Sound designers rebuilt the microwave beep on its own cheap buzzer — code only

The microwave beep has tortured kitchens for decades, and WIRED's Jeremy White confessed to sprinting across the room to silence his before it sounds. The reason it never improves is economics: nearly all microwaves use a passive piezoelectric buzzer, a membrane that vibrates when charged, first mass-marketed by Japanese manufacturers in the 1970s. It costs cents, and beyond frequency and duration it can do almost nothing — so in ferociously competitive white goods, no manufacturer swaps it for a speaker.

The reason the beep never improves is economics: nearly all microwaves use a passive piezoelectric buzzer — a membrane that vibrates when charged — first mass-marketed by Japanese manufacturers in the 1970s. It costs cents and can do little beyond frequency and duration, so in ferociously competitive white goods no manufacturer swaps it for a speaker.

The deepest trick is faked dynamics: sweeping through the frequencies where the buzzer naturally resonates makes it sound louder and softer without any volume change. To soften menu navigation, Starling recreated the soft premium click of a smartphone keyboard as a 4,000-Hz beep lasting one millisecond. 'Oh, that actually sounds pretty good,' Corelitz recalls thinking. 'It would make a microwave infinitely more satisfying to use.'

The demand side may already be there. Coogan said the recurring Instagram comment was 'Oh my God, I would pay $50 more for a microwave that sounded better.' A manufacturer would only need to accept new code on its existing chip.

Keep the buzzer, change the code: trills, sweeps and arpeggios on the same piezo chip — even fake volume, by sweeping through the buzzer's resonant frequency.

Cost competition kept the cents-cheap piezo buzzer in place, freezing microwave sound in the 1970s.

The buzzer's limits — frequency and duration only — defined a small enough design space to attack completely in software.

Resonance sweeping turns the buzzer's physical preference into an expressive resource, faking dynamics the hardware can't produce.

Consumer response on Instagram signaled willingness to pay, framing better sound as margin, not charity.

Treat the cheapest component as a programmable instrument: the constraint that locked the ugly beep in place is also the path around it, at zero hardware cost.

As of the July 2026 WIRED piece, the sounds existed on Starling's test rig and in its Microwave Sound Bench software; shipping them requires a manufacturer to load new code onto the existing microcontroller, and none had been announced. Coogan reported readers repeatedly saying they would pay $50 more for a better-sounding microwave.