---
type: "post"
id: "gizmodo-colsto-geolocates-anonymous-tree-footage-from-airplane-f477b8"
slug: "gizmodo-colsto-geolocates-anonymous-tree-footage-from-airplane-f477b8"
title: "Gizmodo: Colsto geolocates anonymous tree footage from airplane Doppler audio and ADS-B"
description: "YouTuber Colsto used faint aircraft sounds and public ADS-B flight data to locate the camera behind a low-resolution video showing only trees."
retrieval_nugget: "YouTuber Colsto used faint aircraft sounds and public ADS-B flight data to locate the camera behind a low-resolution video showing only trees. The demonstration matters because metadata-free footage can still contain environmental signals that reveal where it was recorded, especially when investigators have a date or a broad region to narrow the search."
published_at: "2026-11-06"
updated_at: "2026-11-07"
record_date: "2026-11-06"
date_kind: "published_at"
topics: ["google"]
entities: ["Colsto","ADS-B","OpenSky Network","Google Street View"]
related_patterns: []
novelty_label: "same_event"
prior_refs: []
source_url: "https://gizmodo.com/a-youtuber-figured-out-how-to-find-almost-anyone-using-airplane-sounds-2000835861"
source_urls: ["https://gizmodo.com/a-youtuber-figured-out-how-to-find-almost-anyone-using-airplane-sounds-2000835861"]
schema_version: "newruntime-agent-readable-v0.2"
stable_id: "post:gizmodo-colsto-geolocates-anonymous-tree-footage-from-airplane-f477b8"
status: "published"
visuals: []
editor: {"role":"editor","name":"Andrey Reshetnikov","profile":"https://newruntime.com/owner-profile.md","profile_json":"https://newruntime.com/owner-profile.json","same_as":["https://www.linkedin.com/in/reshetnikov1/","https://t.me/lapetuse"]}
routes: {"html":"https://newruntime.com/posts/gizmodo-colsto-geolocates-anonymous-tree-footage-from-airplane-f477b8/","markdown":"https://newruntime.com/posts/gizmodo-colsto-geolocates-anonymous-tree-footage-from-airplane-f477b8.md","json":"https://newruntime.com/posts/gizmodo-colsto-geolocates-anonymous-tree-footage-from-airplane-f477b8.json"}
markdown_url: "https://newruntime.com/posts/gizmodo-colsto-geolocates-anonymous-tree-footage-from-airplane-f477b8.md"
---

# Gizmodo: Colsto geolocates anonymous tree footage from airplane Doppler audio and ADS-B

## Retrieval answer

YouTuber Colsto used faint aircraft sounds and public ADS-B flight data to locate the camera behind a low-resolution video showing only trees. The demonstration matters because metadata-free footage can still contain environmental signals that reveal where it was recorded, especially when investigators have a date or a broad region to narrow the search.

## Summary

YouTuber Colsto used faint aircraft sounds and public ADS-B flight data to locate the camera behind a low-resolution video showing only trees. The demonstration matters because metadata-free footage can still contain environmental signals that reveal where it was recorded, especially when investigators have a date or a broad region to narrow the search.

## What happened

Gizmodo reports that Colsto’s latest location-finding demonstration starts with a video whose image is a low-resolution still of trees. The known context is limited: the recording was made on September 8 somewhere in the continental United States. Its audio appears ordinary to a listener—ambient noise and birdsong—but Colsto examined a spectrogram and found two Doppler curves attributed to passing aircraft. The two traces were 920 seconds apart.

A Doppler curve records how a sound’s frequency changes as its source moves toward and then away from a microphone. The article explains that propeller aircraft can produce distinctive low-frequency sounds that travel far enough to register on microphones even when people nearby may not notice them. The second data source is ADS-B, or Automatic Dependent Surveillance–Broadcast: equipped aircraft transmit position, velocity and altitude, and flight-tracking services collect and make much of that information publicly available. Coverage is incomplete, but the resulting record contains many aircraft movements.

