curl -X POST http://<hostname>:<port>/api/settings/apply
```
# Events
# Speaker Monitor
The API endpoint `/audio/speaker` provides access to the audio stream played by the terminal through the speaker in real-time. This API is used by the web-based user interface to "listen in" on the audio played by the terminal. This can be used, for example, by experiment designers to remotely experiment with various kinds of impairments.
To start streaming, send a HTTP GET to the API endpoint. The terminal will respond with a 200 OK with the Transfer-Encoding header set to "chunked". Close the connection to stop streaming. The format of the data stream can be controlled with the following query string parameters:
***type** - Content (stream) type. If unset or set to 'audio/wav', the terminal will include a wav header at the beginning of the stream. If set to 'audio/vnd.lmr.pcm', the header will be omitted and the terminal will stream raw audio data. If the wav header is included, the total length will be set to 0 in the header.
***channels** - Number of audio channels between 1 and 32. The default value is 1.
***rate** - Sample rate. The value must be between 8000 and 96000. The default value is 22050.
***format** - Sample format. One of u8, u-law, A-law, S16LE, S16BE, Float32LE, Float32BE, S32LE, S32BE, S24LE, S24BE, S24_32LE, S24_32BE. The default value is S16LE. Only u8 and S16LE can be used with stream type 'audio/wav'.
***fragmentSize** - This parameter can be used to control the size of the data fragment, i.e., how often the audio subsystem wakes up the application. The value indirectly controls latency. Please note that the value is advisory only. The actual fragment size will be determined based on a variety of factors and can differ significantly (in both directions) from the value specified by the client.
