0.2.24

set write_cache_file = 0 when disableTtsCache

lib/synth-audio.js

@@ -53,9 +53,6 @@ const EXPIRES = JAMBONES_TTS_CACHE_DURATION_MINS;
 const OpenAI = require('openai');
 const getAwsAuthToken = require('./get-aws-sts-token');
 
-/* counter for playback id */
-let playbackIdCounter = 0;
-
 const trimTrailingSilence = (buffer) => {
   assert.ok(buffer instanceof Buffer, 'trimTrailingSilence - argument is not a Buffer');
 
@@ -207,74 +204,81 @@ async function synthAudio(client, createHash, retrieveHash, logger, stats, { acc
     const startAt = process.hrtime();
     switch (vendor) {
       case 'google':
-        audioData = await synthGoogle(logger, {credentials, stats, language, voice, gender, text});
+        audioData = await synthGoogle(logger, {credentials, stats, language, voice, gender, key, text});
         break;
       case 'aws':
       case 'polly':
         vendorLabel = 'aws';
         audioData = await synthPolly(createHash, retrieveHash, logger,
-          {credentials, stats, language, voice, text, engine, renderForCaching, disableTtsStreaming});
+          {credentials, stats, language, voice, key, text, engine, renderForCaching, disableTtsStreaming,
+            disableTtsCache});
         break;
       case 'azure':
       case 'microsoft':
         vendorLabel = 'microsoft';
-        audioData = await synthMicrosoft(logger, {credentials, stats, language, voice, text, deploymentId,
+        audioData = await synthMicrosoft(logger, {credentials, stats, language, voice, key, text, deploymentId,
-          renderForCaching, disableTtsStreaming});
+          renderForCaching, disableTtsStreaming, disableTtsCache});
         break;
       case 'nuance':
         model = model || 'enhanced';
-        audioData = await synthNuance(client, logger, {credentials, stats, voice, model, text});
+        audioData = await synthNuance(client, logger, {credentials, stats, voice, model, key, text});
         break;
       case 'nvidia':
-        audioData = await synthNvidia(client, logger, {credentials, stats, language, voice, model, text,
+        audioData = await synthNvidia(client, logger, {credentials, stats, language, voice, model, key, text,
-          renderForCaching, disableTtsStreaming});
+          renderForCaching, disableTtsStreaming, disableTtsCache});
         break;
       case 'ibm':
-        audioData = await synthIbm(logger, {credentials, stats, voice, text});
+        audioData = await synthIbm(logger, {credentials, stats, voice, key, text});
         break;
       case 'wellsaid':
-        audioData = await synthWellSaid(logger, {credentials, stats, language, voice, text});
+        audioData = await synthWellSaid(logger, {credentials, stats, language, voice, key, text});
         break;
       case 'elevenlabs':
         audioData = await synthElevenlabs(logger, {
-          credentials, options, stats, language, voice, text, renderForCaching, disableTtsStreaming});
+          credentials, options, stats, language, voice, key, text, renderForCaching, disableTtsStreaming,
+          disableTtsCache});
         break;
       case 'playht':
         audioData = await synthPlayHT(client, logger, {
-          credentials, options, stats, language, voice, text, renderForCaching, disableTtsStreaming});
+          credentials, options, stats, language, voice, key, text, renderForCaching, disableTtsStreaming,
+          disableTtsCache});
         break;
       case 'cartesia':
         audioData = await synthCartesia(logger, {
-          credentials, options, stats, language, voice, text, renderForCaching, disableTtsStreaming});
+          credentials, options, stats, language, voice, key, text, renderForCaching, disableTtsStreaming,
+          disableTtsCache});
         break;
       case 'inworld':
         audioData = await synthInworld(logger, {
-          credentials, options, stats, language, voice, text, renderForCaching, disableTtsStreaming});
+          credentials, options, stats, language, voice, key, text, renderForCaching, disableTtsStreaming,
+          disableTtsCache});
         break;
       case 'rimelabs':
         audioData = await synthRimelabs(logger, {
-          credentials, options, stats, language, voice, text, renderForCaching, disableTtsStreaming});
+          credentials, options, stats, language, voice, key, text, renderForCaching, disableTtsStreaming,
+          disableTtsCache});
         break;
       case 'whisper':
         audioData = await synthWhisper(logger, {
-          credentials, stats, voice, text, instructions, renderForCaching, disableTtsStreaming});
+          credentials, stats, voice, key, text, instructions, renderForCaching, disableTtsStreaming,
+          disableTtsCache});
         break;
       case 'verbio':
         audioData = await synthVerbio(client, logger, {
-          credentials, stats, voice, text, renderForCaching, disableTtsStreaming});
+          credentials, stats, voice, key, text, renderForCaching, disableTtsStreaming, disableTtsCache});
         if (audioData?.filePath) return audioData;
         break;
       case 'deepgram':
-        audioData = await synthDeepgram(logger, {credentials, stats, model, text,
+        audioData = await synthDeepgram(logger, {credentials, stats, model, key, text,
-          renderForCaching, disableTtsStreaming});
+          renderForCaching, disableTtsStreaming, disableTtsCache});
         break;
       case 'resemble':
         audioData = await synthResemble(logger, {
-          credentials, stats, voice, text, options, renderForCaching, disableTtsStreaming});
+          credentials, stats, voice, key, text, options, renderForCaching, disableTtsStreaming, disableTtsCache});
         break;
       case vendor.startsWith('custom') ? vendor : 'cant_match_value':
         audioData = await synthCustomVendor(logger,
-          {credentials, stats, language, voice, text});
+          {credentials, stats, language, voice, key, text, renderForCaching, disableTtsStreaming, disableTtsCache});
         break;
       default:
         assert(`synthAudio: unsupported speech vendor ${vendor}`);
@@ -309,13 +313,14 @@ async function synthAudio(client, createHash, retrieveHash, logger, stats, { acc
 }
 
