New Chip Makes Digital Synths Sound Cleaner Than Ever

If you've ever heard an electronic keyboard or a synth app that sounded slightly "digital" or harsh, the culprit might be a glitch called aliasing. When computers recreate analog sounds, they sample them as discrete numbers. But fast-moving waveforms can outpace the sampling, creating unwanted artifacts—like jagged edges on a photo. This new research paper from engineers at ETH Zürich tackles exactly that problem.

The solution: instead of brute-forcing the math, they use additive synthesis—building waveforms from pure sine wave layers, like mixing paint from primary colors. That lets them apply a clever transformation that stays clean even with complex sounds. They also built a dedicated chip, called HASY, that runs this calculation in real time. At just 6 square millimeters, it's tinier than a fingernail and fits easily inside portable instruments.

Why does this matter to you? It means future synthesizers and music production gear could deliver the warmth of classic analog synths without the bulk, cost, or digital fuzz. The chip supports up to 512 harmonics—enough to recreate rich, detailed tones—and processes a sample in just five audio-clock periods. That's fast enough for live performances or recording sessions without noticeable lag.

There's a catch: the chip is a prototype, so you won't hear it in store-bought gear tomorrow. And musical taste is subjective—some people genuinely like the grit of digital artifacts. But for anyone who wants pristine, lifelike synth sounds, this research points to a future where the technology quietly disappears, leaving only the music.