses Aztec and Your Menu Uses QR

Denso Wave released QR in 1994 — one year before Aztec. The two formats are essentially contemporaries, but they were designed for different printing environments.

QR was designed inside Toyota's supply chain to track auto parts on cardboard labels. The scanning conditions were predictable: industrial-grade scanners, controlled lighting, decent print quality. QR's three-corner finder pattern works well when you have those conditions and you want fast, omnidirectional decoding at scale on a factory line.

Aztec was designed at Welch Allyn for the opposite environment: low-quality printing, low contrast, partial damage, no margin for error. Specifically, the design brief was about ticketing and ID — applications where the code might be printed on cheap thermal stock, displayed on a low-resolution mobile screen, or partially smudged by handling. Two design choices fall out of that brief.

First, the bullseye finder. A central concentric-ring pattern is rotation-invariant — the scanner does not need to figure out which corner is which to orient the code. That makes Aztec resilient to skew, rotation, and partial occlusion of the corners. Second, Aztec has no required quiet zone. QR requires a 4-module-wide white margin on all sides to be readable; print a QR too close to the edge of a thermal boarding pass and decoders fail. Aztec can be printed flush against other content and still decode.

Those two choices — no quiet zone, rotation-invariant finder — are why airlines picked Aztec. The boarding-pass print area is roughly 2 inches square on a thermal-printed strip with no margin to spare. QR would have needed either a larger print area (bigger ticket) or smaller, lower-capacity codes. Aztec fits.

The central bullseye is more than aesthetic. It is the structural reason Aztec outperforms QR in a specific failure mode: skewed, rotated, or partially damaged prints.

When a scanner sees a QR code, it has to locate three large corner finders, calculate the perspective transform from those three reference points, and then walk the grid. If one of the three finders is damaged or obscured, the scanner fails. The alignment patterns in larger QR versions help recover from minor distortion, but a missing corner finder is fatal.

When a scanner sees an Aztec code, it locates the central bullseye and then walks outward in a spiral. The reference grid is implicit in the bullseye geometry. Damage to any single edge of the code does not break the decoder, because the decoder is not anchored to that edge. Aztec also includes a Reed-Solomon error correction layer that is configurable in a wide range — typically 23% by default, but adjustable from roughly 5% to 95% of the code's area depending on the application.

The practical consequence: Aztec at high ECC can be partially torn, partially burned, partially erased, and still decode. For applications where the code has to survive being folded into a wallet, soaked by rain, or rubbed against a turnstile reader, that resilience matters. For applications where the code lives on a clean printed surface and is scanned once, the resilience is overkill — and QR's simpler structure and broader smartphone-decoder coverage wins.

The headline differences in one place. Numbers below are from the ISO standards and the published format specifications.

Aztec has carved out a small set of industries where it is the default and QR is almost never seen. The biggest is air travel.

IATA Bar Coded Boarding Pass (BCBP). The IATA BCBP standard — the global specification for airline boarding passes — names PDF417 and Aztec as the permitted 2D symbologies. Aztec is now the dominant choice across most major carriers because the bullseye finder reads reliably on the worst-quality thermal printers at airport check-in counters and the no-quiet-zone property fits the cramped print area of a boarding pass strip. Every paper boarding pass you have scanned at a gate in the last decade was almost certainly an Aztec.

European rail. Deutsche Bahn, SNCF (France), Eurostar, Trenitalia, OBB (Austria), SBB (Switzerland), and several other European rail operators standardized on Aztec for printed and mobile-displayed tickets. The format is specified in the UIC 918.3 standard governing rail e-ticketing across the European interoperability framework. When you scan a mobile rail ticket at a turnstile or against a conductor's handheld reader, that is Aztec.

Government ID and document codes. Some Polish ID cards, certain Russian and Ukrainian passports, and various government-issued forms use Aztec for machine-readable supplementary data. The choice is driven by the same logic as airlines: harsh printing conditions, partial-damage tolerance, no quiet-zone budget.

Pharmaceutical and medical samples (occasionally). Aztec shows up on some lab specimen labels and pharmaceutical packaging where the printing surface is curved or the label may be smudged. Most pharma uses Data Matrix (a different format again, ISO/IEC 16022) for regulatory compliance, but Aztec appears in some niche applications.

