A report from Beaumont-sur-Sarthe on 22 March 2025 contains two stages: a light noticed while driving and a later object visible in a phone recording. GEIPAN identified the recorded feature as an optical reflection, while explicitly stating that the earlier visual observation lacked enough detail to identify. UKPX keeps both stages in one case and assigns Insufficient evidence to the whole account, preserving the narrower explanation of the video.
The drive and the later recording
In the anonymised questionnaire, the witness describes a journey between Saint-Marceau and Beaumont-sur-Sarthe during the early hours of 22 March. A glow seemed like a small moving patch of mist. The observer chose to concentrate on driving and planned to check again after arriving home, several minutes later. The form places the road observation around 02:07 and the recording at about 02:14 local time.
The filming took place outdoors near the witness’s home. The written account asks whether an object visible around eleven seconds into the recording might be a drone. That question is an interpretation, not an identification. The form gives no other witness and concentrates its later descriptive answers on what appeared in the video. It does not provide a comparably detailed record of the first glow. A report of seeing something again is therefore not enough to demonstrate that both stages had the same cause.
What could be analysed
The three-page photographic annex records a significant limitation in the submitted file: re-encoding had altered its metadata and removed usable capture-time information. A watermark also showed editing. The investigator requested the original on 9 September 2025 but received no reply. The phone make and model could not be recovered from the available metadata.
GEIPAN nevertheless continued with a geometric interpretation of the visible feature. Its IPACO analysis considered a strong streetlamp within the frame and a reflection about the optical centre. An initial mismatch with the expected position was attributed to imperfect alignment between optics and sensor; the overlay was then adjusted interactively. The annex concludes that the recorded feature is a reflection. UKPX reports that conclusion and its method without claiming to have independently calibrated the camera or repeated the analysis on an original file.
Why two experiences cannot be silently merged
The official case page distinguishes the mist-like visual report from the later drone-like appearance on the recording. It says that the first observation could not be identified because the necessary information was missing. Its category A classification is expressly limited to the video. That limitation is part of the published finding and must travel with the explanation whenever the case is summarised.
The distinction works in both directions. Identifying a reflection in a recording does not establish that the witness’s earlier naked-eye experience was itself a camera effect. Equally, an unidentified earlier impression does not transform the later recorded reflection into an unknown aerial craft. The evidential status of each stage depends on what was actually recorded about it, rather than on the assumption that a continuous personal experience must have one continuous external source.
The witness associated the two stages, and GEIPAN assessed that association while limiting its identification to the recording. Reading the observations together preserves that relationship and makes the scope of the investigation clear.
File history and interpretation
Re-encoding limits the claims that can be made about a file’s original capture circumstances. It does not automatically show that someone fabricated an image. The relevant source question is what information survived and whether the proposed analysis depends on information that was lost. Here the annex argues that the visible optical relationship could still be examined, while acknowledging that formal authentication and camera identification were not possible from the supplied file.
The interactive adjustment is also worth reporting plainly. A geometric overlay is an analytical reconstruction, not an independently recorded flight path. Readers can assess the strength of GEIPAN’s explanation more fairly when both the proposed optical relationship and the missing original are stated. Suppressing either fact would make the result appear simpler than the surviving evidence permits.
Whole-case outcome
UKPX’s Insufficient evidence outcome was checked on 13 November 2026 through primary-source review. The video has an attributed conventional explanation; the earlier observation has no sufficiently detailed surviving description for identification. This is not a strong unexplained case, and the official video finding is not extended to a different, unrecorded visual impression.
A meaningful reassessment would require material that addresses a specific gap: the original recording and reliable capture information for the filmed stage, or a fuller contemporaneous account of the road observation. A later retelling that simply repeats the association between the two stages would not supply those missing measurements.
Location and photograph
The pin is an approximate Beaumont-sur-Sarthe locality reference, rounded from the checked town coordinates on France-Voyage. It anchors the reported home-area filming without publishing the private residential parking location. It is not a reconstructed position for the earlier glow or the object in the recording, and the case is not offered as a visitor attraction.
The lead photograph shows Beaumont-sur-Sarthe from the D338 on 25 August 2021. It provides locality context only: it does not show the witness’s parking place and is not evidence of either stage of the 2025 report.
Sources
GEIPAN case 2025-03-51657: classification and explicit scope limit. Photographic annex: file limitations and optical analysis. Witness questionnaire: first-person chronology. Beaumont-sur-Sarthe geographical reference: representative locality coordinates. The source video and analytical illustrations are linked, not reproduced.