@andresaya/edge-tts 1.6.1 → 1.7.0

package/dist/config/constants.js

@@ -40,5 +40,46 @@ exports.Constants = {
         'Sec-Fetch-Site': 'none',
         'Sec-Fetch-Mode': 'cors',
         'Sec-Fetch-Dest': 'empty'
+    },
+    // https://learn.microsoft.com/en-us/azure/ai-services/speech-service/rest-text-to-speech?tabs=nonstreaming
+    OUTPUT_FORMAT: {
+        // Streaming        
+        AMR_WB_16000HZ: 'amr-wb-16000hz',
+        AUDIO_16KHZ_16BIT_32KBPS_MONO_OPUS: 'audio-16khz-16bit-32kbps-mono-opus',
+        AUDIO_16KHZ_32KBITRATE_MONO_MP3: 'audio-16khz-32kbitrate-mono-mp3',
+        AUDIO_16KHZ_64KBITRATE_MONO_MP3: 'audio-16khz-64kbitrate-mono-mp3',
+        AUDIO_16KHZ_128KBITRATE_MONO_MP3: 'audio-16khz-128kbitrate-mono-mp3',
+        AUDIO_24KHZ_16BIT_24KBPS_MONO_OPUS: 'audio-24khz-16bit-24kbps-mono-opus',
+        AUDIO_24KHZ_16BIT_48KBPS_MONO_OPUS: 'audio-24khz-16bit-48kbps-mono-opus',
+        AUDIO_24KHZ_48KBITRATE_MONO_MP3: 'audio-24khz-48kbitrate-mono-mp3',
+        AUDIO_24KHZ_96KBITRATE_MONO_MP3: 'audio-24khz-96kbitrate-mono-mp3',
+        AUDIO_24KHZ_160KBITRATE_MONO_MP3: 'audio-24khz-160kbitrate-mono-mp3',
+        AUDIO_48KHZ_96KBITRATE_MONO_MP3: 'audio-48khz-96kbitrate-mono-mp3',
+        AUDIO_48KHZ_192KBITRATE_MONO_MP3: 'audio-48khz-192kbitrate-mono-mp3',
+        G722_16KHZ_64KBPS: 'g722-16khz-64kbps',
+        OGG_16KHZ_16BIT_MONO_OPUS: 'ogg-16khz-16bit-mono-opus',
+        OGG_24KHZ_16BIT_MONO_OPUS: 'ogg-24khz-16bit-mono-opus',
+        OGG_48KHZ_16BIT_MONO_OPUS: 'ogg-48khz-16bit-mono-opus',
+        RAW_8KHZ_8BIT_MONO_ALAW: 'raw-8khz-8bit-mono-alaw',
+        RAW_8KHZ_8BIT_MONO_MULAW: 'raw-8khz-8bit-mono-mulaw',
+        RAW_8KHZ_16BIT_MONO_PCM: 'raw-8khz-16bit-mono-pcm',
+        RAW_16KHZ_16BIT_MONO_PCM: 'raw-16khz-16bit-mono-pcm',
+        RAW_16KHZ_16BIT_MONO_TRUESILK: 'raw-16khz-16bit-mono-truesilk',
+        RAW_22050HZ_16BIT_MONO_PCM: 'raw-22050hz-16bit-mono-pcm',
+        RAW_24KHZ_16BIT_MONO_PCM: 'raw-24khz-16bit-mono-pcm',
+        RAW_24KHZ_16BIT_MONO_TRUESILK: 'raw-24khz-16bit-mono-truesilk',
+        RAW_44100HZ_16BIT_MONO_PCM: 'raw-44100hz-16bit-mono-pcm',
+        RAW_48KHZ_16BIT_MONO_PCM: 'raw-48khz-16bit-mono-pcm',
+        WEBM_16KHZ_16BIT_MONO_OPUS: 'webm-16khz-16bit-mono-opus',
+        WEBM_24KHZ_16BIT_24KBPS_MONO_OPUS: 'webm-24khz-16bit-24kbps-mono-opus',
+        WEBM_24KHZ_16BIT_MONO_OPUS: 'webm-24khz-16bit-mono-opus',
+        // NonStreaming        
+        RIFF_8KHZ_8BIT_MONO_ALAW: 'riff-8khz-8bit-mono-alaw',
+        RIFF_8KHZ_8BIT_MONO_MULAW: 'riff-8khz-8bit-mono-mulaw',
+        RIFF_8KHZ_16BIT_MONO_PCM: 'riff-8khz-16bit-mono-pcm',
+        RIFF_22050HZ_16BIT_MONO_PCM: 'riff-22050hz-16bit-mono-pcm',
+        RIFF_24KHZ_16BIT_MONO_PCM: 'riff-24khz-16bit-mono-pcm',
+        RIFF_44100HZ_16BIT_MONO_PCM: 'riff-44100hz-16bit-mono-pcm',
+        RIFF_48KHZ_16BIT_MONO_PCM: 'riff-48khz-16bit-mono-pcm'
     }
 };

