The Desk That Knows When the Sky Is Busy
Something chimed on my desk at 14:23 and I looked up from the wrong screen.
My hobby is collecting hobbies, and hobby number thirty-eight is Orbital Decay Audio Notifications — a microcontroller that tracks satellite passes and announces each one with a short generative phrase. The ISS had just cleared my northern horizon, climbing toward a 72-degree pass, and I’d missed the first note because I was debugging the code that was supposed to play the first note.
This is the thirty-eighth time I’ve done this to myself.
Punch Cards and Countdowns
The tracking data comes from TLE files — Two-Line Element sets — and they carry history I didn’t expect. NORAD developed the format in the early 1970s to fit 80-column punch cards, and for years, getting them electronically meant waiting for T.S. Kelso to manually transcribe NASA’s printed bulletins into his BBS. While doing that transcription work, he discovered a checksum bug caused by UNIVAC’s switch from BCD to EBCDIC encoding. The format survived. The bug got fixed. The 80-column constraint remains, decades after anyone used a punch card for anything.
My ESP32 pulls fresh TLEs from CelesTrak every morning, runs SGP4 propagation to compute the next passes, and waits. When a satellite clears the horizon — AOS, Acquisition of Signal — the synth wakes up.
Elevation sets the register. A high pass gets bright upper partials; a grazing 15-degree flyover stays in the bass. Doppler shift maps to vibrato, and here’s where I got the physics wrong for most of the morning: I assumed maximum elevation meant maximum signal distortion. It’s the opposite. Doppler shift is worst at the horizon, when the satellite’s relative velocity toward or away from you is highest. By the time it’s overhead, it’s moving sideways, not toward you, and the frequency stabilizes. So the vibrato peaks at AOS and LOS, settling in the middle.
That curve — fast shimmer, steady centre, fast shimmer — gives each pass a shape I can hear without watching the countdown.
The Invisible Traffic
By late afternoon I’d heard four ISS passes. Two of them happened during daylight, when the station was invisible against the blue, and one came through clouds so thick I couldn’t have seen Sirius. The desk chimed anyway. Turns out the sky is busier than I thought, and most of the traffic never announces itself unless you’re listening on the right frequency.
This connects to yesterday’s meteor-scatter triggers, where the sky clicked back at my radio whenever ionized debris crossed the beam. There, the atmosphere decided when to open the shutter. Here, orbital mechanics decides when to play the chime. Both share the same instinct: let invisible phenomena surface through a different sense.
Around 17:40, something in the amateur satellite group played a phrase I didn’t recognize. Lower register, longer duration, more vibrato at the edges. I checked the pass list — AO-91, an FM repeater satellite I’d never consciously tracked. Its orbit is lower than the ISS, decaying faster, and the prediction software flagged it yellow for “stale TLE.” The phrase sounded hurried. I don’t know if that’s real or projection, but I wrote down the timestamp anyway.
What the Ear Learns
The satisfying thing about sonification — and I keep returning to this thread, from METAR chords to stratospheric balloon logs — is that ears learn patterns faster than eyes. After three ISS passes I stopped checking the screen. The rising tone, the brief steady middle, the fade at LOS: that’s the station. Magnitude −5.9, brighter than Venus, carrying six people at 7.67 km/s, and it sounds like a wind chime that knows when to quit.
Operation Moonwatch recruited amateur astronomers with binoculars and stopwatches in 1956, modelled after WWII bomber-watching networks. They were the original satellite spotters. I’m doing the same thing, except the spotting is auditory and the bomber is a research station that reboosts itself every few months to stay ahead of atmospheric drag.
The desk is quiet now. Next pass is in forty-three minutes — a NOAA weather satellite, low elevation, probably bass-heavy with shaky edges. I’ll know by the sound whether the prediction was right.