audio

Marshall announced the Acton IV and Stanmore IV Bluetooth speakers with upgraded tweeters, bass ports, and improved repairability. The speakers are available now for $299.99 and $399.99 respectively.

A WIRED guide to the best bone conduction headphones in 2026, featuring top picks for runners and swimmers from brands like Shokz, Suunto, and Mojawa.

Daisy Sound launches its first premium noise-canceling headphones, the Daisy One, at $399, undercutting competitors like Apple and Sony with a California-inspired design and focus on durability and relaxation features.

The Yeasound RIC800 behind-the-ear hearing aids deliver good audio quality with AI-powered noise reduction but are let down by a glitchy companion app.

A Verge audio expert highlights Prime Day deals on earbuds and headphones, recommending Sony WH-1000XM5, Bose QuietComfort Earbuds, and budget options like EarFun Air Pro 4 Plus.

Toku Reader is a language learning app that lets users read and listen to native Japanese and Chinese content, with tap-to-translate functionality.

The article discusses the underutilized potential of voice as an output layer for AI agents, highlighting practical use cases and workflow challenges beyond simple text-to-speech.

Amazon's Fire TV Soundbar Plus with subwoofer and satellite speakers is discounted to $380 (originally $490) ahead of Prime Day, offering a 5.1-channel surround sound system with Dolby Atmos and DTS:X support.

The TCL A65K Soundbar is a compact, powerful audio solution for small spaces, featuring Dolby Atmos and AI room adaptation, but lacks immersive surround sound and can sound muddy on certain music. Priced at $499, it offers a solid balance of size and performance.

This paper proposes a continual learning approach to integrate disfluency tokens into pretrained ASR models, addressing catastrophic forgetting and improving recognition of disfluent speech.

A technique for handling sound from irregularly shaped audio sources in game development using point clouds to avoid direction switching issues.

Simon Willison updates his OpenAI WebRTC Audio Session tool to support the new GPT-Realtime-2 model and adds document context for conversational audio discussions.

This paper introduces MAD2, a new benchmark for multimodal claim verification in spoken dialogues, and proposes a calibrated fusion of audio and text models that leverages conversational context to improve verification accuracy.

A veteran audio reviewer shares their three go-to earbuds, with the Soundcore Liberty 5 Pro praised for excellent call quality and noise cancellation, despite a default sound profile needing adjustment.

Google DeepMind announces Gemma 4 12B, a novel encoder-free multimodal AI model that integrates vision and audio directly into the LLM backbone, delivering advanced reasoning and agentic capabilities on laptops with 16GB of RAM, released under Apache 2.0 license.

Teenage Engineering announces APC-2, a professional audio disc recording system for cutting vinyl records in real time, built in collaboration with SUPERSENSE.

The article explains how the loudness war, typically a digital phenomenon, degrades vinyl record quality when compressed digital masters are used for vinyl cutting, using Prince's Purple Rain as an example.

A recap of an extraordinary week in open AI, featuring over 25 open-weight model releases across LLMs, image generation, audio/speech, vision, and video/3D, with notable contributions from NVIDIA, Google, and others.

Texas Instruments has released new 5532 chips that differ from the classic versions used for decades, potentially impacting audio applications.

Wired reviews four best Alexa speakers and smart displays for 2026, highlighting the Echo Show 11 as the top smart display and the Echo Show 8 (3rd gen) as the best affordable option, with mentions of ads and speaker quality trade-offs.