Back

YouTuber Used Airplane Sounds to Track Down an Almost Invisible Location

Fintech Digital Marketing

Blog

Nov 06, 2026

2 views

Technology

YouTuber Used Airplane Sounds to Track Down an Almost Invisible Location

A video does not need a recognizable skyline, a street sign, or a GPS tag to reveal where it was recorded. Sometimes, the most valuable clue is something almost nobody notices. A faint aircraft passing overhead. A subtle change in pitch. A sound buried beneath birdsong and ordinary environmental noise.

That is the premise behind a remarkable experiment by YouTuber Colsto, who demonstrated how aircraft sounds can help identify the location of an otherwise unremarkable recording. By combining audio analysis with publicly available flight-tracking information, he narrowed an enormous field of possible locations to the specific security camera that captured a video.

The experiment is an impressive demonstration of technical problem-solving. It is also a warning about digital privacy. In an increasingly connected world, information does not have to be visible to be useful. It only has to be detectable, measurable, and connected to another source of data.

An Ordinary Recording Becomes a Location Investigation

Colsto has built a reputation for finding locations that are difficult to identify, including places associated with album artwork and videos. His latest demonstration takes that investigative approach in a different direction. Instead of relying primarily on visible landmarks, he examines the sound accompanying a video.

The test begins with remarkably little information. The video contains a low-resolution still image of trees. There are no obvious buildings, road signs, or other distinctive visual features. The only initial clues are that the recording was made somewhere in the continental United States on September 8.

To a casual viewer, the audio sounds equally unremarkable. There is ambient outdoor noise and birdsong, but nothing that immediately identifies a city or neighborhood.

Colsto approaches the recording as a technical dataset rather than simply a piece of media. He examines a spectrogram, a visual representation of sound frequencies over time, searching for patterns that might correspond to aircraft passing overhead.

He finds two distinctive patterns associated with passing planes. The events occur 920 seconds apart, or 15 minutes and 20 seconds.

That interval becomes the first major clue in an investigation that demonstrates how apparently insignificant environmental details can acquire considerable value when combined with external information.

The Doppler Effect Provides the First Breakthrough

The central scientific principle behind the technique is the Doppler effect. Most people have experienced it when an emergency vehicle approaches with a high-pitched siren, passes nearby, and then moves away with a lower-pitched sound.

The perceived frequency changes because of the relative motion between the sound source and the observer. The same principle applies to aircraft.

As a plane approaches a microphone and then moves away, its sound creates a characteristic pattern in the recording. When that pattern is visualized on a spectrogram, it becomes possible to examine changes in frequency over time.

These patterns are known as Doppler curves. Their shapes can provide clues about the aircraft's motion, including its approximate speed and distance from the microphone. More subtle characteristics can also help distinguish one flight from another, including changes associated with acceleration, deceleration, climbing, descending, or changing direction.

The important distinction is that the recording does not need to contain a loud, clearly audible airplane. Aircraft can produce low-frequency sounds that travel considerable distances. A microphone may capture traces of those sounds even when a person watching the video would not consciously notice them.

That makes background audio a potentially valuable source of information. What sounds like meaningless noise to a human listener may contain patterns that specialized software can measure and compare.

Public Flight Data Turns Sound Into Location Intelligence

The audio alone is not enough to establish a location. Colsto needs another dataset against which he can compare the Doppler curves.

That is where Automatic Dependent Surveillance–Broadcast, commonly known as ADS-B, becomes important.

Aircraft equipped with ADS-B transmit information about their position, velocity, and altitude. Flight-tracking services collect this information and make much of it publicly accessible. The system does not cover every aircraft, but it provides a substantial record of aircraft movements.

Colsto uses that information to reconstruct which planes were flying at the relevant time and determine which flight paths could plausibly explain the sounds in the recording.

The method combines two different forms of evidence.

The audio provides measurements of the sounds captured at the unknown location. The flight data provides information about aircraft movements across a much larger geographic area. By matching the two, Colsto can eliminate locations that do not fit the observed patterns.

