Crate gestura_core_audio 

Audio capture, speech processing, noise reduction, and STT abstractions.

gestura-core-audio owns the platform-independent audio domain used by the workspace. It covers microphone capture, noise cancellation, speech processing helpers, and speech-to-text provider abstractions that can be consumed by both CLI and GUI entry points.

Main entry points

• audio_capture: recording helpers, input-device discovery, and AudioCaptureConfig
• noise_cancellation: configurable noise reduction processors and stats
• speech: speech-focused processing utilities
• stt_provider: the SttProvider trait plus concrete providers such as OpenAI-compatible STT and local Whisper

Architecture role

This crate owns the reusable audio primitives and provider interfaces. GUI permission prompts, tray interactions, and other presentation-layer concerns remain outside this crate.

Stable import paths

Most application code should import these surfaces through:

• gestura_core::audio::*
• gestura_core::audio_capture::*
• gestura_core::stt_provider::*

Modules

audio_capture

Audio capture module for microphone input Uses cpal for cross-platform audio recording with voice activity detection (VAD)

noise_cancellation

Noise cancellation and audio enhancement for Gestura.app Reduces background noise and enhances speech quality

speech

Speech processing module for Gestura

stt_provider

Speech-to-text provider abstraction.