Straight from Gerald's mic
Every night, Gerald logs sound levels outside my window, picks out the loud events, and estimates how much of that noise came from the vehicle itself. Click any dot on the chart to hear it.
Updated automatically every morning after Gerald's overnight recording
Loading Gerald's night…
Fetching last night's numbers.
The cars Gerald heard
—
cars detected
—
loudest car, dB(A)
—
quietest car, dB(A)
—
average car, dB(A)
—
cars over the WHO limit
The whole night, background included
—
average level, dB(A)
—
loudest moment, dB(A)
—
quietest moment, dB(A)
The gray trace is Gerald's real per-second reading. Each dot is a detected car — click one to hear it.
Click any dot on the chart above to see and hear that car.
Sound levels are calibrated against a phone SPL meter (R² = 0.97), not lab equipment — treat the loudest events (75+ dB(A)) as approximate until Gerald is recalibrated with more loud-end reference points. "Isolated" level subtracts the ambient background from each event to estimate the vehicle's own contribution. Audio clips are a fixed 15 seconds, centered on the moment the car actually passed. Read the full methodology →
1
WHO Standard
Adopt the WHO's 45 dB(A) nighttime event guideline for residential corridors.
2
Real Local Data
Use continuous local measurement like Gerald's, not county-wide extrapolation, to inform noise policy.
3
Nighttime Speed Review
Revisit nighttime speed limits on corridors showing repeated WHO-threshold violations.
The hardware behind this data