 const synthPolly = async(createHash, retrieveHash, logger,
-  {credentials, stats, language, voice, engine, text, renderForCaching, disableTtsStreaming}) => {
+  {credentials, stats, language, voice, engine, key, text, renderForCaching, disableTtsStreaming, disableTtsCache}) => {
   const {region, accessKeyId, secretAccessKey, roleArn} = credentials;
   if (!JAMBONES_DISABLE_TTS_STREAMING && !renderForCaching && !disableTtsStreaming) {
 
     let params = '{';
     params += `language=${language}`;
-    params += ',write_cache_file=1';
+    params += `,playback_id=${key}`;
+    params += `,write_cache_file=${disableTtsCache ? 0 : 1}`;
     params += ',vendor=aws';
     if (accessKeyId && secretAccessKey) {
       if (accessKeyId) params += `,accessKeyId=${accessKeyId}`;
@@ -562,9 +567,11 @@ const synthMicrosoft = async(logger, {
   stats,
   language,
   voice,
+  key,
   text,
   renderForCaching,
-  disableTtsStreaming
+  disableTtsStreaming,
+  disableTtsCache
 }) => {
   try {
     const {api_key: apiKey, region, use_custom_tts, custom_tts_endpoint, custom_tts_endpoint_url} = credentials;
@@ -593,10 +600,11 @@ const synthMicrosoft = async(logger, {
       !renderForCaching && !disableTtsStreaming) {
       let params = '{';
       params += `api_key=${apiKey}`;
+      params += `,playback_id=${key}`;
       params += `,language=${language}`;
       params += ',vendor=microsoft';
       params += `,voice=${voice}`;
-      params += ',write_cache_file=1';
+      params += `,write_cache_file=${disableTtsCache ? 0 : 1}`;
       if (region) params += `,region=${region}`;
       if (custom_tts_endpoint) params += `,endpointId=${custom_tts_endpoint}`;
       if (custom_tts_endpoint_url) params += `,endpoint=${custom_tts_endpoint_url}`;
@@ -769,15 +777,16 @@ const synthNuance = async(client, logger, {credentials, stats, voice, model, tex
 };
 
