Reading Speed, Image Quality Ratings, and Comfort Ratings in Augmented Reality
Over 11,000 reading speeds and 6,000 comfort ratings reveal how AR headsets handle text.
For tech professionals, text readability in augmented reality (AR) remains a critical but under-benchmarked metric. A new paper from Kim, Ghahghaei Nezamabadi, Lian, and Singh introduces Read-AR, a comprehensive dataset of over 11,000 reading speed measurements and nearly 6,000 subjective quality and comfort ratings. All data was collected on the same experimental setup across more than 80 distinct conditions—varying font size, contrast, rendering method, and display parameters—eliminating the typical confounds of cross-study comparisons. The dataset is available on arXiv (ID: 2604.27203) and is intended to serve as a reference for evaluating different AR headset architectures, making it possible to directly compare how each design affects real-world text legibility and eye strain.
The implications extend beyond academia. As AR moves into productivity, navigation, and remote assistance, text is the primary interface. Read-AR provides a rigorous, repeatable methodology for manufacturers to test headset prototypes against a baseline, potentially accelerating design iterations for better contrast, resolution, and ergonomic comfort. The authors emphasize that the consistent experimental framework allows for apples-to-apples comparisons—a major step forward from isolated studies. With the dataset now public, developers and hardware teams can benchmark their own systems against these 80+ conditions, reducing guesswork in display optimization and improving user experience across current and next-generation AR devices.
- Read-AR dataset contains 11,031 reading speeds and 5,966 quality/comfort ratings across 80+ controlled AR conditions.
- All data was collected on the same experimental setup, enabling direct cross-architecture benchmarking for headset designs.
- The dataset is publicly available on arXiv (2604.27203) and targets HCI researchers and AR hardware engineers.
Why It Matters
First standardized AR text readability benchmark enables data-driven display optimization for next-gen headsets.