Skip to content

Speaker

A speaker object represents the physical speaker or audio backchannel of a device integrated into Netsocs. Its function is to receive audio from the platform and play it on the physical device, enabling voice communication in the operator → device direction.

It complements the microphone object: while the microphone captures audio from the environment, the speaker allows an operator to speak towards it. Together they form the basis of a two-way communication system.

Main characteristics

  • Real-time talkback: Receives RTP packets from DriversHub and forwards them to the device audio channel.
  • Managed sessions: Each talkback session has a unique session_id that allows them to be managed independently.
  • Output configuration: Exposes audio output parameters from the device (codec, sample rate, volume level).

Common use cases

  • IP intercoms and PA systems.
  • Real-time voice instructions to monitored areas.
  • Automated audio warnings triggered by automations.
  • Two-way communication on IP cameras with built-in speaker.

States

State Description
speaker.state.idle The speaker is available but not playing audio.
speaker.state.talkback An active talkback session is in progress.

Actions

Action Description
speaker.action.start_talkback Starts a talkback session. Requires session_id in the payload.
speaker.action.stop_talkback Stops the active talkback session for the given session_id.

State properties

Property Description
profile_token Device media profile token.
output_token Audio output token on the device.
decoder_token Audio decoder token for the backchannel.
codec Audio codec in use (e.g. G711).
sample_rate Sampling frequency in Hz (e.g. 8000).
output_level Output volume level from 0 to 100.