 const synthNvidia = async(client, logger, {
-  credentials, stats, language,  voice, model, text, renderForCaching, disableTtsStreaming
+  credentials, stats, language,  voice, model, key, text, renderForCaching, disableTtsStreaming, disableTtsCache
 }) => {
   const {riva_server_uri} = credentials;
   if (!JAMBONES_DISABLE_TTS_STREAMING && !renderForCaching && !disableTtsStreaming) {
     let params = '';
     params += `{riva_server_uri=${riva_server_uri}`;
+    params += `,playback_id=${key}`;
     params += `,voice=${voice}`;
     params += `,language=${language}`;
-    params += ',write_cache_file=1';
+    params += `,write_cache_file=${disableTtsCache ? 0 : 1}`;
     params += '}';
 
     return {
@@ -817,9 +826,27 @@ const synthNvidia = async(client, logger, {
 };
 
 
-const synthCustomVendor = async(logger, {credentials, stats, language, voice, text, filePath}) => {
+const synthCustomVendor = async(logger, {credentials, stats, language, voice,
+  text, filePath, renderForCaching, disableTtsStreaming, key, disableTtsCache}) => {
   const {vendor, auth_token, custom_tts_url} = credentials;
 
+  if (!JAMBONES_DISABLE_TTS_STREAMING && !renderForCaching && !disableTtsStreaming) {
+    let params = '{';
+    params += `auth_token=${auth_token}`;
+    params += `,playback_id=${key}`;
+    params += `,custom_tts_url=${custom_tts_url}`;
+    params += ',vendor=custom';
+    params += `,voice=${voice}`;
+    params += `,write_cache_file=${disableTtsCache ? 0 : 1}`;
+    params += '}';
+
+    return {
+      filePath: `say:${params}${text.replace(/\n/g, ' ').replace(/\r/g, ' ')}`,
+      servedFromCache: false,
+      rtt: 0
+    };
+  }
+
   try {
     const post = bent('POST', {
       'Authorization': `Bearer ${auth_token}`,
@@ -848,7 +875,7 @@ const synthCustomVendor = async(logger, {credentials, stats, language, voice, te
 };
 
 const synthElevenlabs = async(logger, {
-  credentials, options, stats, voice, text, renderForCaching, disableTtsStreaming
+  credentials, options, stats, voice, key, text, renderForCaching, disableTtsStreaming, disableTtsCache
 }) => {
   const {api_key, model_id, options: credOpts} = credentials;
   const opts = !!options && Object.keys(options).length !== 0 ? options : JSON.parse(credOpts || '{}');
@@ -857,12 +884,14 @@ const synthElevenlabs = async(logger, {
   if (!JAMBONES_DISABLE_TTS_STREAMING && !renderForCaching && !disableTtsStreaming) {
     let params = '{';
     params += `api_key=${api_key}`;
-    params += `,playback_id=${playbackIdCounter++}`;
+    params += `,playback_id=${key}`;
     params += ',vendor=elevenlabs';
     params += `,voice=${voice}`;
     params += `,model_id=${model_id}`;
-    params += `,optimize_streaming_latency=${opts.optimize_streaming_latency || 2}`;
+    params += `,write_cache_file=${disableTtsCache ? 0 : 1}`;
-    params += ',write_cache_file=1';
+    if (opts.optimize_streaming_latency !== null && opts.optimize_streaming_latency !== undefined) {
+      params += `,optimize_streaming_latency=${opts.optimize_streaming_latency}`;
+    }
     if (opts.voice_settings?.similarity_boost) params += `,similarity_boost=${opts.voice_settings.similarity_boost}`;
     if (opts.voice_settings?.stability) params += `,stability=${opts.voice_settings.stability}`;
     if (opts.voice_settings?.style) params += `,style=${opts.voice_settings.style}`;
@@ -912,7 +941,7 @@ const synthElevenlabs = async(logger, {
 };
 
