Add an option to play thinking sound

Dockerfile

@@ -0,0 +1,15 @@
+FROM python:3.13-slim-trixie
+
+ENV LANG=C.UTF-8
+ENV DEBIAN_FRONTEND=noninteractive
+
+RUN apt-get update && \
+    apt-get install --yes --no-install-recommends \
+    avahi-utils alsa-utils libportaudio2 portaudio19-dev \
+    build-essential libmpv-dev pulseaudio
+
+WORKDIR /srv
+COPY . ./
+RUN ./script/setup
+
+ENTRYPOINT ["./script/run"]

docker-compose.yaml

@@ -0,0 +1,32 @@
+version: "3.8"
+services:
+  linux-voice-assistant:
+    build: .
+    image: ohf/linux-voice-assistant:local
+    container_name: linux-voice-assistant
+    restart: unless-stopped
+    devices:
+      - "/dev/snd:/dev/snd"
+    group_add:
+      - audio
+    environment:
+      - FIXED_MAC_ADDRESS=device_mac_address
+      # - XDG_RUNTIME_DIR=/run/user/1000
+      # - PULSE_SERVER=unix:/run/user/1000/pulse/native
+      - AUDIO_OUTPUT_DEVICE=default
+    volumes:
+      - ./config:/app/.config
+      - ./wakewords:/app/wakewords
+      - ./sounds:/app/sounds
+      # - /var/run/pulse/native:/var/run/pulse/native
+      # - /run/user/1000:/run/user/1000
+    ports:
+      - "6053:6053"
+    tty: true
+    command:
+      - "--name"
+      - "voice-satellite"
+      - "--audio-input-device"
+      - "default"
+      - "--wake-model"
+      - "hey_jarvis"

linux_voice_assistant/__main__.py

@@ -80,6 +80,10 @@ async def main() -> None:
     parser.add_argument(
         "--timer-finished-sound", default=str(_SOUNDS_DIR / "timer_finished.flac")
     )
+    parser.add_argument(
+        "--processing-sound", default=str(_SOUNDS_DIR / "processing.wav"),
+        help="Short sound to play while assistant is processing (thinking)"
+    )
     #
     parser.add_argument("--preferences-file", default=_REPO_DIR / "preferences.json")
     #
@@ -184,6 +188,7 @@ async def main() -> None:
         tts_player=MpvMediaPlayer(device=args.audio_output_device),
         wakeup_sound=args.wakeup_sound,
         timer_finished_sound=args.timer_finished_sound,
+        processing_sound=args.processing_sound,
         preferences=preferences,
         preferences_path=preferences_path,
         libtensorflowlite_c_path=libtensorflowlite_c_path,

linux_voice_assistant/models.py

@@ -70,6 +70,7 @@ class ServerState:
     music_player: "MpvMediaPlayer"
     tts_player: "MpvMediaPlayer"
     wakeup_sound: str
+    processing_sound: str
     timer_finished_sound: str
     preferences: Preferences
     preferences_path: Path

linux_voice_assistant/satellite.py

@@ -65,6 +65,7 @@ def __init__(self, state: ServerState) -> None:
         self._tts_played = False
         self._continue_conversation = False
         self._timer_finished = False
+        self._processing = False
 
     def handle_voice_event(
         self, event_type: VoiceAssistantEventType, data: Dict[str, str]
@@ -75,6 +76,15 @@ def handle_voice_event(
             self._tts_url = data.get("url")
             self._tts_played = False
             self._continue_conversation = False
+        elif event_type == VoiceAssistantEventType.VOICE_ASSISTANT_INTENT_START:
+            # Play short "thinking/processing" sound if configured
+            processing = getattr(self.state, "processing_sound", None)
+            if processing:
+                _LOGGER.debug("Playing processing sound: %s", processing)
+                self.state.stop_word.is_active = True
+                self._processing = True
+                self.duck()
+                self.state.tts_player.play(self.state.processing_sound)            
         elif event_type in (
             VoiceAssistantEventType.VOICE_ASSISTANT_STT_VAD_END,
             VoiceAssistantEventType.VOICE_ASSISTANT_STT_END,

linux_voice_assistant/util.py

@@ -4,9 +4,15 @@
 import uuid
 from collections.abc import Callable
 from typing import Optional
+import os
 
 
 def get_mac() -> str:
+    # Check for fixed MAC in environment variable
+    mac_env = os.getenv("FIXED_MAC_ADDRESS")
+    if mac_env:
+        return mac_env.lower()
+    # Fallback to uuid.getnode()
     mac = uuid.getnode()
     mac_str = ":".join(f"{(mac >> i) & 0xff:02x}" for i in range(40, -1, -8))
     return mac_str