@jambonz/speech-utils 0.0.28 → 0.0.30

package/lib/get-aws-sts-token.js

@@ -27,7 +27,7 @@ async function getAwsAuthToken(
     const credentials = {
       accessKeyId: data.Credentials.AccessKeyId,
       secretAccessKey: data.Credentials.SecretAccessKey,
-      sessionToken: data.Credentials.SessionToken
+      securityToken: data.Credentials.SessionToken
     };
 
     /* expire 10 minutes before the hour, so we don't lose the use of it during a call */

package/lib/synth-audio.js

@@ -76,7 +76,7 @@ const trimTrailingSilence = (buffer) => {
  * the synthesized audio, and a variable indicating whether it was served from cache
  */
 async function synthAudio(client, logger, stats, { account_sid,
-  vendor, language, voice, gender, text, engine, salt, model, credentials, deploymentId, disableTtsCache
+  vendor, language, voice, gender, text, engine, salt, model, credentials, deploymentId, disableTtsCache, options
 }) {
   let audioBuffer;
   let servedFromCache = false;
@@ -194,7 +194,7 @@ async function synthAudio(client, logger, stats, { account_sid,
         audioBuffer = await synthWellSaid(logger, {credentials, stats, language, voice, text, filePath});
         break;
       case 'elevenlabs':
-        audioBuffer = await synthElevenlabs(logger, {credentials, stats, language, voice, text, filePath});
+        audioBuffer = await synthElevenlabs(logger, {credentials, options, stats, language, voice, text, filePath});
         break;
       case 'whisper':
         audioBuffer = await synthWhisper(logger, {credentials, stats, voice, text});
@@ -585,21 +585,25 @@ const synthCustomVendor = async(logger, {credentials, stats, language, voice, te
   }
 };
 
-const synthElevenlabs = async(logger, {credentials, stats, language, voice, text}) => {
-  const {api_key, model_id} = credentials;
+const synthElevenlabs = async(logger, {credentials, options, stats, language, voice, text}) => {
+  const {api_key, model_id, options: credOpts} = credentials;
+  const opts = !!options && Object.keys(options).length !== 0 ? options : JSON.parse(credOpts || '{}');
+  const optimize_streaming_latency = opts.optimize_streaming_latency ?
+    `?optimize_streaming_latency=${opts.optimize_streaming_latency}` : '';
   try {
     const post = bent('https://api.elevenlabs.io', 'POST', 'buffer', {
       'xi-api-key': api_key,
       'Accept': 'audio/mpeg',
       'Content-Type': 'application/json'
     });
-    const mp3 = await post(`/v1/text-to-speech/${voice}`, {
+    const mp3 = await post(`/v1/text-to-speech/${voice}${optimize_streaming_latency}`, {
       text,
       model_id,
       voice_settings: {
         stability: 0.5,
         similarity_boost: 0.5
-      }
+      },
+      ...opts
     });
     return mp3;
   } catch (err) {

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@jambonz/speech-utils",
-  "version": "0.0.28",
+  "version": "0.0.30",
   "description": "TTS-related speech utilities for jambonz",
   "main": "index.js",
   "author": "Dave Horton",

package/test/aws.js

@@ -21,12 +21,12 @@ test('AWS - create and cache auth token', async(t) => {
   try {
     let obj = await getAwsAuthToken(process.env.AWS_ACCESS_KEY_ID, process.env.AWS_SECRET_ACCESS_KEY, process.env.AWS_REGION);
     //console.log({obj}, 'received auth token from AWS');
-    t.ok(obj.sessionToken && !obj.servedFromCache, 'successfullY generated auth token from AWS');
+    t.ok(obj.securityToken && !obj.servedFromCache, 'successfullY generated auth token from AWS');
 
     await sleep(250);
     obj = await getAwsAuthToken(process.env.AWS_ACCESS_KEY_ID, process.env.AWS_SECRET_ACCESS_KEY, process.env.AWS_REGION);
     //console.log({obj}, 'received auth token from AWS - second request');
-    t.ok(obj.sessionToken && obj.servedFromCache, 'successfully received access token from cache');
+    t.ok(obj.securityToken && obj.servedFromCache, 'successfully received access token from cache');
  
     await client.flushall();
     t.end();

package/test/synth.js

@@ -459,7 +459,16 @@ test('Elevenlabs speech synth tests', async(t) => {
       vendor: 'elevenlabs',
       credentials: {
         api_key: process.env.ELEVENLABS_API_KEY,
-        model_id: process.env.ELEVENLABS_MODEL_ID
+        model_id: process.env.ELEVENLABS_MODEL_ID,
+        options: JSON.stringify({
+          optimize_streaming_latency: 1,
+          voice_settings: {
+            similarity_boost: 1,
+            stability: 0.8,
+            style: 1,
+            use_speaker_boost: true
+          }
+        })
       },
       language: 'en-US',
       voice: process.env.ELEVENLABS_VOICE_ID,