The object was not a witnessed fly-by. It was discovered on playback.
On Sunday afternoon, 20 October 2002 — placed at about 4:20 p.m. EDT in later reconstructions — Mowry was recording aircraft and weather footage near Albany International Airport in Colonie, New York. During later review, station staff noticed a narrow, elongated object crossing the sky near an aircraft.
That distinction matters. Mowry did not report seeing a missile-shaped craft streak past him in real time. The case begins as a video anomaly embedded in ordinary airport footage.
Fox 23 reporter Dan Bazile contacted the National UFO Reporting Center the following morning. NUFORC's contemporaneous account says the station offered stills and sought outside help identifying the object.
The story developed over forty-eight hours.
“About 4:20 p.m.” is useful — but it is a reconstruction, not the cleanest first-generation timestamp.
Later reconstructions place Mowry's recording at approximately 4:20 p.m. EDT on 20 October 2002. The contemporaneous NUFORC database entry is less precise: it carries an approximate noon placeholder and explicitly notes uncertainty in the actual date/time.
That means the strongest public treatment is not to silently choose one clock time. The page should distinguish the reconstructed ~4:20 p.m. figure from the weaker database timestamp and preserve the fact that exact station metadata has not yet been recovered.
The tape was not demonstrably “seized.”
The most dramatic version of Albany says four FBI agents confiscated the original tape. The contemporaneous NUFORC record itself contains the correction.
Fox 23 news director Dave Brown told Peter Davenport that the station voluntarily cooperated with law enforcement and that he personally provided the FBI with video. Brown said he did not know whether the Bureau received the literal original or a copy.
NUFORC then contacted the Albany FBI office. Public-affairs spokeswoman Lisa Massaroni said, to her knowledge, the office had a copy and intended to analyse it, but she did not know the status.
No completed FBI technical finding has been recovered in the present public source chain.
Angular speed is visible. True speed is not.
What we can say
The elongated feature crosses several frames quickly relative to the camera field. It appears coherent enough that early analysts treated it as a real image component rather than an obvious edit.
What we cannot recover from web copies
Physical size, distance, altitude and true velocity require original frame timing, lens focal length/field of view, camera position, object range and a first-generation recording.
Do not convert “fast across the image” into a missile-like physical velocity. A small object close to the lens can traverse a large angle very quickly while moving at ordinary speed.
Useful evidence — but analysis of a derivative copy.
Brian Huff told NUFORC he had captured the footage from the Fox 23 website and subjected it to inversion, contrast, sharpening and edge-definition techniques. He concluded that the object was genuinely embedded in the image rather than a pasted or chroma-keyed overlay.
He also argued that the object was behind clouds and therefore distant, which would imply considerable size and speed.
The first conclusion is narrower and more defensible than the second. Image-processing may help detect crude compositing. But establishing that an unresolved feature passes “behind” cloud structure from a compressed web derivative is much harder, particularly where interlacing, motion blur, compression and low contrast are involved.
Range has to be demonstrated, not inferred from a processed web copy.
If Huff's cloud-occlusion reading is correct, it would be the most important technical fact in the case because it would place the object beyond a known atmospheric layer. But proving that requires source media and geometry.
1 · First-generation frames
Use the original camera tape/file or earliest station master. Web compression, deinterlacing and sharpening can create false edge relationships between a moving object and cloud texture.
2 · Exact frame timing
Recover native frame rate, shutter/exposure behaviour and interlace field order. Without these, elongation and motion across “seven frames” cannot be translated reliably into motion.
3 · Lens and sensor geometry
Identify focal length, zoom setting, sensor format and camera orientation. These determine field of view and angular displacement.
4 · Cloud-layer altitude
Recover local weather observations and cloud-base estimates for the exact recording window. A “behind cloud” claim has no range value unless the cloud layer itself is constrained.
5 · Demonstrated occlusion
Show frame-by-frame that cloud structure genuinely passes in front of the object rather than appearing to do so because of blur, compression or contrast changes.
6 · Only then calculate size/speed
If distance is established, angular size and frame-to-frame motion can be converted into physical dimensions and velocity with stated uncertainties.
The close-to-lens problem has to be tested before “missile speed.”
Bird / insect close to lens
A nearby fast-moving animal can become elongated by exposure time, interlacing and motion blur. This is especially relevant because there was no live visual sighting.
Aircraft at unusual aspect
A conventional aircraft can appear featureless in a few blurred frames. Matching one requires traffic records and geometry, not visual resemblance alone.
Airborne debris
Debris or material near the camera can cross rapidly through the field. The original tape would be needed to test focus, blur and depth cues.
Genuine distant object
This remains possible if first-generation footage establishes the feature is beyond cloud layers or otherwise distant. Current derivative clips do not settle that geometry.
No radar return is a constraint, not a verdict.
Contemporary summaries report that the object did not appear on Albany airport radar. If correct, that matters — but it cuts in more than one direction.
A small close-to-lens object would not be expected on airport radar. A distant target might also be missed depending on size, material, altitude, coverage and radar filtering. Conversely, a large solid missile-like craft crossing controlled airspace would make the absence of a documented radar track more important.
The useful question is not “radar saw nothing, therefore fake.” It is whether original ATC/radar records exist for the exact recording window and what those systems were capable of detecting.
The archive is not just a reference. It contains Albany-specific case material.
The University of Wyoming's Richard F. Haines collection explicitly lists multiple Albany holdings connected to the 20 October 2002 case. That makes it the strongest surviving route back toward first-generation evidence.
“Aircraft UC video – Albany, New York – October 20, 2002”
A dedicated Albany video item is listed in the archive. This is the first place to look for a higher-generation version than the compressed web copies in circulation.
“ALB UAP”
An Albany-specific UAP file grouping survives in the collection, indicating that the case generated its own analysis/correspondence trail rather than existing only as a web clip.
“NARCAP case 7: Albany, New York”
The event appears as a formal NARCAP case entry, which may preserve investigator notes, technical analysis, witness correspondence and source-media references.
“Brandon Mowry – Fox 23 WXXA – UFO video”
The named videographer and station are indexed directly in the archive, creating a clear path toward the operator/station evidence chain.
This is where the case should be rebuilt from.
The strongest research route is not another compressed web clip. It is the Richard F. Haines/NARCAP archival material, whose catalogue explicitly lists Albany-specific video, a formal NARCAP case file, an “ALB UAP” grouping and Brandon Mowry/Fox 23 material.