Skip to main content

Text-to-Speech Example

Send text to the device and have it spoken through the built-in speakers — useful for prompting or instructing the wearer during a session.

Script: device_tts.py (export the samples to get it)

Prerequisites​

One-Time Setup: Device Authentication Required

Before running this example you must authenticate your device by running aria_gen2 auth pair (see Device Authentication). This only needs to be done once per device-PC combination, but all device operations will fail without it.

Run it​

python ~/Downloads/projectaria_client_sdk_samples_gen2/device_tts.py --text "Hello from Aria"

--text is required. The TTS engine handles numbers and punctuation naturally, so "You have 3 notifications. Check your device." reads the way you would expect.

Worth knowing​

render_tts() returns before the audio finishes. The call sends the text and returns immediately; the device synthesizes and plays asynchronously. A script that exits right after calling it may cut the utterance short, and a long utterance keeps playing while your code moves on.

stop_tts() interrupts playback:

device.render_tts("This is a long announcement that we will interrupt")
time.sleep(1)
device.stop_tts()

The CLI equivalents are aria_gen2 device tts start --tts-text "..." and aria_gen2 device tts stop.

render_tts() synthesizes on the device. To play audio you produced — a recorded clip, your own TTS engine's output, a synthesized tone — use push_audio() over a WebRTC session instead. See WebRTC Streaming Example → Pushing Audio to the Glasses.

Troubleshooting​

Command succeeds but the device is silent. Check that device audio is not muted and the volume is up in the Companion App. A very low battery can also affect audio output.

Connection drops when sending TTS. Verify the USB connection is stable and re-check aria_gen2 auth check. Reconnect and retry:

try:
    device.render_tts(text)
except Exception as e:
    print(f"Connection lost: {e}")
    device = device_client.connect()
    device.render_tts(text)

Next steps​