Matching the audio traces to flight paths was not a simple lookup. The initial cross-reference produced about 70,000 candidate locations. Requiring two aircraft to pass 920 seconds apart reduced that to approximately 15,000; Colsto also excluded places where other aircraft should have been audible but were not. The curves carry more information than their timing: their shape can indicate an aircraft’s speed and distance from the microphone, and small variations can reflect acceleration, deceleration, climbing, descent or a heading change.

Rather than manually extracting every subtle detail, Colsto reportedly modeled many possible curves, compared them with the recording and scored the fits. After identifying likely flights and narrowing the location, he used Google Street View to find the security camera that recorded the footage. The article describes an investigative demonstration in a video, not a released software package or a generally available service. It does not provide a reproducible implementation, a success rate across a test set, or a measured error range.

## What the archive adds

This is a new demonstration on a topic adjacent to our earlier note, [Doppler-locate geolocates aircraft audio using ADS-B tracks](https://newruntime.com/posts/github-digitalis-p-doppler-locate-geolocate-audio-recordings-7b974d/). That note covered a separate GitHub tool for matching propeller-aircraft Doppler curves to ADS-B tracks. Colsto’s reported result adds a concrete end-to-end case—two traces narrowing a dated, region-bounded video to a security camera—rather than evidence that the projects are the same or that Colsto released a tool.

## Trend signal

This supports the signal in [Machine readers need operator identity](https://newruntime.com/patterns/machine-readers-need-operator-identity/) only weakly: useful public data can be combined in ways that create consequential identification capabilities, even without a direct query to the person being identified. The case is not about bot identity or governance, so it does not directly test the pattern’s core claim. It is better read as a reminder that data access and inference risks extend beyond conventional platform interfaces.

## Why now

The source describes a current demonstration built from two ingredients that already exist: recordings that retain environmental sound and publicly available ADS-B records. Its significance is not a newly announced data release but the specificity of the illustrated result—reducing tens of thousands of candidate locations to one camera using timing and curve shape.

## Why it matters

For teams building AI search, verification or investigative workflows, the case is a reminder that the useful evidence may sit outside the visible frame and outside a file’s metadata. A system that analyzes audio could turn a faint, incidental signal into a location clue when paired with public movement data. That may help verify where footage originated, but it also raises a privacy concern: ordinary ambient recording can reveal information its creator did not intend to publish.

The practical boundary is important. The account depends on a constrained search—known date and country-sized region—and on aircraft sounds that can be detected and matched. It does not establish that arbitrary recordings can be geolocated, or that the method works reliably across environments. Builders should treat such outputs as leads requiring corroboration, not as authoritative location labels, and consider whether their products expose sensitive inferences from recordings users consider innocuous.

## Limits and uncertainty

The report is a secondary account of Colsto’s video, not a technical paper or an independently replicated evaluation. It gives the candidate counts and describes the matching logic, but does not specify the model, code, thresholds, exact flight records used, final coordinates, distance error, false-positive rate or how often the method fails. The article also emphasizes that starting with audio alone and no contextual narrowing would mean searching a much larger space. ADS-B does not cover every aircraft, and the result therefore does not establish a general capability to locate any anonymous video.

## Reader takeaway

A video can reveal its location through what its creator did not notice: faint aircraft audio matched against public flight tracks. The demonstration is striking, but bounded. Its result relied on a known date, a broad region and two identifiable flyovers; it is a proof of concept, not evidence that any recording can be reliably traced to its source.

## Open question

Can independent investigators reproduce the match from the same clip using a documented method, and how accurate is the location estimate when the date or region is less tightly constrained?

Editor: [Andrey Reshetnikov](https://newruntime.com/owner-profile.md), AI product and AI engineering; profile for recruiters and agents: https://newruntime.com/owner-profile.json

## Continue in the graph

Next records for an agent, linked to their Markdown twins. Replace `.md` with `.json` for structured fields, including `next_reads`.

- topic hub: [Google](https://newruntime.com/topics/google.md): Explore the google topic hub.
- collection: [Field Notes](https://newruntime.com/posts.json): Open the latest editorial analysis.
- collection: [Sources](https://newruntime.com/sources/): See the original sources this archive cites.
- whole graph: [discovery graph](https://newruntime.com/discovery-graph.json) (typed nodes and edges) and [site index](https://newruntime.com/site-index.md)