# QFT Roadmap ## v0.4 physical validation 1. Use **450 B / single QR** as the baseline and measure real-phone distinct new-frame/s at Windows 30 / 45 / 60 FPS. 2. Record lock acquisition time, locked ROI area, camera FPS, successful decode/s, distinct new-frame/s and reacquire time per device. 3. Only after real-device data exists, test nearby DATA payloads (e.g. 390 / 450 / 510 B) and keep the value with the highest *effective* file throughput rather than the highest theoretical bytes/frame. 4. Add optional Windows FPS presets chosen from observed Android effective new-frame/s; the optical transport itself remains strictly one-way. 5. Tune focus/exposure behavior across LCD/OLED and 60/120 Hz panels. ## Later - Fountain/RaptorQ to reduce whole-playlist replay cost without a reverse ACK channel. - Native SIMD/custom visual decoder if ZXing remains the dominant bottleneck. - Optional custom QFT Visual Code after the single-QR baseline is measured. - iOS/macOS/Linux clients.