Microphone Plugins

March 12, 2026 · View on GitHub

Entry point group: opm.microphone Template: ovos_plugin_manager.templates.microphone Factory: ovos_plugin_manager.microphone

Microphone plugins provide a raw audio stream to ovos-dinkum-listener. They abstract over different audio sources: ALSA, PulseAudio, network streams, files, etc.

`OVOSMicrophone` base class

from ovos_plugin_manager.templates.microphone import OVOSMicrophone

Constructor

OVOSMicrophone(config: Optional[dict] = None)

Attribute	Type	Description
`sample_rate`	`int`	Samples per second (default 16000).
`sample_width`	`int`	Bytes per sample (default 2 = 16-bit).
`sample_channels`	`int`	Number of channels (default 1 = mono).
`chunk_size`	`int`	Bytes per read call (default 4096).

Abstract methods (must implement)

`open()`

Open the audio source and prepare for reading.

`close()`

Release the audio source.

`read_chunk() -> Optional[bytes]`

Return the next chunk of raw PCM audio bytes, or None if no data is available.

Provided methods

`enter() / exit()`

Context manager support — calls open() / close().

Module-level helpers

from ovos_plugin_manager.microphone import (
    find_microphone_plugins,
    load_microphone_plugin,
    get_microphone_config,
    OVOSMicrophoneFactory,
)

`find_microphone_plugins() -> dict`

Return {entry_point_name: class} for all installed microphone plugins.

`load_microphone_plugin(module_name: str) -> Type[OVOSMicrophone]`

Return the uninstantiated class.

`get_microphone_config(config: dict = None) -> dict`

Resolve microphone config from the global configuration (listener.microphone section).

`OVOSMicrophoneFactory.create(config=None) -> OVOSMicrophone`

Instantiate and return a microphone based on configuration.

Configuration

{
  "listener": {
    "microphone": {
      "module": "ovos-microphone-plugin-alsa",
      "ovos-microphone-plugin-alsa": {
        "device": "default",
        "sample_rate": 16000
      }
    }
  }
}