Experimental color matrix code
QuadQR.
A different way to encode information. Four colors, a denser data layer, and optional encrypted payloads in a QR-inspired format.
What if a code could carry more?
I built QuadQR to explore a question: could modern cameras read more information from the same area using color? The result is an independent matrix-code format with its own encoder, renderer, image and camera scanner, and browser demo.
- RGBW encoding and optional experimental Triangle16 high-density mode
- Reed–Solomon error correction and perspective-aware scanning
- Optional AES-256-GCM encryption and Ed25519 signing
- Compression, benchmarks, and image reliability tools
QuadQR is an experimental custom format. It requires a QuadQR-aware scanner; standard QR readers cannot decode it.
Color carries information
The colors encode data: four RGBW states carry two raw bits per cell. Experimental Triangle16 splits body cells into two colored triangles for four raw bits. That mode is off by default and needs real-world testing.
Recovery informed by the scanner
Spectrum ECC uses color confidence to guide Reed–Solomon recovery and try bounded alternative readings. Spatial interleaving spreads neighboring data across the symbol to reduce concentrated damage. Recovered candidates still face error-correction and CRC checks.
Protect the payload
Optional AES-256-GCM encryption protects contents using a password or application key. Ed25519 signatures let applications verify payloads against a trusted public key. Encryption and signing can be combined; secret keys stay outside the symbol.
Measure the idea
The browser demo brings generation, camera scanning, compression, and image stress tests together. Capacity benchmarks compare equal matrix dimensions, but matching ECC labels do not establish equal damage tolerance. Raw density is only part of usable capacity.