 const synthPlayHT = async(client, logger, {
-  credentials, options, stats, voice, language, text, renderForCaching, disableTtsStreaming
+  credentials, options, stats, voice, language, key, text, renderForCaching, disableTtsStreaming, disableTtsCache
 }) => {
   const {api_key, user_id, voice_engine, playht_tts_uri, options: credOpts} = credentials;
   const opts = !!options && Object.keys(options).length !== 0 ? options : JSON.parse(credOpts || '{}');
@@ -948,13 +977,13 @@ const synthPlayHT = async(client, logger, {
   if (!JAMBONES_DISABLE_TTS_STREAMING && !renderForCaching && !disableTtsStreaming) {
     let params = '{';
     params += `api_key=${api_key}`;
-    params += `,playback_id=${playbackIdCounter++}`;
+    params += `,playback_id=${key}`;
     params += `,user_id=${user_id}`;
     params += ',vendor=playht';
     params += `,voice=${voice}`;
     params += `,voice_engine=${voice_engine}`;
     params += `,synthesize_url=${synthesizeUrl}`;
-    params += ',write_cache_file=1';
+    params += `,write_cache_file=${disableTtsCache ? 0 : 1}`;
     params += `,language=${language}`;
     if (opts.quality) params += `,quality=${opts.quality}`;
     if (opts.speed) params += `,speed=${opts.speed}`;
@@ -1007,7 +1036,7 @@ const synthPlayHT = async(client, logger, {
 };
 
 const synthInworld = async(logger, {
-  credentials, options, stats, voice, text, renderForCaching, disableTtsStreaming
+  credentials, options, stats, voice, key, text, renderForCaching, disableTtsStreaming, disableTtsCache
 }) => {
   const {api_key, model_id, options: credOpts} = credentials;
   const opts = !!options && Object.keys(options).length !== 0 ? options : JSON.parse(credOpts || '{}');
@@ -1016,11 +1045,11 @@ const synthInworld = async(logger, {
   if (!JAMBONES_DISABLE_TTS_STREAMING && !renderForCaching && !disableTtsStreaming) {
     let params = '{';
     params += `api_key=${api_key}`;
-    params += `,playback_id=${playbackIdCounter++}`;
+    params += `,playback_id=${key}`;
     params += `,model_id=${model_id}`;
     params += ',vendor=inworld';
     params += `,voice=${voice}`;
-    params += ',write_cache_file=1';
+    params += `,write_cache_file=${disableTtsCache ? 0 : 1}`;
     if (opts.temperature) params += `,temperature=${opts.temperature}`;
     if (opts.audioConfig?.pitch) params += `,pitch=${opts.pitch}`;
     if (opts.audioConfig?.speakingRate) params += `,speakingRate=${opts.speakingRate}`;
@@ -1072,7 +1101,7 @@ const synthInworld = async(logger, {
 };
 
 const synthRimelabs = async(logger, {
-  credentials, options, stats, language, voice, text, renderForCaching, disableTtsStreaming
+  credentials, options, stats, language, voice, key, text, renderForCaching, disableTtsStreaming, disableTtsCache
 }) => {
   const {api_key, model_id, options: credOpts} = credentials;
   const opts = !!options && Object.keys(options).length !== 0 ? options : JSON.parse(credOpts || '{}');
@@ -1081,12 +1110,12 @@ const synthRimelabs = async(logger, {
   if (!JAMBONES_DISABLE_TTS_STREAMING && !renderForCaching && !disableTtsStreaming) {
     let params = '{';
     params += `api_key=${api_key}`;
-    params += `,playback_id=${playbackIdCounter++}`;
+    params += `,playback_id=${key}`;
     params += `,model_id=${model_id}`;
     params += ',vendor=rimelabs';
     params += `,language=${language}`;
     params += `,voice=${voice}`;
-    params += ',write_cache_file=1';
+    params += `,write_cache_file=${disableTtsCache ? 0 : 1}`;
     if (opts.speedAlpha) params += `,speed_alpha=${opts.speedAlpha}`;
     if (opts.reduceLatency) params += `,reduce_latency=${opts.reduceLatency}`;
     // Arcana model parameters
@@ -1129,7 +1158,9 @@ const synthRimelabs = async(logger, {
     throw err;
   }
 };
-const synthVerbio = async(client, logger, {credentials, stats, voice, text, renderForCaching, disableTtsStreaming}) => {
+const synthVerbio = async(client, logger, {
+  credentials, stats, voice, key, text, renderForCaching, disableTtsStreaming, disableTtsCache
+}) => {
   //https://doc.speechcenter.verbio.com/#tag/Text-To-Speech-REST-API
   if (text.length > 2000) {
     throw new Error('Verbio cannot synthesize for the text length larger than 2000 characters');
@@ -1138,10 +1169,10 @@ const synthVerbio = async(client, logger, {credentials, stats, voice, text, rend
   if (!process.env.JAMBONES_DISABLE_TTS_STREAMING && !renderForCaching && !disableTtsStreaming) {
     let params = '{';
     params += `access_token=${token.access_token}`;
-    params += `,playback_id=${playbackIdCounter++}`;
+    params += `,playback_id=${key}`;
     params += ',vendor=verbio';
     params += `,voice=${voice}`;
-    params += ',write_cache_file=1';
+    params += `,write_cache_file=${disableTtsCache ? 0 : 1}`;
     params += '}';
 
