Audio Output and Devices

Devices are listed and changed with the methods of Device management, micDevices, speakerDevices, changeMicDevice and changeSpeakerDevice. What a phone changes is what those lists hold.

Loudspeaker or earpiece

A phone has one decision to make about output: whether the voice comes out loud or against the ear. The two routes are what speakerDevices holds here, so the same code that picks an output in a browser works.

import {
  EARPIECE_DEVICE,
  Micdrop,
  SPEAKER_DEVICE,
} from '@micdrop/react-native'

await Micdrop.changeSpeakerDevice(EARPIECE_DEVICE)
await Micdrop.changeSpeakerDevice(SPEAKER_DEVICE)

The current one is in the state:

const state = useMicdropState()
const onEarpiece = state.speakerDeviceId === EARPIECE_DEVICE

const handlePress = () =>
  Micdrop.changeSpeakerDevice(onEarpiece ? SPEAKER_DEVICE : EARPIECE_DEVICE)

Headphones and Bluetooth are handled by the operating system: plugging them in moves the call over, whatever was chosen here.

The audio played here is PCM streamed by the server, which is one of two ways to get a voice out of a React Native app.

Choosing the microphone

The inputs the OS reports are in the state, each with an id to hand back and a label to show, and the call can be moved to another one without interrupting it:

const state = useMicdropState()

state.micDevices.map((device) => (
  <Button
    key={device.id}
    title={device.label}
    onPress={() => Micdrop.changeMicDevice(device.id)}
  />
))

The list is filled once the call has started, since the audio session has to be active for the system to answer.

Echo

The audio session is configured as a voice call, which is what tells iOS and Android to apply echo cancellation. Without it, the microphone would hear the assistant and the call would interrupt itself.

On a device where cancellation is weak, turn interruption off so the microphone is ignored while the assistant speaks:

await Micdrop.start({
  url: 'wss://example.com/call',
  disableInterruption: true,
})