Notice the pattern. Every industry where Aztec dominates picked it for the same reason: printing conditions that QR would struggle with, plus an institutional standards body that could mandate the choice across the industry. Where there is no central standards body and consumers are scanning with smartphones, QR wins by default.

The airline story is worth a closer look because it shows how a technical decision propagates into a global standard.

In the early 2000s, IATA needed to standardize machine-readable boarding passes across thousands of airports and dozens of airlines. The technical brief had three constraints. First, the code had to print legibly on cheap thermal printers — the same kind of printer that prints receipts, which produces faded, low-contrast output that degrades within months in a wallet. Second, the print area was fixed by the existing boarding-pass form factor: roughly 2 inches square on the right edge of a paper strip, with no margin to spare for a quiet zone. Third, the code had to scan reliably under fluorescent gate lighting on whatever handheld scanner the airport had bought a decade earlier.

QR failed the second constraint outright. The 4-module quiet zone is mandatory; without it, scanners reject the code. Reformatting the boarding pass to give QR its margin would have meant a smaller code (less capacity for the encoded passenger record) or a larger ticket (cost increase across billions of annual print runs).

Aztec passed all three. No quiet zone meant the code could fill the available print area. The configurable ECC let airlines crank correction up to 25-50% for boarding passes, which absorbed thermal-print degradation. The bullseye finder read reliably on the cheap scanners.

IATA ratified Aztec (alongside PDF417, a 1D-stacked format) as the BCBP 2D symbology around 2005. By 2010, Aztec was the dominant choice. Today, every major airline uses Aztec on paper boarding passes; mobile boarding passes displayed in Apple Wallet or Google Wallet also use Aztec for compatibility with existing airport scanner fleets. The format is locked in because every airport in the world is provisioned to decode it.

Outside airlines, European rail, and a handful of government-document use cases, QR is the default 2D format on planet Earth. The reasons are structural.

Smartphone mindshare. Apple's Camera app gained native QR decoding in iOS 11 (September 2017). Google Lens and the stock Android Camera support QR across every major OEM. When consumers see a square pattern in a marketing context, they hold up their phone and assume "QR." That mental model is worth more than any technical capability difference.

Marketing and packaging. Every restaurant menu QR, every retail product-information code, every smart-packaging QR you scan in a store is a QR code. The QR codes for packaging and labels guide covers the print durability and sizing rules — and assumes QR, because that is what packaging-print houses are configured to produce. The best QR code generators roundup and the permanent QR code generator analysis both treat QR as the default; Aztec generators are a niche corner of the tooling market.

Payments. EMVCo merchant-presented QR (MPQR) is QR-based. Alipay, WeChat Pay, UPI in India, PIX in Brazil, SPEI in Mexico — every major QR-payment scheme uses QR, not Aztec. The standards bodies that could have picked Aztec did not, because consumer smartphone scanning was the binding constraint, not print resilience.

Events and check-in. Conference badges, ticket QRs, event check-in flows — all QR. The events and conferences guide and the event QR codes industry page cover the workflow patterns, again assuming QR because the readers (attendees with phones, staff with handheld scanners loaded with QR-decoding SDKs) are QR-native.

Hospitality. Hotel room QRs, restaurant ordering QRs, spa intake QRs — all QR. The hotels and hospitality guide treats Aztec as nonexistent because in that vertical it effectively is.

If your project involves a smartphone user scanning a code in a normal printing condition, QR is the answer. The only reason to consider Aztec is a constrained-print-area or industry-mandated use case.

A specific point because we get asked about it. Modern iOS and Android cameras decode Aztec natively. iOS 11 added support for QR, PDF417, Data Matrix, and Aztec simultaneously. Google Lens supports the same set on Android. Third-party scanner SDKs (Scandit, Cognex, ZXing) all handle Aztec out of the box.

So technically, a marketing flyer with an Aztec code on it would work — most people's phones would decode it correctly when they pointed the camera at it.

In practice, it would not work, because the consumer behaviour layer breaks before the decoder runs. People associate the word "scan" with QR specifically. The square-with-three-corner-finders shape is the mental image of "scannable code." An Aztec's central bullseye looks different enough that a meaningful share of users would assume it is decorative, broken, or "not the kind of code my phone reads." Even users who try would frequently hold the camera in QR-scan mode (slightly far away, expecting the camera to recognize a corner pattern) rather than the framing that produces the cleanest Aztec read.

