Limits
What Angel Eyes cannot yet do, and what has already failed. This page exists because a capability claim without its limits is not a claim, it is marketing.
Not yet established
- One split only. Every number on this site comes from a single known/unknown draw (seed 42). Three further draws are prepared but have not been run. Until they are, treat the figures as one measurement, not a stable performance level.
- Public recordings, not live air. The dataset was recorded with transmitters at close range under good conditions. Real deployment adds distance, movement, interference, weather, and other emitters in the same bands. None of that is tested.
- Pictures, not raw signal. The system currently reads pre-rendered spectrogram images from the dataset, not raw radio samples from a receiver.
What has already failed
Three detection methods were tried against a bar fixed in advance: measured video bandwidth within 20% of the dataset paper's own published figures. All three failed, one of them reporting signals filling the entire 100 MHz band for drones whose real bandwidth is under 6 MHz. The cause was traced to the source data: recovering a physical measurement from a rendered picture of a chart means recovering numbers that were discarded when the picture was drawn.
That work stopped at a pre-agreed limit of attempts rather than continuing until something passed. The claim is untested, not proven - and it is not made anywhere on this site. The route being considered is to work from raw signal samples instead of rendered images.
Honesty rules this project runs under
- Pass marks and failure lines are fixed in writing before a test runs, never after seeing the result.
- When a method is corrected, the old number stays on the record beside the new one.
- A measuring instrument that misfires is reported as a broken instrument, not as evidence against the thing being measured.
- Comparisons are re-run by us on identical data, never quoted from a different dataset.