@fonoster/apiserver 0.15.21 → 0.16.1

package/dist/voice/{VoiceClientImpl.d.ts → client/VoiceClientImpl.d.ts}

@@ -17,37 +17,34 @@
  * limitations under the License.
  */
 import { Stream } from "stream";
-import { SayOptions, VoiceClientConfig, VoiceIn, VoiceSessionStreamClient } from "@fonoster/common";
-import { AudioSocket } from "@fonoster/streams";
+import { SayOptions, VoiceClientConfig, VoiceIn } from "@fonoster/common";
 import { Bridge, Client } from "ari-client";
-import { SpeechResult } from "./stt/types";
-import { GRPCClient, SpeechToText, TextToSpeech, VoiceClient } from "./types";
+import { SpeechResult } from "../stt/types";
+import { SpeechToText, TextToSpeech, VoiceClient } from "../types";
 declare class VoiceClientImpl implements VoiceClient {
     config: VoiceClientConfig;
     verbsStream: Stream;
     transcriptionsStream: Stream;
-    voice: VoiceSessionStreamClient;
-    tts: TextToSpeech;
-    stt: SpeechToText;
-    grpcClient: GRPCClient;
-    audioSocket: AudioSocket;
-    asStream: Stream;
     ari: Client;
     bridge: Bridge;
-    filesServer: any;
+    private authHandler;
+    private audioSocketHandler;
+    private externalMediaHandler;
+    private grpcHandler;
+    private speechHandler;
+    private _tts;
+    private _stt;
     constructor(params: {
         ari: Client;
         config: VoiceClientConfig;
         tts: TextToSpeech;
         stt: SpeechToText;
-    }, filesServer: any);
+    });
     connect(): Promise<void>;
-    setupAudioSocket(port: number): void;
-    on(type: string, callback: (data: VoiceIn) => void): void;
     sendResponse(response: VoiceIn): void;
-    getTranscriptionsStream(): Stream;
-    setupExternalMedia(port: number): Promise<void>;
+    on(type: string, callback: (data: VoiceIn) => void): void;
     synthesize(text: string, options: SayOptions): Promise<string>;
+    stopSynthesis(): Promise<void>;
     transcribe(): Promise<SpeechResult>;
     startSpeechGather(callback: (stream: {
         speech: string;
@@ -56,7 +53,6 @@ declare class VoiceClientImpl implements VoiceClient {
     startDtmfGather(sessionRef: string, callback: (event: {
         digit: string;
     }) => void): Promise<void>;
-    stopStreamGather(): Promise<void>;
     waitForDtmf(params: {
         sessionRef: string;
         finishOnKey: string;
@@ -66,6 +62,10 @@ declare class VoiceClientImpl implements VoiceClient {
     }): Promise<{
         digits: string;
     }>;
+    stopStreamGather(): void;
+    getTranscriptionsStream(): Stream;
     close(): void;
+    get tts(): TextToSpeech;
+    get stt(): SpeechToText;
 }
 export { VoiceClientImpl };