     return {
@@ -1175,18 +1206,18 @@ const synthVerbio = async(client, logger, {credentials, stats, voice, text, rend
   }
 };
 
-const synthWhisper = async(logger, {credentials, stats, voice, text, instructions,
+const synthWhisper = async(logger, {credentials, stats, voice, key, text, instructions,
-  renderForCaching, disableTtsStreaming}) => {
+  renderForCaching, disableTtsStreaming, disableTtsCache}) => {
   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 (!JAMBONES_DISABLE_TTS_STREAMING && !renderForCaching && !disableTtsStreaming) {
     let params = '{';
     params += `api_key=${api_key}`;
-    params += `,playback_id=${playbackIdCounter++}`;
+    params += `,playback_id=${key}`;
     params += `,model_id=${model_id}`;
     params += ',vendor=whisper';
     params += `,voice=${voice}`;
-    params += ',write_cache_file=1';
+    params += `,write_cache_file=${disableTtsCache ? 0 : 1}`;
     if (speed) params += `,speed=${speed}`;
     // comma is used to separated parameters in freeswitch tts module
     if (instructions) params += `,instructions=${instructions.replace(/\n/g, ' ').replace(/,/g, ';')}`;
@@ -1224,15 +1255,16 @@ const synthWhisper = async(logger, {credentials, stats, voice, text, instruction
   }
 };
 
-const synthDeepgram = async(logger, {credentials, stats, model, text, renderForCaching, disableTtsStreaming}) => {
+const synthDeepgram = async(logger, {credentials, stats, model, key, text, renderForCaching,
+  disableTtsStreaming, disableTtsCache}) => {
   const {api_key, deepgram_tts_uri} = credentials;
   if (!JAMBONES_DISABLE_TTS_STREAMING && !renderForCaching && !disableTtsStreaming) {
     let params = '{';
     params += `api_key=${api_key}`;
-    params += `,playback_id=${playbackIdCounter++}`;
+    params += `,playback_id=${key}`;
     params += ',vendor=deepgram';
     params += `,voice=${model}`;
-    params += ',write_cache_file=1';
+    params += `,write_cache_file=${disableTtsCache ? 0 : 1}`;
     if (deepgram_tts_uri) params += `,endpoint=${deepgram_tts_uri}`;
     params += '}';
 
@@ -1265,7 +1297,7 @@ const synthDeepgram = async(logger, {credentials, stats, model, text, renderForC
 };
 