package/dist/services/EdgeTTS.d.ts

@@ -12,6 +12,7 @@ export interface SynthesisOptions {
     rate?: string | number;
     volume?: string | number;
     inputType?: 'auto' | 'ssml' | 'text';
+    outputFormat?: string;
 }
 export interface WordBoundary {
     type: "WordBoundary";
@@ -22,6 +23,7 @@ export interface WordBoundary {
 export declare class EdgeTTS {
     private audio_stream;
     private audio_format;
+    private output_format;
     private word_boundaries;
     private ws;
     getVoices(): Promise<Voice[]>;
@@ -43,6 +45,7 @@ export declare class EdgeTTS {
     private parseMetadata;
     generateSecMsGec: (trustedClientToken: string) => Promise<string>;
     getDuration(): number;
+    private getFileExtension;
     getAudioInfo(): {
         size: number;
         format: string;

package/dist/services/EdgeTTS.js

@@ -27,6 +27,7 @@ function ensureBuffer(data) {
 class EdgeTTS {
     audio_stream = [];
     audio_format = 'mp3';
+    output_format = 'audio-24khz-48kbitrate-mono-mp3';
     word_boundaries = [];
     ws;
     async getVoices() {
@@ -114,38 +115,49 @@ class EdgeTTS {
             this.audio_stream = [];
             const reqId = this.generateUUID();
             const url = `${constants_1.Constants.WSS_URL}?Ocp-Apim-Subscription-Key=${constants_1.Constants.TRUSTED_CLIENT_TOKEN}&Sec-MS-GEC=${secMsGEC}&Sec-MS-GEC-Version=${constants_1.Constants.VERSION_MS_GEC}&ConnectionId=${reqId}`;
-            console.log('WebSocket URL:', url);
             this.ws = new ws_1.default(url, {
                 headers: constants_1.Constants.getBaseHeaders(),
                 rejectUnauthorized: false
             });
             const SSML_text = this.getSSML(text, voice, options);
-            const timeout = setTimeout(() => {
-                if (this.ws && this.ws.readyState === ws_1.default.OPEN) {
-                    this.ws.close();
-                }
-                reject(new Error("Synthesis timeout"));
-            }, 30000);
+            const outputFormat = options.outputFormat || 'audio-24khz-48kbitrate-mono-mp3';
+            this.output_format = outputFormat;
+            let timedOut = false;
+            let inactivityTimeout;
+            const resetInactivityTimeout = () => {
+                clearTimeout(inactivityTimeout);
+                inactivityTimeout = setTimeout(() => {
+                    timedOut = true;
+                    if (this.ws && this.ws.readyState === ws_1.default.OPEN) {
+                        this.ws.close();
+                    }
+                    reject(new Error("WebSocket inactivity timeout - no response from server"));
+                }, 30000); // 30 seconds of inactivity
+            };
             this.ws.on('open', () => {
-                const message = this.buildTTSConfigMessage();
+                resetInactivityTimeout(); // start the inactivity timeout
+                const message = this.buildTTSConfigMessage(outputFormat);
                 this.ws.send(message);
                 const timestamp = this.nowRFC1123();
                 const speechMessage = `X-RequestId:${reqId}\r\nContent-Type:application/ssml+xml\r\nX-Timestamp:${timestamp}\r\nPath:ssml\r\n\r\n${SSML_text}`;
                 this.ws.send(speechMessage);
             });
             this.ws.on('message', (data) => {
+                resetInactivityTimeout(); // restart inactivity timeout
                 this.processAudioData(data);
             });
             this.ws.on('error', (err) => {
-                clearTimeout(timeout);
+                clearTimeout(inactivityTimeout);
                 if (this.ws && this.ws.readyState === ws_1.default.OPEN) {
                     this.ws.close();
                 }
                 reject(err);
             });
             this.ws.on('close', () => {
-                clearTimeout(timeout);
-                resolve();
+                clearTimeout(inactivityTimeout);
+                if (!timedOut) {
+                    resolve();
+                }
             });
         });
     }
@@ -245,10 +257,10 @@ class EdgeTTS {
     parseRFC1123(rfcStr) {
         return new Date(rfcStr);
     }
-    buildTTSConfigMessage() {
+    buildTTSConfigMessage(outputFormat = 'audio-24khz-48kbitrate-mono-mp3') {
         const timestamp = this.nowRFC1123();
         return `X-Timestamp:${timestamp}\r\nContent-Type:application/json; charset=utf-8\r\nPath:speech.config\r\n\r\n` +
-            `{"context":{"synthesis":{"audio":{"metadataoptions":{"sentenceBoundaryEnabled":false,"wordBoundaryEnabled":true},"outputFormat":"audio-24khz-48kbitrate-mono-mp3"}}}}`;
