What calibration drift actually looks like

The mental model most people carry is that an instrument either works or it does not. In practice, the failure mode we see most often across a population of measurement microphones is a slow, monotonic sensitivity shift that stays comfortably within the instrument's tolerance for a long time, and then does not.

This matters because of what sits downstream. A field team calibrates against a sound calibrator before each survey, which corrects the overall level — so a drifting microphone sensitivity is masked at the calibration frequency and only shows up as a frequency response error elsewhere in the spectrum. The A-weighted number still looks plausible. The one-third-octave data quietly is not.

This is the argument for keeping calibration history rather than just the current certificate. One certificate tells you the instrument passed on the day. A run of certificates tells you which direction it is going, how fast, and whether this year's result is the start of a trend or a one-off. That history is the single most useful thing an annual calibration plan produces, and it is why we retain and issue it.