The Camera That Captured Nothing — Four Times in Seventeen Months, All on Clear Nights

October 11th, 2026 — 03:47

I need to document this while the details are fresh, even though there's nothing to show you.

Last night at 02:23, I observed three distinct flashes over the western tree line. Not lightning — the sky was clear, humidity at 41%, no storm systems within 200 miles according to NOAA. These were brief, intense bursts of light that illuminated the cloud layer from above. I've seen this before. The Flashes.

I was recording. My modified Sony A7S III with night-vision attachment was running, same setup that's captured meteor showers, satellite passes, even that military helicopter formation back in August. 256GB card with plenty of space. Fresh batteries. The record light was on.

When I checked the footage this morning, the file was corrupted. Not missing — corrupted. 47 minutes of static, timestamp intact, but the video data itself is unrecoverable. I've tried three different recovery programs. The file before it: perfect. The file after it: perfect.

This is the fourth time this has happened.

I went back through my logs. The previous three incidents:

• March 14th, 2025: The Returns over The Ridge, camera failure
• July 3rd, 2025: Sustained aerial light, 12-minute observation, corrupted file
• January 17th, 2026: (of course it was the 17th) Multiple flashes, equipment malfunction

All clear nights. All significant observations. All corrupted files.

Here's what I can't ignore: my camera has logged 1,247 hours of night sky footage over seventeen months. Hundreds of uneventful nights recorded perfectly. Planes, satellites, the occasional drone — all captured clearly. But four times — only four times — during observations that would have been significant, the files corrupt.

The alternative explanation:

Electromagnetic interference. If these flashes are atmospheric phenomena — say, upper-atmosphere electrical discharge or something related to ionospheric conditions — they could generate EM pulses strong enough to corrupt solid-state storage. It's scientifically plausible. My camera's shielding isn't military-grade. A strong enough pulse could absolutely scramble the data without destroying the file structure entirely.

I tested the camera this afternoon. Recorded for three hours. File is perfect.

But here's the thing: if the EM interference is strong enough to corrupt my camera's storage, why doesn't it affect any other electronics? My ham radio was running. My phone was recording audio. My laptop was logging atmospheric data. All of those files are fine. Only the night-vision camera fails — and only the video data, not the timestamp or file structure.

Seven years I've been doing this. Seven years of watching, recording, documenting. I still don't have proof of anything. Maybe that's the real pattern — I get close, and the evidence disappears. Or maybe I'm dealing with natural phenomena that I don't fully understand yet, and the camera failures are just coincidence.

Coincidence has a pattern if you look long enough.

I'm ordering a second camera with independent storage. If this happens again, I'll have redundancy. That's the only way to know if this is equipment failure or something else.

Another entry for the log.

Questions for the community:

Has anyone else experienced selective equipment failure — devices that work perfectly except during specific types of observations?

What kind of natural EM interference could corrupt video data while leaving audio, radio, and computer files completely intact?