Hardware & Devices

Sonos Tackles Muffled Dialogue with AI-Powered Speech Enhancement

Credit: sonos.com (edited)

Key Points

  • Sonos introduces an AI-powered Speech Enhancement feature for its Arc Ultra soundbar, improving dialogue clarity in movies and TV shows.
  • Developed in collaboration with the Royal National Institute for Deaf People, the feature uses 20,000 hours of custom audio for training.
  • This launch may prompt competitors to develop more advanced dialogue enhancement tools, addressing a widespread issue.

The common frustration of muffled dialogue in movies and TV shows often plagues the modern home entertainment experience. It's a problem made worse by cinematic audio mixes that prioritize immersive sound over vocal clarity, not to mention the steady rise of hearing loss, particularly among older adults.

Audio manufacturer Sonos has confronted this challenge with the debut of its AI-powered Speech Enhancement feature for its premium Sonos Arc Ultra soundbar. The system dissects sound in real-time, employing "source separation" to isolate and sharpen speech while preserving the overall dynamic range and Dolby Atmos immersion, a frequent casualty of older dialogue enhancement methods.

Smarter sound: Rolled out as a free software update, the new Sonos tool provides four distinct levels of dialogue boost—Low, Medium, High, and Max—designed for a range of listener preferences, from those wanting a subtle nudge to individuals with more pronounced hearing difficulties. This tiered system, especially the "Max" setting crafted for the hard of hearing, sets Sonos apart in a market where competitors like Samsung and Sony often provide simpler, on/off speech modes.

Expert collaboration: This new feature comes from a four-year development effort, heavily influenced by a year-long alliance with the Royal National Institute for Deaf People (RNID), the UK's top hearing loss charity, and input from award-winning film sound mixer Chris Jenkins.

"This is the first time that we've been embedded so early in the process," stated Lauren Ward, Lead RNID Researcher, highlighting the benefit of their early involvement. The AI learned its craft from 20,000 hours of custom-designed audio, not copyrighted films, to skillfully distinguish dialogue.

Designed for all: Sonos's move makes a statement, tackling a widespread issue with a refined, user-focused answer, and raising the bar for inclusive design in home audio. "Speech enhancement isn't just about those with hearing loss. It's about making sure that everybody can engage with the content they're watching on their terms," Matt Benatan, Principal Audio Researcher at Sonos, pointed out.

The feature's easy access in the Sonos app, not buried in an accessibility submenu, shows its design for all users and could bring back viewers who previously sidestepped dialogue-rich or action-filled shows.

Power demands: The machine learning algorithms that make this real-time speech separation possible require substantial processing muscle, confining the feature, for now, to the high-end Sonos Arc Ultra. The limitation shows a hurdle in bringing such advanced AI to a broader array of consumer gear. Still, its effective deployment on the Arc Ultra could clear the path for its integration into future Sonos devices as hardware improves.

Market ripples: The launch reflects a wider industry movement toward using AI for more responsive and personalized audio, potentially spurring rivals to create more sophisticated and adjustable dialogue tools. "One in three adults in the UK experience hearing loss, and it is reported that just under one in four adults in the USA do too," Lauren Ward, Lead RNID Researcher, mentioned, illustrating the large audience for such improvements.