+            `{"context":{"synthesis":{"audio":{"metadataoptions":{"sentenceBoundaryEnabled":false,"wordBoundaryEnabled":true},"outputFormat":"${outputFormat}"}}}}`;
     }
     async *synthesizeStream(text, voice = 'en-US-AnaNeural', options = {}) {
         this.audio_stream = [];
@@ -260,6 +272,8 @@ class EdgeTTS {
             rejectUnauthorized: false
         });
         const SSML_text = this.getSSML(text, voice, options);
+        const outputFormat = options.outputFormat || 'audio-24khz-48kbitrate-mono-mp3';
+        this.output_format = outputFormat;
         const queue = [];
         let done = false;
         let error = null;
@@ -271,19 +285,33 @@ class EdgeTTS {
                 notify = null;
             }
         };
-        const timeout = setTimeout(() => {
-            if (this.ws && this.ws.readyState === ws_1.default.OPEN) {
-                this.ws.close();
-            }
-        }, 30000);
+        let timedOut = false;
+        let inactivityTimeout;
+        const resetInactivityTimeout = () => {
+            clearTimeout(inactivityTimeout);
+            inactivityTimeout = setTimeout(() => {
+                timedOut = true;
+                error = new Error("WebSocket inactivity timeout - no response from server");
+                done = true;
+                if (this.ws && this.ws.readyState === ws_1.default.OPEN) {
+                    this.ws.close();
+                }
+                if (notify) {
+                    notify();
+                    notify = null;
+                }
+            }, 30000); // 30 seconds of inactivity
+        };
         this.ws.on('open', () => {
-            const message = this.buildTTSConfigMessage();
+            resetInactivityTimeout(); // start the inactivity timeout
+            const message = this.buildTTSConfigMessage(outputFormat);
             this.ws.send(message);
             const timestamp = this.nowRFC1123();
             const speechMessage = `X-RequestId:${reqId}\r\nContent-Type:application/ssml+xml\r\nX-Timestamp:${timestamp}\r\nPath:ssml\r\n\r\n${SSML_text}`;
             this.ws.send(speechMessage);
         });
         this.ws.on('message', (data) => {
+            resetInactivityTimeout(); // restart inactivity timeout
             const buffer = ensureBuffer(data);
             const needle = buffer_1.Buffer.from('Path:audio\r\n');
             const audioStartIndex = buffer.indexOf(new Uint8Array(needle));
@@ -307,6 +335,7 @@ class EdgeTTS {
             }
         });
         this.ws.on('error', (err) => {
+            clearTimeout(inactivityTimeout);
             error = err;
             done = true;
             if (notify) {
@@ -315,7 +344,7 @@ class EdgeTTS {
             }
         });
         this.ws.on('close', () => {
-            clearTimeout(timeout);
+            clearTimeout(inactivityTimeout);
             done = true;
             if (notify) {
                 notify();
@@ -405,17 +434,41 @@ class EdgeTTS {
         const estimatedDuration = bufferSize / (24000 * 3); // 24000 Hz sample rate, 3 bytes per sample (16-bit stereo)
         return estimatedDuration;
     }
+    getFileExtension(format) {
+        if (format.includes('mp3'))
+            return 'mp3';
+        if (format.includes('opus') && format.includes('webm'))
+            return 'webm';
+        if (format.includes('opus') && format.includes('ogg'))
+            return 'ogg';
+        if (format.includes('wav') || format.includes('riff'))
+            return 'wav';
+        if (format.includes('pcm') && format.includes('raw'))
+            return 'pcm';
+        if (format.includes('alaw'))
+            return 'alaw';
+        if (format.includes('mulaw'))
+            return 'mulaw';
+        if (format.includes('truesilk'))
+            return 'silk';
+        if (format.includes('g722'))
+            return 'g722';
+        if (format.includes('amr'))
+            return 'amr';
+        return 'audio';
+    }
     getAudioInfo() {
         const buffer = this.toBuffer();
         return {
             size: buffer.length,
-            format: this.audio_format,
+            format: this.getFileExtension(this.output_format),
             estimatedDuration: this.getDuration()
         };
     }
-    async toFile(outputPath, format = this.audio_format) {
-        if (!format || typeof format !== 'string')
-            format = this.audio_format;
+    async toFile(outputPath, format) {
+        if (!format) {
+            format = this.getFileExtension(this.output_format);
+        }
         const audioBuffer = this.toBuffer();
         const finalPath = `${outputPath}.${format}`;
         await (0, promises_1.writeFile)(finalPath, new Uint8Array(audioBuffer));

