Behind the Scenes: How We Test Photorealistic QR Codes for Real-World Advertising

← Back to Insights

A StoryCode has two jobs: make people want to look, and let them scan. Creating a photorealistic image that performs both jobs takes more than a compelling prompt. It takes specialised technology, creative judgement and hands-on testing.

At StoryCodes^®, we use advanced AI tools developed internally to create images that scan like QR codes while belonging to the campaign creative. But an image that works inside a generation tool is not automatically ready for a billboard, bus shelter or connected TV Pause Ad.

That is why our process continues beyond the software — with real phones, different viewing distances and changing light.

Advanced AI, built around a demanding creative challenge

A conventional QR code prioritises its machine-readable structure. A StoryCode must preserve that function while creating a visual that makes sense for the brand: a product, a character, a destination or a photorealistic scene.

Our internally developed generation tools combine AI models, specialised methods and adjustable parameters to explore that balance. They allow our team to create and refine different directions, then select promising designs as potential production candidates.

The technology expands what we can create. Testing helps determine what we should use.

“AI gives us endless creative possibilities. But hands-on testing with real phones, at different distances and under different lighting conditions, tells us whether the result is ready to be integrated into a real advertisement — potentially much larger than the image displayed on our computer screen.”

Frederik Rademakers, Founder of StoryCodes^®

Why software checks are only part of the answer

A tool evaluates an image under defined conditions. A viewer encounters it through a camera, on a particular phone, from a particular distance.

In advertising, those conditions vary. Someone may scan a bus shelter while standing nearby, see a billboard from further away or reach for their phone when a Pause Ad appears on the television.

Hands-on testing exposes the design to more of that variation.

Real phones, different ways to scan

We test selected StoryCodes^® with both iPhones and Samsung phones. On each, we use the native camera as well as a QR code reader app.

This gives us more than one device and one scanning method against which to evaluate a design. We do not assume that success with a single phone tells the whole story.

In the photo below, a team member frames a food-ad StoryCode displayed on a desktop monitor. The salad is the central advertising visual — and also the element being tested for scanning.

Tap to enlarge

Changing the distance, not just the device

We display production candidates on computer screens and scan them from different distances. Moving further away changes how much of the phone’s camera view the StoryCode occupies — and whether its scannable structure remains readable.

This matters because advertising is rarely viewed at a fixed distance. A design intended for out-of-home advertising or a TV Pause Ad needs to be considered beyond a close-up scan at a desk.

“Our StoryCodes^® generator has over 100 parameters in addition to the prompts we use. Carefully calibrating these parameters helps us create scannable images. But being scannable does not automatically make an image suitable for advertising — and scanning successfully up close does not mean it will scan reliably from further away, or vice versa. Sometimes, we need to generate more than 50 images and scan-test 10 of them before finding one that is both visually attractive and reliably scannable under different conditions.”

Frederik Rademakers, Founder of StoryCodes^®

The second photo illustrates this wider-distance approach. A team member stands back from a large display showing a tropical-island StoryCode, rather than scanning directly in front of a desktop monitor. The tropical island at this size would not scan from a distance of 30cm, but scans perfectly when scanned from 1 to 5 metres.

Tap to enlarge

Sunlight in. Curtains closed.

We also vary the lighting conditions. Sometimes sunlight enters the office; at other times, we close the curtains and test in a darker environment.

The aim is to avoid selecting a design solely because it scans under one favourable setup. Changing the light gives us another practical way to challenge a candidate before it becomes part of campaign creative.

Distance and lighting tests complement one another. Together with multiple phones and scanning methods, they give our team a broader picture of how a design behaves.

Office testing prepares the design — not every possible placement

These tests are an important preparation step, but a monitor is not a printed billboard or a backlit bus shelter. Nor does one office display represent every television.

The final creative size, reproduction quality, display characteristics and intended scanning position still need to be considered for each execution. Where possible, the finished advertisement should also be checked on a representative print or screen before launch.

We therefore distinguish between testing a StoryCode candidate and validating its final placement. A successful office scan is evidence about that test — not a guarantee that the design will work everywhere.