We build sound based speaker cleaners so you never need a pin

We work with audio: recording, measurement, the unglamorous business of working out why one speaker sounds different from another. Somewhere along the way we noticed a pattern in the questions friends and family brought us. A phone gets rained on, a laptop lives in a dusty bag, a pair of earbuds spends a summer in a gym pocket, and the sound goes dull. The advice they had found online was almost always worse than the problem: a safety pin, a cotton bud soaked in alcohol, a can of compressed air pressed flat against the mesh, a bowl of rice.

Every one of those can turn a blocked grille into a torn membrane. Meanwhile the speaker itself is a device built to vibrate, and vibration is exactly what moves a water droplet or a plug of lint back out the way it came. Smartwatches have shipped a water eject tone for years. There was no reason that had to be locked inside one company's hardware.

So we built it as a web page instead. Then people started asking for versions tuned to their specific device, because a MacBook driver and an AirPod driver are not remotely the same thing, and the library grew from there. Today SonicSweep is a set of cleaners for specific devices plus a blog of guides we actually want our own families to read.

Picking a cleaning tone is not guesswork, but it is not one number either. A phone bottom speaker, a laptop woofer, a tablet array and an earbud driver differ in size by an order of magnitude, and the frequency that produces the largest useful excursion moves with them. Big drivers respond lower. Tiny drivers need a higher band and far less level before they start distorting rather than moving air.

• Start from the driver. We set each page's default frequency near the band where that class of speaker moves most air without audible strain.
• Listen on hardware, not just a scope. Every tone gets checked on a spread of real devices we have on hand, across current and several year old models.
• Watch for distortion, not just loudness. If a tone makes a small speaker rattle or clip instead of pumping cleanly, it is the wrong tone, however impressive it sounds.
• Keep cycles short. Cycle lengths are set so a full run is long enough to work and short enough that nobody is tempted to leave a speaker screaming for ten minutes.
• Re-check when something changes. New browser audio behaviour or a new device layout means the affected pages get another pass.

We are deliberately not going to claim a laboratory, a standards body or a certification we do not have. This is careful, repeatable listening on a range of devices, plus the physics of how a diaphragm moves. It is enough to choose a good frequency, and we would rather describe it accurately than inflate it. You can hear the result for yourself on the speaker sound test before and after a cycle.

The blog exists because a cleaning tone is only ever part of the answer. A crackle at high volume, one channel quieter than the other, sound that cuts out when the case flexes: these are different faults with different fixes, and some of them are not fixes you can do at home.

So we hold every guide to the same test before it goes live.

1. It names the safe method first, and names the unsafe shortcuts explicitly so you can recognise the bad advice elsewhere.
2. It gives you a way to tell whether the problem is a blockage or hardware, usually in under two minutes.
3. It states plainly where the do it yourself route ends and a repair shop begins, including the rough cost of being wrong.
4. It separates what we have verified from what is general guidance, and it does not pad either with filler.
5. It gets corrected when a reader shows us we were wrong. Corrections are edits to the guide, not quiet deletions.

If you spot something inaccurate, we genuinely want to hear about it. The fastest route is the contact page, and a specific note beats a general complaint every time.