package/dist/voice/client/VoiceClientImpl.js

@@ -0,0 +1,158 @@
+"use strict";
+var __awaiter = (this && this.__awaiter) || function (thisArg, _arguments, P, generator) {
+    function adopt(value) { return value instanceof P ? value : new P(function (resolve) { resolve(value); }); }
+    return new (P || (P = Promise))(function (resolve, reject) {
+        function fulfilled(value) { try { step(generator.next(value)); } catch (e) { reject(e); } }
+        function rejected(value) { try { step(generator["throw"](value)); } catch (e) { reject(e); } }
+        function step(result) { result.done ? resolve(result.value) : adopt(result.value).then(fulfilled, rejected); }
+        step((generator = generator.apply(thisArg, _arguments || [])).next());
+    });
+};
+Object.defineProperty(exports, "__esModule", { value: true });
+exports.VoiceClientImpl = void 0;
+/**
+ * Copyright (C) 2025 by Fonoster Inc (https://fonoster.com)
+ * http://github.com/fonoster/fonoster
+ *
+ * This file is part of Fonoster
+ *
+ * Licensed under the MIT License (the "License");
+ * you may not use this file except in compliance with
+ * the License. You may obtain a copy of the License at
+ *
+ *    https://opensource.org/licenses/MIT
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+const stream_1 = require("stream");
+const logger_1 = require("@fonoster/logger");
+const pick_port_1 = require("pick-port");
+const AudioSocketHandler_1 = require("./AudioSocketHandler");
+const AuthorizationHandler_1 = require("./AuthorizationHandler");
+const ExternalMediaHandler_1 = require("./ExternalMediaHandler");
+const GrpcClientHandler_1 = require("./GrpcClientHandler");
+const SpeechHandler_1 = require("./SpeechHandler");
+const logger = (0, logger_1.getLogger)({ service: "apiserver", filePath: __filename });
+class VoiceClientImpl {
+    constructor(params) {
+        const { config, tts, stt, ari } = params;
+        this.config = config;
+        this.ari = ari;
+        this._tts = tts;
+        this._stt = stt;
+        this.verbsStream = new stream_1.Stream();
+        this.transcriptionsStream = new stream_1.Stream();
+        // Initialize component handlers
+        this.authHandler = new AuthorizationHandler_1.AuthorizationHandler({
+            config: this.config,
+            ari: this.ari
+        });
+        this.audioSocketHandler = new AudioSocketHandler_1.AudioSocketHandler({
+            transcriptionsStream: this.transcriptionsStream,
+            config: this.config
+        });
+        this.externalMediaHandler = new ExternalMediaHandler_1.ExternalMediaHandler({
+            ari: this.ari,
+            config: this.config
+        });
+        this.grpcHandler = new GrpcClientHandler_1.GrpcClientHandler({
+            config: this.config,
+            verbsStream: this.verbsStream
+        });
+    }
+    connect() {
+        return __awaiter(this, void 0, void 0, function* () {
+            // Check authorization
+            const isAuthorized = yield this.authHandler.checkAuthorization();
+            if (!isAuthorized) {
+                return;
+            }
+            // Set up GRPC client
+            yield this.grpcHandler.setupGrpcClient();
+            // Set up audio socket and external media
+            const externalMediaPort = yield (0, pick_port_1.pickPort)({ type: "tcp" });
+            logger.verbose("picked external media port", { port: externalMediaPort });
+            // Wait for both audio socket and external media setup to complete
+            yield Promise.all([
+                this.audioSocketHandler.setupAudioSocket(externalMediaPort),
+                this.externalMediaHandler.setupExternalMedia(externalMediaPort)
+            ]);
+            // Get the bridge from the external media handler
+            this.bridge = this.externalMediaHandler.getBridge();
+            // Initialize speech handler now that we have the audio stream
+            this.speechHandler = new SpeechHandler_1.SpeechHandler({
+                tts: this._tts,
+                stt: this._stt,
+                ari: this.ari,
+                transcriptionsStream: this.transcriptionsStream,
+                audioStream: this.audioSocketHandler.getAudioStream(),
+                sessionRef: this.config.sessionRef
+            });
+            logger.verbose("voice client setup completed");
+        });
+    }
+    // Public API methods required by VoiceClient interface
+    sendResponse(response) {
+        this.grpcHandler.sendResponse(response);
+    }
+    on(type, callback) {
+        this.verbsStream.on(type.toString(), (data) => {
+            callback(data[type]);
+        });
+    }
+    synthesize(text, options) {
+        return __awaiter(this, void 0, void 0, function* () {
+            return this.speechHandler.synthesize(text, options);
+        });
+    }
+    stopSynthesis() {
+        return __awaiter(this, void 0, void 0, function* () {
+            return this.speechHandler.stopSynthesis();
+        });
+    }
+    transcribe() {
+        return __awaiter(this, void 0, void 0, function* () {
+            return this.speechHandler.transcribe();
+        });
+    }
+    startSpeechGather(callback) {
+        this.speechHandler.startSpeechGather(callback);
+    }
+    startDtmfGather(sessionRef, callback) {
+        return __awaiter(this, void 0, void 0, function* () {
+            return this.speechHandler.startDtmfGather(callback);
+        });
+    }
+    waitForDtmf(params) {
+        return __awaiter(this, void 0, void 0, function* () {
+            const { finishOnKey, maxDigits, timeout, onDigitReceived } = params;
+            return this.speechHandler.waitForDtmf({
+                finishOnKey,
+                maxDigits,
+                timeout,
+                onDigitReceived
+            });
+        });
+    }
+    stopStreamGather() {
+        this.speechHandler.stopStreamGather();
+    }
+    getTranscriptionsStream() {
+        return this.transcriptionsStream;
+    }
+    close() {
+        this.grpcHandler.close();
+        this.audioSocketHandler.close();
+    }
+    get tts() {
+        return this._tts;
+    }
+    get stt() {
+        return this._stt;
+    }
+}
+exports.VoiceClientImpl = VoiceClientImpl;