If you put Aztec on a consumer marketing surface, you are paying a recognition tax that no technical capability is going to offset. The recommendation is straightforward: outside the industry contexts where Aztec is the standard, do not use it. The QR code best practices guide covers the consumer-recognition rules in more depth.

The most legitimate reason to consider Aztec is extreme print survivability. QR error correction tops out at level H — roughly 30% of the code area can be damaged or obscured and the decoder will still recover the data. Aztec error correction is configurable up to roughly 95% of the code area. That is a meaningful gap for harsh-environment printing.

Useful comparisons. On a clean office-printed flyer scanned with a clean phone in good lighting, the difference is invisible — both formats decode every time. On a shipping label that has been rained on, oil-stained, and then crumpled, QR-H at 30% ECC may fail while an Aztec at 75-95% ECC succeeds. On a sun-faded outdoor sign after 18 months of UV exposure, the high-ECC Aztec has a survival margin QR cannot match. The QR code error correction levels deep dive covers the L/M/Q/H mechanics — which is the direct comparator here, and which makes the gap concrete.

The trade-off: higher ECC consumes more module area, so the same physical code holds less data at higher correction. An Aztec at 95% ECC is mostly correction redundancy with a small payload. For most use cases, QR-H is enough; the cost-per-character efficiency of QR at L or M is dramatically better than Aztec at 75%+.

For production-print scenarios where the codes will be exposed to harsh handling (logistics labels, industrial-environment signage, ID cards in field use), the high-ECC capability is worth modelling. Run the math on expected damage rates and required scan-success rates, then decide if the extra survivability is worth the format-recognition cost (since Aztec readers in your field of use must support it) and the data-density loss.

The narrow set of cases where Aztec is the right choice. Be honest about which one you are in — if none of these apply, use QR.

Industry-mandated standards compliance. If you are an airline implementing BCBP, a European rail operator implementing UIC 918.3, or a government issuer working under a spec that names Aztec, the format is dictated by the standard. No decision to make.

Thermal-printed tickets with no quiet-zone budget. A small ticketing or event company printing thousands of paper tickets on receipt-style thermal printers where the print area is tight. Aztec's no-quiet-zone property genuinely matters here. For most event ticketing, though, the codes are displayed on mobile screens with plenty of quiet zone — see the events guide. QR is fine for mobile event tickets.

Government or healthcare ID where the standard already specifies Aztec. Some national ID cards, lab specimen labels, and medical-document workflows have specs that pre-date or supersede QR adoption. If the spec says Aztec, do Aztec.

Extreme harsh-print scenarios with high-ECC requirements and controlled readers. Industrial logistics, military field documents, harsh-environment signage where you control both the print process and the reader fleet. Aztec's configurable 95% ECC justifies itself when QR-H is not enough and you do not depend on consumer-smartphone scanning.

That is the whole list. If your project is restaurant menus, marketing campaigns, packaging, business cards, event signage scanned by attendee phones, payment QRs, or any other consumer-facing surface, Aztec is the wrong choice for reasons that have nothing to do with technical capability and everything to do with how humans scan codes.

If you are reading this because you saw an Aztec code on your boarding pass and wondered whether you should use one for your business, the answer is no.

QR has won every consumer-scanning market. Smartphone defaults are tuned for QR. Print houses are configured for QR. The permanent QR code generator analysis and the best-of generator roundup cover the tooling and the longevity questions you should actually be answering — which generator, which plan, which cancellation policy. Format selection is a non-question for the use cases EZQR readers care about.

For a URL QR code, generate it free on EZQR — no signup, no watermark, no expiration on static codes. For dynamic redirect codes where you may need to change the destination later, see the pricing page or the deep-dive on which dynamic generators actually keep your codes alive after cancellation. For anything where the QR has to survive a harsh print run, optimize ECC to level H and use the QR best practices guide for sizing and placement.

Aztec is a beautifully designed format that solved an important problem in 1995 and still owns its industries 30 years later. It just is not the format for your business card.