A military infrared video that appeared to show six small objects moving at extraordinary speed now has a testable and potentially ordinary explanation. Researchers analysing the released footage found that the objects repeatedly brightened and dimmed at frequencies resembling the wingbeats of flying animals. The result does not prove every unidentified object is a bird. It demonstrates how measurable details in a contested recording can challenge impressions created by dramatic speed estimates.
The analysis was released as an October 6, 2026 research preprint by Jacob Haqq-Misra, Ravi Kopparapu, Richard Cloete, Omer Eldadi, Michael Shermer and Abraham Loeb. It examines an infrared recording released by the United States Department of War, accompanied by a military report that described six compact objects and cited an estimated ground-projected speed of about 480 miles per hour.
Rather than focus first on what the objects looked like, the researchers measured their brightness across the video frame by frame. All six sources showed periodic changes, each cycling at roughly 7.1 to 7.9 times per second. The pulses did not all occur together, which led the team to investigate whether they might come from separately flapping wings rather than a single camera or encoding artifact.
A crucial part of the study was its attempt to rule out simple technical explanations. The researchers inserted artificial constant-brightness objects into comparable footage and checked whether compression or sensor effects could generate a similar rhythm. They also compared the military recording with infrared videos of flying animals collected by the Galileo Project near Las Vegas.
The comparison footage showed animal wingbeat-related rhythms over a broad range that overlapped the frequencies measured in the military video. Brightness changes accompanied changes in the apparent area of the flying objects, another feature compatible with the varying silhouette of moving wings.
The claimed speed also deserves scrutiny. A sensor observing a distant object from a moving aircraft needs information about range to recover the object’s true position and velocity. Without distance, a speed derived from where the camera’s line of sight intersects the ground can be very different from the movement of an object closer to the aircraft.
The authors argue that, under plausible geometric conditions, objects much nearer to the platform could account for the reported large ground-projected speed without traveling at anything like that speed themselves. This is an example of the familiar parallax problem: the motion of the observer and the distance to the target can change the apparent movement dramatically.
The bird hypothesis is not without limits. Flicker at a wingbeat-like frequency is evidence for a particular kind of repeated physical movement, not a species identification. Other rotating or oscillating objects could produce periodic signatures, and the released recording does not include all the range and platform data that would settle the geometry.
Some details, including the circumstances of the flight and the particular wingbeat frequencies of local bird species, require further evaluation. The paper is also a preprint, which means readers should distinguish the authors’ findings from an independently established scientific consensus or an official reclassification by the Department of War.
An important strength of the work is that the researchers have made a reproduction package available with video sources, analysis scripts, comparison material and derived measurements. Those resources allow critics to test the calculation and make disagreement more concrete. A reproducible argument is not automatically correct, but it is more informative than an explanation that depends entirely on intuition.
The cultural fascination with unexplained aerial phenomena often begins with uncertainty and quickly moves toward extraordinary conclusions. That leap is not required. An object can be unidentified because a recording lacks information; unidentified does not mean extraterrestrial, and a plausible natural explanation does not dismiss the possibility that other cases require different answers.
The broader lesson concerns evidence, perception and intellectual humility. People are excellent at recognizing familiar stories in ambiguous images, whether those stories concern secret technology or visitors from elsewhere. The disciplined response is to ask what the recording measures, what it omits and which competing hypotheses remain testable.
For now, the six objects provide a case study in how a mystery can become more tractable when the question changes from ‘What does this look like?’ to ‘What physical signal can be measured?’ The bird interpretation is a scientifically grounded candidate, while the ultimate identification remains contingent on evidence that has not yet been released.
References
https://arxiv.org/abs/2610.08692 — Haqq-Misra and colleagues, October 6, 2026, research preprint.
https://zenodo.org/records/23021686 — Primary data and code repository for the analysis, October 6, 2026.