Playing a cleaning tone from a video platform can work, but several platform behaviours reduce its effectiveness — and they are easy to check.
For a complete overview, see clearing trapped water properly.
Re-encoding alters the waveform
Video platforms re-encode uploaded audio with lossy compression optimised for how music and speech sound to
human ears, not for preserving a mathematically clean sine wave. The result can include harmonic artefacts and
altered peaks — precisely the waveform precision the mechanism depends on.
Loudness normalisation
Most platforms normalise playback loudness across content. A pure low-frequency tone measures differently from
music, so normalisation may quietly reduce its level. You set 80 percent and receive noticeably less at the
driver, with nothing to indicate it.
Autoplay is a real problem
A cycle interrupted halfway by the platform moving to the next video wastes the attempt, and it is easy not to
notice while holding the phone grille-down. Disable autoplay before starting.
Playback quality settings
Lower quality settings compress audio harder. Set quality to the highest available, since audio bitrate scales
with video quality on most platforms.
Titles are frequently wrong
There is no verification of frequency claims, and speaker cleaning is a high-traffic category that attracts
low-effort uploads. Videos claiming 30,000 Hz exist in quantity, and that frequency cannot work.
The check that matters
Turn Bluetooth off entirely first — video apps route to paired devices readily, and this catches more people
here than anywhere else. Then confirm you can hear an obvious low hum coming from the device itself. If you
cannot, nothing useful is happening.
The alternative
A browser-generated tone avoids compression, normalisation, autoplay and mislabelling, and works offline once
loaded.