An open observatory · Mission control
An open lab for radio signals in public telescope data. Explore what we measured, how we tested it, and what the evidence can actually tell us.
The working observatory · Archive processing
Real telescope recordings, processed in bounded passes. Every completed window adds an inspectable record. This is archival analysis, not a live telescope feed.
5 / 8 windows completed
3 awaiting a pass
0 blocked after three attempts
Last processing check: 2026-10-20 21:55 UTC. Scheduled every three hours; starts may be delayed.
The field guide · Volume 01
Follow a known signal, inspect a real archive search, and explore the limits of our detector. Every story opens its evidence.
Recorded telescope data
Find a known spacecraft in a real telescope recording. Start with an answer we can check.
Open investigation
Recorded telescope data
Trace 273 labelled rows through a real ON/OFF comparison, including the questions the run leaves open.
Open investigation
Synthetic control
Plant signals with known properties. Explore where the detector succeeds—and where it misses them.
Open investigationRecorded evidence · Observational totals exclude synthetic practice runs. This page is a snapshot; reload to see a newer build.
5 measured windows across 1 cadence(s). The report and original guided investigations retain their frozen evidence.
State built 2026-10-20 21:55 UTCEvery count opens its supporting run manifests.
Start with something we know. A recorded Voyager carrier tests whether the pipeline can recover a human transmitter. A positive control establishes a starting point, not a universal sensitivity limit.
EXP-001 · Known transmitter
The waterfall shows received power across frequency and time. Follow the narrow carrier, then inspect the exact measurement.
Frequency 8419.297028 MHz
Drift -0.378 Hz/s
SNR 1244.8
Open the run manifestThe method
A signal in an OFF scan fails the direction test. A signal confined to ON scans still needs further checks.
Method illustration · not a measured waterfall. Surviving this comparison does not establish a source or origin.
Observe Read a bounded frequency window from public recordings.
Compare Check drift, repeatability, OFF scans and known transmitters.
Reproduce Keep the inputs, parameters and result together in a run record.
Each point opens a recorded cadence. The map shows where we looked, not how much of the sky we have ruled out. Synthetic coordinates are excluded from the observational view.
A detector must earn its null results. We inject signals with known strengths and drifts, then measure how often they return. The grid shows a recorded calibration run.
Synthetic setigen injections only — not telescope observations. Recovery is computed from the recorded grid.
EXP-002 · 20261012T190848550861Z-exp-002-floor
At 0 Hz/s, full recovery begins at SNR ≥ 8; best nonzero-drift recovery was 100%. No tested SNR reached full recovery across ±3 Hz/s.
Each cell records the fraction of injected signals recovered; no cell is observational evidence.
Open the injection run manifestFollow the result
All recorded verdicts, including hits outside the preview gallery. Each bar links to the supporting run records.
One first-fitting verdict per hit. A zero means none in these records; it is not a limit on signals elsewhere.
Selected measured hits, newest first. Each classification describes the recorded checks; it does not identify a physical source. A survivor needs further investigation.
Also present in OFF scans
1700.000002 MHz · +0.010 Hz/s · SNR 592.8
The line is present in an OFF scan, so the ON/OFF filter does not pass.
Open the hit page Open run manifest
Also present in OFF scans
1700.999944 MHz · +0.316 Hz/s · SNR 75.3
The line is present in an OFF scan, so the ON/OFF filter does not pass.
Open the hit page Open run manifest
Also present in OFF scans
1700.032004 MHz · +0.010 Hz/s · SNR 21.0
The line is present in an OFF scan, so the ON/OFF filter does not pass.
Open the hit page Open run manifest
Known transmitter: GLONASS L1 FDMA
1600.004786 MHz · +0.010 Hz/s · SNR 26.7
The frequency falls inside GLONASS L1 FDMA.
Open the hit page Open run manifest
Also present in OFF scans
1500.000000 MHz · +0.010 Hz/s · SNR 331032.5
The line is present in an OFF scan, so the ON/OFF filter does not pass.
Open the hit page Open run manifest
Also present in OFF scans
1500.999928 MHz · +0.367 Hz/s · SNR 50480.4
The line is present in an OFF scan, so the ON/OFF filter does not pass.
Open the hit page Open run manifest
Also present in OFF scans
1500.999891 MHz · +0.500 Hz/s · SNR 26416.1
The line is present in an OFF scan, so the ON/OFF filter does not pass.
Open the hit page Open run manifest
Also present in OFF scans
1420.898438 MHz · +0.010 Hz/s · SNR 6255.5
The line is present in an OFF scan, so the ON/OFF filter does not pass.
Open the hit page Open run manifestTwo early research tracks use the same rule: a source, a method, and a record of what remains uncertain.
Archive lab · Early work
Project Blue Book document extraction. These runs measure extraction work; an incident map and corpus-level validation are still pending.
Run 20261012T231509202130Z-exp-010-count outcome ocr_pages_counted. Pages processed: 137.
Open the run manifestContext lab · Early work
Public fireball and geomagnetic records for a historical window. Coverage gaps remain; absence in a catalogue does not explain an event.
Run 20261012T224610003865Z-exp-020-baseline outcome baseline_measured. Fireballs recorded: 0. Kp max recorded: 4.0.
Open the run manifestAn invitation to investigate
Inspect a run, challenge a check, or help improve a source reader. Null results and documented failures belong here too.
Counts are computed from run manifests. A “survivor” has passed the checks recorded so far; it is not a conclusion. “Unexplained” is reserved for a reviewed record with the required sign-offs. Every displayed result links to its supporting run manifest.