├── tencent_virtual_human_completeV1.py   # Main script (Python)
├── trtc_player.html                       # TRTC protocol H5 playback page
├── webrtc_player.html                     # WebRTC protocol H5 playback page
└── TcPlayer-2.4.5.js                      # TCPlayerLite SDK (WebRTC playback dependency)

pip install requests websocket-client pydub

# macOS
brew install ffmpeg

# Ubuntu/Debian
sudo apt install ffmpeg

# Windows
# Download https://ffmpeg.org/download.html and add it to PATH

CONFIG = {
    "appkey": "your_appkey",
    "accesstoken": "your_accesstoken",
    "asset_virtualman_key": "your_asset_key",   # Image Asset ID (preferred)
    "virtualman_project_id": "",                  # Project ID (used when asset_key is empty)
    "protocol": "rtmp",                           # Streaming protocol: "rtmp" / "trtc" / "webrtc"
    "protocol_option": None,                      # Protocol option (see Section 5)
}

python tencent_virtual_human_completeV1.py

1️⃣ Create a session (automatically select the stream creation method)
2️⃣ Wait for session readiness (polling status, maximum 120 seconds)
3️⃣ Enable session
🎬 Automatically launch the H5 player (for TRTC/WebRTC protocols)
4️⃣ Create WebSocket connection
5️⃣ Enter interactive mode
6️⃣ Close session

📋 Driver mode selection:
  1 - text-driven (input text to make Digital Humans speak)
  2 - audio-driven (selecting audio files to drive the Digital Human)
  q - Exit

CONFIG = {
    "asset_virtualman_key": "your_asset_key",
    "virtualman_project_id": "",  # Leave blank
}

CONFIG = {
    "asset_virtualman_key": "",   # Leave blank
    "virtualman_project_id": "your_project_id",
}

def create_session(self) -> Tuple[bool, str]:
    if self.asset_virtualman_key:
        return self.create_session_by_asset()    # Priority
    elif self.virtualman_project_id:
        return self.create_session_by_project()  # Fallback
    else:
        return False, "Both parameters are empty, unable to create stream"

CONFIG = {
    "protocol": "rtmp",
    "protocol_option": None,
}

# Debug Mode (using the platform-unified AppId, no additional configuration required)
CONFIG = {
    "protocol": "trtc",
    "protocol_option": None,
}

CONFIG = {
    "protocol": "webrtc",
    "protocol_option": None,
}

CONFIG = {
    "protocol": "trtc",
    "protocol_option": {
        "TrtcUseExternalApp": True,
        "TrtcAppId": "1400xxxxxx",
        "TrtcRoomId": 12345,
        "TrtcUserSig": "eJw8js0Kgk...",
        "TrtcPrivateMapKey": "dummy"
    },
}

CONFIG = {
    "protocol": "rtmp",
    "protocol_option": {
        "CssCustomPushUrl": "rtmp://domain/appName/streamName?txSecret={0}&txTime={1}"
    },
}

"ExtraInfo": {"AlphaChannelEnable":True} # Enable Alpha channel (if transparent background is required)

// Enter the room (pull stream only as audience, no pushing)
await trtc.enterRoom({
    sdkAppId: config.appId,
    userId: config.userId,
    userSig: config.userSig,
    roomId: config.roomId,
    scene: 'live',
    role: 'audience'
});

// Listen for and play remote video
trtc.on(TRTC.EVENT.REMOTE_VIDEO_AVAILABLE, async (event) => {
    await trtc.startRemoteVideo({
        userId: event.userId,
        streamType: event.streamType,
        view: 'remote-video',
        option: { objectFit: 'contain' }
    });
});

http://localhost:8080/trtc_player.html?appId=1400695865&roomId=402183450&userId=user_xxx&userSig=eJw8...&virtualManUserId=402183450_ivh_anchor

player = new TcPlayer('player-container', {
    "webrtc": webrtcUrl,
    "width": '100%',
    "height": '540',
    "autoplay": true,
    "live": true,
    "controls": "none",
    "webrtcConfig": {
        "streamType": "auto"
    },
    "listener": function (msg) {
        handlePlayerEvent(msg);
    }
});

http://localhost:8080/webrtc_player.html?url=webrtc://liveplay.ivh.qq.com/live/m11533590420520971383?min_delay_ms=100

def _generate_signature(self, parameters: Dict[str, str]) -> str:
    sorted_params = sorted(parameters.items())
    signing_content = '&'.join(f'{k}={v}' for k, v in sorted_params)
    h = hmac.new(
        self.accesstoken.encode('utf-8'),
        signing_content.encode('utf-8'),
        hashlib.sha256
    )
    hash_in_base64 = base64.b64encode(h.digest()).decode('utf-8')
    return quote(hash_in_base64)

┌─────────────┐
│  Create session     │  create_session()
└──────┬──────┘
       ▼
┌─────────────┐
│  Wait for readiness     │  wait_for_session_ready()  ← Poll SessionStatus
└──────┬──────┘       SessionStatus=3 → Preparing
       ▼              SessionStatus=1 → Ready
┌─────────────┐
│  Enable session     │  start_session()
└──────┬──────┘
       ▼
┌─────────────┐
│  H5 Player   │  start_h5_player()  ← Automatically launched for TRTC/WebRTC
└──────┬──────┘
       ▼
┌─────────────┐
│  WebSocket   │  create_websocket_connection()
│  Persistent connection channel   │
└──────┬──────┘
       ▼
┌─────────────┐
│  Interactive Drive     │  send_text_drive() / send_audio_drive()
│  (Loop)     │
└──────┬──────┘
       ▼
┌─────────────┐
│  Close session     │  close_session()
└─────────────┘

drive_cmd = {
    "Header": {},
    "Payload": {
        "ReqId": req_id,
        "SessionId": self.session_id,
        "Command": "SEND_TEXT",
        "Data": {
            "Text": "Hello, welcome to Tencent Cloud Digital Human Platform",
            "ChatCommand": "NotUseChat"
        }
    }
}
self.ws.send(json.dumps(drive_cmd, ensure_ascii=False))

audio = AudioSegment.from_file(audio_file_path)
audio = audio.set_channels(1).set_frame_rate(16000).set_sample_width(2)
pcm_data = audio.raw_data

_ffmpeg_path = shutil.which('ffmpeg')

h5_url = self.start_h5_player(h5_port=9090)