 const synthCartesia = async(logger, {
-  credentials, options, stats, voice, language, text, renderForCaching, disableTtsStreaming
+  credentials, options, stats, voice, language, key, text, renderForCaching, disableTtsStreaming, disableTtsCache
 }) => {
   const {api_key, model_id, embedding, options: credOpts} = credentials;
   const opts = !!options && Object.keys(options).length !== 0 ? options : JSON.parse(credOpts || '{}');
@@ -1273,11 +1305,11 @@ const synthCartesia = async(logger, {
   if (!JAMBONES_DISABLE_TTS_STREAMING && !renderForCaching && !disableTtsStreaming) {
     let params = '{';
     params += `api_key=${api_key}`;
-    params += `,playback_id=${playbackIdCounter++}`;
+    params += `,playback_id=${key}`;
     params += `,model_id=${model_id}`;
     params += ',vendor=cartesia';
     params += `,voice=${voice}`;
-    params += ',write_cache_file=1';
+    params += `,write_cache_file=${disableTtsCache ? 0 : 1}`;
     params += `,language=${language}`;
     params += `,voice_mode=${embedding ? 'embedding' : 'id'}`;
     if (embedding) params += `,embedding=${embedding}`;
@@ -1336,7 +1368,7 @@ const synthCartesia = async(logger, {
 };
 
 const synthResemble = async(logger, {
-  credentials, options, stats, voice, text, renderForCaching, disableTtsStreaming
+  credentials, options, stats, voice, key, text, renderForCaching, disableTtsStreaming, disableTtsCache
 }) => {
   const {api_key, resemble_tts_uri, resemble_tts_use_tls} = credentials;
   const {project_uuid, use_hd} = options || {};
@@ -1345,10 +1377,10 @@ const synthResemble = async(logger, {
   if (!JAMBONES_DISABLE_TTS_STREAMING && !renderForCaching && !disableTtsStreaming) {
     let params = '{';
     params += `api_key=${api_key}`;
-    params += `,playback_id=${playbackIdCounter++}`;
+    params += `,playback_id=${key}`;
     params += ',vendor=resemble';
     params += `,voice=${voice}`;
-    params += ',write_cache_file=1';
+    params += `,write_cache_file=${disableTtsCache ? 0 : 1}`;
     if (project_uuid) params += `,project_uuid=${project_uuid}`;
     if (use_hd) params += `,use_hd=${use_hd}`;
     if (resemble_tts_uri) params += `,endpoint=${resemble_tts_uri}`;

package-lock.json

@@ -1,12 +1,12 @@
 {
   "name": "@jambonz/speech-utils",
-  "version": "0.2.20",
+  "version": "0.2.24",
   "lockfileVersion": 2,
   "requires": true,
   "packages": {
     "": {
       "name": "@jambonz/speech-utils",
-      "version": "0.2.20",
+      "version": "0.2.24",
       "license": "MIT",
       "dependencies": {
         "23": "^0.0.0",

package.json

@@ -1,6 +1,6 @@
 {
   "name": "@jambonz/speech-utils",
-  "version": "0.2.20",
+  "version": "0.2.24",
   "description": "TTS-related speech utilities for jambonz",
   "main": "index.js",
   "author": "Dave Horton",

test/synth.js

@@ -554,6 +554,7 @@ test('Custom Vendor speech synth tests', async(t) => {
       language: 'en-US',
       voice: 'English-US.Female-1',
       text: 'This is a test.  This is only a test',
+      renderForCaching: true
     });
     t.ok(!opts.servedFromCache, `successfully synthesized custom vendor audio to ${opts.filePath}`);
     t.ok(opts.filePath.endsWith('wav'), 'audio is cached as wav file');
@@ -575,6 +576,7 @@ test('Custom Vendor speech synth tests', async(t) => {
       language: 'en-US',
       voice: 'English-US.Female-1',
       text: 'This is a test.  This is only a test',
+      renderForCaching: true
     });
     t.ok(opts.servedFromCache, `successfully get custom vendor cached audio to ${opts.filePath}`);
     t.ok(opts.filePath.endsWith('wav'), 'audio is cached as wav file');
@@ -589,6 +591,7 @@ test('Custom Vendor speech synth tests', async(t) => {
       language: 'en-US',
       voice: 'English-US.Female-1',
       text: '<speak>This is a test.  This is only a test</speak>',
+      renderForCaching: true
     });
     t.ok(!opts.servedFromCache, `successfully synthesized Custom Vendor audio to ${opts.filePath}`);
     obj = await getJSON(`http://127.0.0.1:3100/lastRequest/somethingnew2`);