The station has moved. Last night, between the last arrivals and the first morning flights, every sensor left the old plastic box and settled into its new 3D-printed home. And the move brought more than tidier cables.
Before installing anything, we spread all the components on the kitchen table and tested the measuring chain one piece at a time. That is where the mystery of the past week was finally solved. Since 18 October, some aircraft peaks looked a few decibels too low, and we had suspected everything: the microphone position, the foam, even the sound level meter itself. The real culprit was much smaller: a faulty earth wire inside the new microphone cable. With the previous cable back in place, the numbers snapped right back.
Then came the moment of truth. With the station in its final position, we fitted our reference calibrator to the microphone. It plays two exact tones, 94 and 114 dB. The station read 94.0 and 113.8 dB. For the first time, the whole chain, from microphone to database, is verified at two points against a reference source.
The new box brought a surprise too. At night, the background level dropped from about 45 to 35 dB. The old box was not as quiet as we thought: the fans of the computer and the particle sensor were humming right next to the microphone. Now the microphone sits outside under a small rain hood, and the fans stay inside. Aircraft peaks are barely affected, but the quiet hours of the neighbourhood can finally be measured as they really are.
One small confession: the cooling fan was first installed backwards, blowing in instead of out. Of course, it was the hardest part to reach. It has been turned around this morning.
Two flights recorded while the sensors were on the kitchen table are flagged in the archive and left out of the statistics. Everything else is back to normal, only better.
Station Captain