@jambonz/speech-utils 0.0.42 → 0.0.44

package/lib/synth-audio.js

@@ -77,7 +77,7 @@ const trimTrailingSilence = (buffer) => {
  */
 async function synthAudio(client, logger, stats, { account_sid,
   vendor, language, voice, gender, text, engine, salt, model, credentials, deploymentId,
-  disableTtsCache, renderForCaching, options
+  disableTtsCache, renderForCaching, disableTtsStreaming, options
 }) {
   let audioBuffer;
   let servedFromCache = false;
@@ -200,21 +200,14 @@ async function synthAudio(client, logger, stats, { account_sid,
         break;
       case 'elevenlabs':
         audioBuffer = await synthElevenlabs(logger, {
-          credentials, options, stats, language, voice, text, renderForCaching, filePath
+          credentials, options, stats, language, voice, text, renderForCaching, disableTtsStreaming, filePath
         });
-        if (typeof audioBuffer === 'object' && audioBuffer.filePath) {
-          return audioBuffer;
-        }
-        else {
-          audioBuffer = await synthElevenlabs(logger, {credentials, options, stats, language, voice, text, filePath});
-        }
+        if (audioBuffer?.filePath) return audioBuffer;
         break;
       case 'whisper':
-        audioBuffer = await synthWhisper(logger, {credentials, stats, voice, text, renderForCaching});
-        if (typeof audioBuffer === 'object' && audioBuffer.filePath) {
-          return audioBuffer;
-        }
-        audioBuffer = await synthWhisper(logger, {credentials, stats, voice, text});
+        audioBuffer = await synthWhisper(logger, {
+          credentials, stats, voice, text, renderForCaching, disableTtsStreaming});
+        if (audioBuffer?.filePath) return audioBuffer;
         break;
       case 'deepgram':
         audioBuffer = await synthDeepgram(logger, {credentials, stats, model, text});
@@ -611,14 +604,18 @@ const synthCustomVendor = async(logger, {credentials, stats, language, voice, te
   }
 };
 
-const synthElevenlabs = async(logger, {credentials, options, stats, language, voice, text, renderForCaching}) => {
+const synthElevenlabs = async(logger, {
+  credentials, options, stats, voice, text, renderForCaching, disableTtsStreaming
+}) => {
   const {api_key, model_id, options: credOpts} = credentials;
   const opts = !!options && Object.keys(options).length !== 0 ? options : JSON.parse(credOpts || '{}');
 
   /* default to using the streaming interface, unless disabled by env var OR we want just a cache file */
-  if (!process.env.JAMBONES_DISABLE_TTS_STREAMING && !renderForCaching) {
+  if (!process.env.JAMBONES_DISABLE_TTS_STREAMING && !renderForCaching && !disableTtsStreaming) {
     let params = '';
     params += `{api_key=${api_key}`;
+    params += ',vendor=elevenlabs';
+    params += `,voice=${voice}`;
     params += `,model_id=${model_id}`;
     params += `,optimize_streaming_latency=${opts.optimize_streaming_latency || 2}`;
     params += ',write_cache_file=1';
@@ -660,13 +657,14 @@ const synthElevenlabs = async(logger, {credentials, options, stats, language, vo
   }
 };
 
-const synthWhisper = async(logger, {credentials, stats, voice, text, renderForCaching}) => {
+const synthWhisper = async(logger, {credentials, stats, voice, text, renderForCaching, disableTtsStreaming}) => {
   const {api_key, model_id, baseURL, timeout, speed} = credentials;
   /* if the env is set to stream then bag out, unless we are specifically rendering to generate a cache file */
-  if (!process.env.JAMBONES_DISABLE_TTS_STREAMING && !renderForCaching) {
+  if (!process.env.JAMBONES_DISABLE_TTS_STREAMING && !renderForCaching && !disableTtsStreaming) {
     let params = '';
     params += `{api_key=${api_key}`;
     params += `,model_id=${model_id}`;
+    params += ',vendor=whisper';
     params += `,voice=${voice}`;
     params += ',write_cache_file=1';
     if (speed) params += `,speed=${speed}`;
@@ -702,7 +700,7 @@ const synthWhisper = async(logger, {credentials, stats, voice, text, renderForCa
 const synthDeepgram = async(logger, {credentials, stats, model, text}) => {
   const {api_key} = credentials;
   try {
-    const post = bent('https://api.beta.deepgram.com', 'POST', 'buffer', {
+    const post = bent('https://api.deepgram.com', 'POST', 'buffer', {
       'Authorization': `Token ${api_key}`,
       'Accept': 'audio/mpeg',
       'Content-Type': 'application/json'

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@jambonz/speech-utils",
-  "version": "0.0.42",
+  "version": "0.0.44",
   "description": "TTS-related speech utilities for jambonz",
   "main": "index.js",
   "author": "Dave Horton",

package/test/synth.js

@@ -212,6 +212,56 @@ test('Azure speech synth tests', async(t) => {
   client.quit();
 });
 
+test('Azure SSML tests', async(t) => {
+  const fn = require('..');
+  const {synthAudio, client} = fn(opts, logger);
+
+  if (!process.env.MICROSOFT_API_KEY || !process.env.MICROSOFT_REGION) {
+    t.pass('skipping Microsoft speech synth tests since MICROSOFT_API_KEY or MICROSOFT_REGION not provided');
+    return t.end();
+  }
+  try {
+    const text = `<speak version="1.0" xmlns="http://www.w3.org/2001/10/synthesis" xmlns:mstts="https://www.w3.org/2001/mstts" xml:lang="en-US">
+    <voice name="en-US-JennyMultilingualNeural">
+    <mstts:express-as style="cheerful" styledegree="2">That'd be just amazing!
+    </mstts:express-as>
+    </voice>
+    </speak>`;
+
+    let opts = await synthAudio(stats, {
+      vendor: 'microsoft',
+      credentials: {
+        api_key: process.env.MICROSOFT_API_KEY,
+        region: process.env.MICROSOFT_REGION,
+      },
+      language: 'en-US',
+      voice: 'en-US-ChristopherNeural',
+      text,
+    });
+    t.ok(!opts.servedFromCache, `successfully synthesized microsoft audio to ${opts.filePath}`);
+    if (process.env.JAMBONES_HTTP_PROXY_IP && process.env.JAMBONES_HTTP_PROXY_PORT) {
+      t.pass('successfully used proxy to reach microsoft tts service');
+    }
+
+    opts = await synthAudio(stats, {
+      vendor: 'microsoft',
+      credentials: {
+        api_key: process.env.MICROSOFT_API_KEY,
+        region: process.env.MICROSOFT_REGION,
+      },
+      language: 'en-US',
+      voice: 'en-US-ChristopherNeural',
+      text,
+    });
+    t.ok(opts.servedFromCache, `successfully retrieved microsoft audio from cache ${opts.filePath}`);
+  } catch (err) {
+    console.error(err);
+    t.end(err);
+  }
+  client.quit();
+});
+
+
 test('Azure custom voice speech synth tests', async(t) => {
   const fn = require('..');
   const {synthAudio, client} = fn(opts, logger);
@@ -537,7 +587,7 @@ test('Deepgram speech synth tests', async(t) => {
       credentials: {
         api_key: process.env.DEEPGRAM_API_KEY
       },
-      model: 'alpha-asteria-en-v2',
+      model: 'aura-asteria-en',
       text,
     });
     t.ok(!opts.servedFromCache, `successfully synthesized deepgram audio to ${opts.filePath}`);