This is a form of data fusion. Neither dataset necessarily reveals the answer on its own, but their combination can be much more informative than either source separately.

The same broad principle appears in many areas of modern intelligence and investigation. A dataset that seems harmless in isolation can become revealing when cross-referenced with another source.

Seventy Thousand Candidates Become One Location

The scale of the search illustrates both the power and the difficulty of the technique.

Initially, matching the Doppler curves against aircraft data produces approximately 70,000 candidate locations. That is far too many to investigate manually.

Colsto reduces the possibilities by examining the timing of the two detected aircraft sounds. He looks for locations where two planes passed overhead approximately 920 seconds apart, while also excluding places where other aircraft should have produced detectable sounds that are absent from the recording.

These filters reduce the candidate pool to approximately 15,000 locations.

The remaining task is more demanding. Thousands of places may experience two aircraft passing at similar intervals, so timing alone cannot identify the correct one.

Colsto instead compares the detailed shapes of the observed Doppler curves with modeled curves generated from potential aircraft movements. Rather than attempting to extract every subtle flight characteristic directly from the noisy recording, he generates possible explanations and scores them according to how closely they match the actual audio.

This approach turns a difficult interpretation problem into a systematic comparison exercise.

Once he finds a strong match, he identifies the relevant flights and narrows the geographic search. A subsequent examination using Google Street View allows him to identify the specific security camera associated with the footage.

The progression is striking. A recording that initially offers almost no obvious visual clues becomes identifiable through two faint sounds, public flight records, and a process of computational elimination.

The figures also reveal an important limitation. The search becomes manageable because Colsto has a starting date, a country, and detectable aircraft sounds. Searching every possible location from an audio recording alone would be a far more difficult problem.

The Real Innovation Is the Combination of Data Sources

It would be easy to describe the demonstration as an unusual trick involving airplanes. That interpretation misses the larger significance.

The real innovation is the combination of information that was not necessarily created for the same purpose.

ADS-B exists to support aircraft surveillance and provide information about flight movements. Audio recordings capture the surrounding environment. Spectrograms make patterns in those recordings easier to analyze. Mapping tools then help connect a likely location to a physical place.

None of these elements needs to be secret. Their value emerges from the relationship between them.

This distinction matters because digital privacy is often understood too narrowly. People may focus on whether a photograph contains identifying information, whether a video includes GPS metadata, or whether a social platform can access their location.

Those are important considerations, but they do not cover every possible source of information.

A recording can disclose clues through its environment even after visible landmarks have been removed. Aircraft, traffic, public announcements, industrial equipment, and other sounds may all provide contextual evidence, depending on the recording and the surrounding data.

That does not mean every recording can be geolocated. It means that deleting obvious metadata does not necessarily eliminate every indirect route to identifying a location.

A New Challenge for Content Creators and Businesses

The implications extend beyond individual creators who want to keep their whereabouts private.

Journalists, activists, researchers, businesses, and organizations may all publish recordings whose locations are sensitive. A video intended to document an event, demonstrate a product, or show an ordinary working environment may unintentionally disclose more than its creator realizes.

For businesses, the risks can involve physical security and operational confidentiality. A recording made near a facility could contain environmental clues about where it was captured. A video from an otherwise anonymous workspace might reveal information through sounds or recurring background patterns.

For journalists and investigators, the same capabilities can have legitimate value. Audio-based geolocation could help verify the origin of footage, investigate disputed events, or assess whether a recording is consistent with a claimed location.

The technology is therefore not inherently beneficial or harmful. Its consequences depend on who uses it, what information is being investigated, and whether the investigation is justified.

The broader issue is that publicly available data can support forms of analysis that were difficult for ordinary users to perform in the past. Technical knowledge and computational tools can turn scattered clues into a coherent picture.

That creates opportunities for verification, but it also makes casual assumptions about anonymity less reliable.

Platform or Product. The Advantage Comes From the Workflow

Colsto's demonstration is not evidence of a single commercial product that can automatically locate every anonymous video. It is better understood as a workflow that combines existing information sources with custom analysis.