package/package.json

@@ -1,54 +1,54 @@
-{
-  "name": "@andresaya/edge-tts",
-  "version": "1.6.1",
-  "description": "Edge TTS is a package that allows access to the online text-to-speech service used by Microsoft Edge without the need for Microsoft Edge, Windows, or an API key.",
-  "main": "dist/index.js",
-  "browser": "dist/browser/edge-tts.esm.js",
-  "unpkg": "dist/browser/edge-tts.umd.min.js",
-  "jsdelivr": "dist/browser/edge-tts.umd.min.js",
-  "keywords": [
-    "edge",
-    "tts",
-    "text-to-speech",
-    "microsoft",
-    "microsoft edge",
-    "speech-synthesis"
-  ],
-  "author": "Andres Aya",
-  "license": "GPL-3.0-only",
-  "homepage": "https://github.com/andresayac/edge-tts",
-  "bugs": {
-    "url": "https://github.com/andresayac/edge-tts/issues"
-  },
-  "scripts": {
-    "build": "tsc",
-    "build:bin": "bun scripts/build.ts",
-    "build:browser": "bun scripts/build-browser.ts",
-    "prepublishOnly": "npm run build && npm run build:bin && npm run build:browser"
-  },
-  "devDependencies": {
-    "@types/bun": "latest",
-    "@types/ws": "^8.18.1",
-    "esbuild": "^0.25.12",
-    "typescript": "^5.6.2"
-  },
-  "dependencies": {
-    "@types/uuid": "^10.0.0",
-    "commander": "^12.1.0",
-    "events": "^3.3.0",
-    "fs": "^0.0.1-security",
-    "ws": "^8.18.3"
-  },
-  "peerDependencies": {
-    "typescript": "^5.0.0"
-  },
-  "files": [
-    "dist"
-  ],
-  "bin": {
-    "edge-tts": "./dist/cli/edge-tts.js"
-  },
-  "engines": {
-    "bun": ">=0.5.0"
-  }
-}
+{
+  "name": "@andresaya/edge-tts",
+  "version": "1.7.0",
+  "description": "Edge TTS is a package that allows access to the online text-to-speech service used by Microsoft Edge without the need for Microsoft Edge, Windows, or an API key.",
+  "main": "dist/index.js",
+  "browser": "dist/browser/edge-tts.esm.js",
+  "unpkg": "dist/browser/edge-tts.umd.min.js",
+  "jsdelivr": "dist/browser/edge-tts.umd.min.js",
+  "keywords": [
+    "edge",
+    "tts",
+    "text-to-speech",
+    "microsoft",
+    "microsoft edge",
+    "speech-synthesis"
+  ],
+  "author": "Andres Aya",
+  "license": "GPL-3.0-only",
+  "homepage": "https://github.com/andresayac/edge-tts",
+  "bugs": {
+    "url": "https://github.com/andresayac/edge-tts/issues"
+  },
+  "scripts": {
+    "build": "tsc",
+    "build:bin": "bun scripts/build.ts",
+    "build:browser": "bun scripts/build-browser.ts",
+    "prepublishOnly": "npm run build && npm run build:bin && npm run build:browser"
+  },
+  "devDependencies": {
+    "@types/bun": "latest",
+    "@types/ws": "^8.18.1",
+    "esbuild": "^0.25.12",
+    "typescript": "^5.6.2"
+  },
+  "dependencies": {
+    "@types/uuid": "^10.0.0",
+    "commander": "^12.1.0",
+    "events": "^3.3.0",
+    "fs": "^0.0.1-security",
+    "ws": "^8.18.3"
+  },
+  "peerDependencies": {
+    "typescript": "^5.0.0"
+  },
+  "files": [
+    "dist"
+  ],
+  "bin": {
+    "edge-tts": "./dist/cli/edge-tts.js"
+  },
+  "engines": {
+    "bun": ">=0.5.0"
+  }
+}

package/dist/browser.d.ts

@@ -1 +0,0 @@
-export { EdgeTTS, type Voice, type SynthesisOptions, type WordBoundary } from './browser/EdgeTTS.browser';

package/dist/browser.js

@@ -1,5 +0,0 @@
-"use strict";
-Object.defineProperty(exports, "__esModule", { value: true });
-exports.EdgeTTS = void 0;
-var EdgeTTS_browser_1 = require("./browser/EdgeTTS.browser");
-Object.defineProperty(exports, "EdgeTTS", { enumerable: true, get: function () { return EdgeTTS_browser_1.EdgeTTS; } });