Discussions

Context:

I recently began working on migrating my organization's custom HeyGen Avatar into LiveAvatar. I have come to the step where I am changing the code that hosts connections to our Avatar, and am encountering road bumps that documentation has not clearly addressed.

Questions:

1. After using the code below to successfully establish a connection (api.js), do I need to create a new reference as shown in index.js via the new LiveAvatarSession? I see this is in our legacy code, and I am confused because I thought I had already established the session in the first step.

   api.jsindex.js

   router.get('/token', (req, res) => {
   	const liveAvatarUrl = url.parse(`${liveAvatarUrl}/v1/sessions/token`)
     const requestBody = JSON.stringify({
       mode: 'LITE',
       avatar_id: avatarId,
       is_sandbox: false,
       avatar_persona: {
         voice_id: voiceId,
         language: 'en'
       },
       interactivity_type: 'CONVERSATIONAL'
     });
   
     const options = {
       hostname: parsedUrl.hostname,
       path: parsedUrl.path,
       port: parsedUrl.port || 443,
       method: "POST",
       headers: {
         accept: 'application/json',
         'content-type': 'application/json',
         'X-API-KEY': liveAvatarKey,
         'Content-Length': Buffer.byteLength(requestBody)
       }
     };
     
   

   import { LiveAvatarSession } from '@heygen/liveavatar-web-sdk'
   
   const response = await fetch(`${HOST_URL}/api/token`);
   sessionId = response.data.session_Id;
   sessionToken = response.data.sessionToken;
   
   const userConfig = {
     voiceChat: true,
   }
   
   liveAvatarSession = new LiveAvatarSession(sessionToken, userConfig);
   

2. How do I use streaming events with the new LiveAvatar? It appears that in the legacy code, we are using session.start() and then defining listeners, is that behavior the same? Reason I ask is because I am able to create a connection and see that in the web-developer tools, but it gets hung up and never begins the session. The snippet below shows how it was done previously, I would like to know how I need to adapt it for the LiveAvatart SDK rather than the previous implementation.

import StreamingAvatar, { AvatarQuality, StreamingEvents, VoiceEmotion } from '@heygen/streaming-avatar'

liveAvatarSession = new LiveAvatarSession(sessionId, sessionToken, userConfig)
liveAvatarSession.on('avatar_start_talking', () => { // How do I do this with LiveAvatar rather than HeyGen?
	console.log('Avatar started talking' 
});