Aircraft Sounds Can Pinpoint Where Videos Were Shot
A single creator's demonstration uses Doppler shifts and public flight-tracking data, with no validated error rate reported yet.
Updated on Nov. 3, 2026 in Geography

YouTuber colsto showed that faint aircraft sounds, inaudible during normal playback, can reveal where a video was recorded. Using flight-tracking data and the Doppler effect, the creator located a security camera in the continental United States to within a few hundred feet.
Why it matters
The technique turns ambient aircraft noise into a location fingerprint, a form of audio forensics that works on ordinary footage. It shows how publicly available flight data can be combined with signal analysis to geolocate recordings that otherwise carry no obvious place clues.
The analysis combined about 50GB of ADS-B flight data with terrain elevation data to compute three-dimensional distances between aircraft and candidate locations. The demonstration narrowed roughly 70,000 candidate points to approximately 15,000 and identified a top-scoring location that matched the original two-aircraft result, though the stub reports no error rate across other recordings.
The players
colsto
A YouTuber whose audio-forensics demonstration combined ADS-B flight tracking with spectrogram analysis to geolocate a security camera.
The details
The method relies on the Doppler effect, the shift in a sound's frequency as its source moves toward or away from a listener. A spectrogram, which plots frequency against time, revealed aircraft patterns in the recording even though the sounds could not be heard during normal playback. The creator used those patterns to predict where two passing aircraft would have sounded at each candidate location and compared the predictions with the original footage. Matching tree shapes and a utility pole in Google Street View then helped confirm the leading location.
Timeline
The recording was filmed on 8 September.
The Tech Race
The technique builds on publicly broadcast ADS-B aircraft positions, which are widely used for flight tracking, and applies them to a forensic problem few tracking services were designed for. It sits in the emerging field of audio and open-source geolocation, where analysts combine unrelated public datasets to recover context from media that carries no explicit location.
The method currently exists as a single creator's demonstration and is not a packaged tool, so no availability, pricing, or platform requirements have been announced. Its practical reach for ordinary viewers is unclear, though anyone with access to public flight-tracking data and basic signal-analysis software could in principle attempt a similar analysis.
The takeaway
Watch whether the method holds up when applied to other recordings and whether colsto publishes error rates or a reproducible workflow. Until then, treat the few-hundred-foot accuracy as a single-case result.
Further reading
For more on how location clues hide in everyday data, explore Geography coverage.