package/dist/voice/client/index.d.ts

@@ -0,0 +1,24 @@
+/**
+ * Copyright (C) 2025 by Fonoster Inc (https://fonoster.com)
+ * http://github.com/fonoster/fonoster
+ *
+ * This file is part of Fonoster
+ *
+ * Licensed under the MIT License (the "License");
+ * you may not use this file except in compliance with
+ * the License. You may obtain a copy of the License at
+ *
+ *    https://opensource.org/licenses/MIT
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+export * from "./VoiceClientImpl";
+export * from "./AudioSocketHandler";
+export * from "./ExternalMediaHandler";
+export * from "./AuthorizationHandler";
+export * from "./GrpcClientHandler";
+export * from "./SpeechHandler";

package/dist/voice/client/index.js

@@ -0,0 +1,40 @@
+"use strict";
+var __createBinding = (this && this.__createBinding) || (Object.create ? (function(o, m, k, k2) {
+    if (k2 === undefined) k2 = k;
+    var desc = Object.getOwnPropertyDescriptor(m, k);
+    if (!desc || ("get" in desc ? !m.__esModule : desc.writable || desc.configurable)) {
+      desc = { enumerable: true, get: function() { return m[k]; } };
+    }
+    Object.defineProperty(o, k2, desc);
+}) : (function(o, m, k, k2) {
+    if (k2 === undefined) k2 = k;
+    o[k2] = m[k];
+}));
+var __exportStar = (this && this.__exportStar) || function(m, exports) {
+    for (var p in m) if (p !== "default" && !Object.prototype.hasOwnProperty.call(exports, p)) __createBinding(exports, m, p);
+};
+Object.defineProperty(exports, "__esModule", { value: true });
+/**
+ * Copyright (C) 2025 by Fonoster Inc (https://fonoster.com)
+ * http://github.com/fonoster/fonoster
+ *
+ * This file is part of Fonoster
+ *
+ * Licensed under the MIT License (the "License");
+ * you may not use this file except in compliance with
+ * the License. You may obtain a copy of the License at
+ *
+ *    https://opensource.org/licenses/MIT
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+__exportStar(require("./VoiceClientImpl"), exports);
+__exportStar(require("./AudioSocketHandler"), exports);
+__exportStar(require("./ExternalMediaHandler"), exports);
+__exportStar(require("./AuthorizationHandler"), exports);
+__exportStar(require("./GrpcClientHandler"), exports);
+__exportStar(require("./SpeechHandler"), exports);

package/dist/voice/connectToAri.js

@@ -64,7 +64,7 @@ function connectToAri(filesServer) {
             logger.info("asterisk is ready");
             const createContainer = (0, integrations_1.createCreateContainer)(db_1.prisma, envs_1.INTEGRATIONS_FILE);
             const nats = yield (0, nats_1.connect)({ servers: envs_1.NATS_URL, maxReconnectAttempts: -1 });
-            const dispatcher = new VoiceDispatcher_1.VoiceDispatcher(ari, nats, (0, createCreateVoiceClient_1.createCreateVoiceClient)(createContainer, filesServer));
+            const dispatcher = new VoiceDispatcher_1.VoiceDispatcher(ari, nats, (0, createCreateVoiceClient_1.createCreateVoiceClient)(createContainer));
             dispatcher.start();
         }
         else {

package/dist/voice/createCreateVoiceClient.d.ts

@@ -1,7 +1,7 @@
 import { Channel, Client, StasisStart } from "ari-client";
 import { CreateContainer } from "./integrations/types";
 import { VoiceClient } from "./types";
-declare function createCreateVoiceClient(createContainer: CreateContainer, filesServer: any): (params: {
+declare function createCreateVoiceClient(createContainer: CreateContainer): (params: {
     ari: Client;
     event: StasisStart;
     channel: Channel;

