StepAudio 3 is a five-model audio family whose most interesting advances concern continuity: conversations that keep moving, scenes assembled on one timeline, and songs planned before they are rendere...

A concise synthesis of recent speech AI releases from OpenAI, Meta, Gradium, Zendesk, and Google Workspace, tracing the industry's shift toward full-duplex audio, ambient acoustic perception, and refe...

Designed by Central Saint Martins graduates Domenico Di Paolo and Kieran Feechan, IMAGO is a tactile audio player that challenges extractive generative audio by running lightweight, artist-consented n...

Drone & Drone is Siedl/Cao's ongoing kinetic installation for 16 pairs of Vietnamese wind-driven flutes on rotating arms, built on the military and musical senses of drone. It is on view at Klangspure...

The Vinyl Factory has repressed two Ryoji Ikeda works that use locked grooves to translate planetary and communication systems into continuous sound.

An IEEE Spectrum interview with Stephen Roddy frames sonification as a perceptual language for complex data, joining computational analysis with human listening.

Microduck uses a seedable synthesizer and each unit's chip serial to create a persistent voice, then carries that identity through calls, gestures, instruments, and group performance.

Stability AI's beta plugin and web app move Stable Audio 3.0 from one-shot generation toward arrangement, section-level revision, mixing, and export.

Sonic Field has organized 631 public STEIM recordings into a searchable research collection, opening its archive relations and beta People profiles to corrections and community contributions.

ElevenLabs, Cartesia, Fish Audio, Pipecat, and BreezeBlue are turning AI speech into programmable infrastructure for realtime agents.

Matic Cues joins local acoustic sensing, gesture recognition, a 3D map, and cloud language parsing so that a spoken request can become a situated physical action.

After the earthquake in western Colombia, sound workers are joining organised rescue efforts to detect faint signs of life while communities help by producing the silence the search requires.

STEIM's 631-video YouTube collection preserves a culture of live electronics while showing why instruments, performances, and their relations need more than recordings to remain legible.

Las cinco ponencias del seminario Políticas de la Escucha 2026 ya pueden verse en línea: dos jornadas sobre las relaciones entre sonido, escucha y poder.

Sean Dockray, James Parker, and Joel Stern turn machine listening into a shared field of critical writing, artistic work, software, pedagogy, and collective study.

A personal account of studying sound across philosophy, art, politics, computation, and music at The New Centre for Research & Practice.

Iben Have examines the synthetic voice of public authority, while attempted vishing attacks against Wall Street firms show how vocal trust can be used as an attack surface.

The new Sonic Field Calendar gathers opportunities and events across sound and listening culture, with date, category, and country filters and three ways to browse.

Audio8, Inflect v2, VibeVoice-ASR-BitNet, MOSS Transcribe-Diarize, Cohere Transcribe Arabic, and GPT-Live point toward speech systems that are smaller, more continuous, and more operational. LISTENING...

A new open-source Python package turns fragmented bioacoustic collections into reproducible infrastructure for studying animal communication across species.

A new Routledge book connects soundscape practice, empathy, and arts-based environmental work across nine chapters on listening and more-than-human relations.

Wojciech Dobry's See the music turns a track into a sequence of visual experiments, making audio analysis and beat-tracking decisions visible in the browser.