The components perform different jobs. Flight-tracking data provides candidate aircraft movements. Audio analysis extracts measurable features. Modeling compares possible explanations. Mapping tools connect the strongest candidates to physical locations.

The advantage comes from integrating those components effectively.

For technology companies, this is a familiar pattern. Competitive value often emerges not from owning every underlying dataset but from knowing how to connect data sources, filter irrelevant results, and turn a difficult problem into a repeatable process.

However, the original demonstration does not establish a commercial business model, pricing strategy, customer base, or revenue stream for this particular technique. It would be misleading to treat the experiment as proof that a mature location-tracking industry has emerged around aircraft audio.

What it does show is that existing public infrastructure can support sophisticated investigative capabilities without requiring access to a secret database or a specialized physical tracking device.

That lowers one barrier to experimentation, even though accurate implementation still requires technical expertise, suitable data, and enough contextual information to narrow the search.

The Limits of the Technique Matter as Much as Its Success

The demonstration is compelling, but its limitations should not be overlooked.

First, the method depends on suitable aircraft sounds being present in the recording. Not every environment will contain a detectable passing plane, and not every microphone will capture a useful signal.

Second, the flight data must provide enough relevant coverage to support meaningful comparisons. ADS-B does not track every aircraft, and incomplete records can make it harder to identify the correct flight.

Third, the search needs useful starting information. Colsto knew the recording came from the continental United States and had a specific date. Those constraints dramatically reduce the search space compared with a completely unknown recording.

Fourth, sound is affected by the environment. Distance, buildings, weather, microphone characteristics, and other background noise can influence what a recording captures. A modeled Doppler curve may resemble several candidates before more detailed evidence separates them.

Finally, a strong mathematical match is not the same thing as certainty in every case. The demonstration's successful result does not establish that the method will work consistently across arbitrary videos, different countries, or recordings with little contextual information.

These qualifications are essential. The experiment proves that the approach can work under favorable conditions, not that anyone can locate anyone else from any recording.

Privacy Protection Must Extend Beyond Metadata

The practical lesson is not that people should stop recording videos or assume every background sound exposes their location. It is that privacy requires more than removing GPS coordinates and checking visible details.

Creators handling sensitive material should consider the entire recording before publication. They can review whether the environment contains distinctive sounds, whether the recording date and broad geographic area are publicly known, and whether multiple pieces of contextual information could be combined to narrow down a location.

Where appropriate, they can remove or replace unnecessary audio, delay publication, or avoid sharing details that reveal when and where a recording was made. Organizations dealing with sensitive facilities or vulnerable individuals may also benefit from a review process that considers indirect clues rather than focusing exclusively on metadata.

No single precaution guarantees anonymity. The goal is to reduce unnecessary exposure and understand the kinds of information a recording may contain.

At the same time, the technique should be assessed in context. Verifying the location of publicly circulated footage can serve an important public interest. Using similar methods to identify private individuals without a legitimate reason raises different ethical and safety concerns.

The capability and the justification for using it are separate questions.

The Wider World Is Becoming More Observable

Colsto's experiment offers a revealing glimpse of a broader technological shift. The world is generating vast amounts of information through aircraft tracking systems, microphones, cameras, mapping services, and other connected infrastructure.

Much of this information was collected for practical purposes unrelated to identifying the source of an anonymous video. Yet, once combined, it can provide insights that the original data sources were never designed to deliver on their own.

For businesses, this creates both an opportunity and an obligation. Data integration can strengthen investigations, improve verification, and help organizations understand complex events. But it can also expose information that people assumed would remain private.

For ordinary users, the central lesson is more straightforward. A video is not just an image. It is a collection of signals, context, and potential clues. Some are obvious. Others become meaningful only when someone knows where to look.

The two airplane sounds in Colsto's demonstration were separated by 15 minutes and 20 seconds. Individually, they appeared insignificant. Combined with flight records and computational modeling, they helped reveal a specific recording location.

That is what makes the experiment both fascinating and unsettling. In the digital age, anonymity is not determined solely by what a person chooses to show. It can also depend on what their surroundings reveal, what public datasets preserve, and how effectively someone connects the two.