package/dist/voice/createCreateVoiceClient.js

@@ -32,13 +32,13 @@ const identity_1 = require("@fonoster/identity");
 const logger_1 = require("@fonoster/logger");
 const identityConfig_1 = require("../core/identityConfig");
 const mapCallDirectionToEnum_1 = require("../events/mapCallDirectionToEnum");
+const client_1 = require("./client");
 const types_1 = require("./types");
 const createGetChannelVarWithoutThrow_1 = require("./utils/createGetChannelVarWithoutThrow");
-const VoiceClientImpl_1 = require("./VoiceClientImpl");
 const logger = (0, logger_1.getLogger)({ service: "apiserver", filePath: __filename });
 const generateCallAccessToken = (0, identity_1.createGenerateCallAccessToken)(identityConfig_1.identityConfig);
 // Note: By the time the call arrives here the owner of the app MUST be authenticated
-function createCreateVoiceClient(createContainer, filesServer) {
+function createCreateVoiceClient(createContainer) {
     return function createVoiceClient(params) {
         return __awaiter(this, void 0, void 0, function* () {
             var _a, _b, _c, _d, _e;
@@ -70,7 +70,7 @@ function createCreateVoiceClient(createContainer, filesServer) {
                 callerNumber,
                 ingressNumber
             });
-            return new VoiceClientImpl_1.VoiceClientImpl({ ari, config, tts, stt }, filesServer);
+            return new client_1.VoiceClientImpl({ ari, config, tts, stt });
         });
     };
 }

package/dist/voice/handlers/createSayHandler.js

@@ -10,34 +10,21 @@ var __awaiter = (this && this.__awaiter) || function (thisArg, _arguments, P, ge
 };
 Object.defineProperty(exports, "__esModule", { value: true });
 exports.createSayHandler = createSayHandler;
-const nanoid_1 = require("nanoid");
 const pb_util_1 = require("pb-util");
 const zod_1 = require("zod");
-const envs_1 = require("../../envs");
-const awaitForPlaybackFinished_1 = require("./utils/awaitForPlaybackFinished");
 const withErrorHandling_1 = require("./utils/withErrorHandling");
 const sayRequestSchema = zod_1.z.object({
     text: zod_1.z.string(),
     sessionRef: zod_1.z.string(),
-    playbackRef: zod_1.z.string().optional(),
     options: zod_1.z.record(zod_1.z.unknown()).optional()
 });
-const getMediaUrl = (filename) => `sound:http://${envs_1.APISERVER_HOST}:${envs_1.HTTP_BRIDGE_PORT}/api/sounds/${filename}.sln16`;
 function createSayHandler(ari, voiceClient) {
     return (0, withErrorHandling_1.withErrorHandling)((request) => __awaiter(this, void 0, void 0, function* () {
-        const { sessionRef: channelId } = request;
         sayRequestSchema.parse(request);
-        const playbackRef = request.playbackRef || (0, nanoid_1.nanoid)(10);
-        const mediaId = yield voiceClient.synthesize(request.text, request.options ? pb_util_1.struct.decode(request.options) : {});
-        yield ari.channels.play({
-            channelId,
-            media: getMediaUrl(mediaId),
-            playbackId: playbackRef
-        });
-        yield (0, awaitForPlaybackFinished_1.awaitForPlaybackFinished)(ari, playbackRef);
+        yield voiceClient.synthesize(request.text, request.options ? pb_util_1.struct.decode(request.options) : {});
         voiceClient.sendResponse({
             sayResponse: {
-                playbackRef
+                sessionRef: request.sessionRef
             }
         });
     }));

package/dist/voice/handlers/createStopSayHandler.d.ts

@@ -0,0 +1,3 @@
+import { VoiceClient } from "../types";
+declare function createStopSayHandler(voiceClient: VoiceClient): (request: import("@fonoster/common").VerbRequest) => Promise<void>;
+export { createStopSayHandler };

package/dist/voice/handlers/createStopSayHandler.js

@@ -0,0 +1,34 @@
+"use strict";
+var __awaiter = (this && this.__awaiter) || function (thisArg, _arguments, P, generator) {
+    function adopt(value) { return value instanceof P ? value : new P(function (resolve) { resolve(value); }); }
+    return new (P || (P = Promise))(function (resolve, reject) {
+        function fulfilled(value) { try { step(generator.next(value)); } catch (e) { reject(e); } }
+        function rejected(value) { try { step(generator["throw"](value)); } catch (e) { reject(e); } }
+        function step(result) { result.done ? resolve(result.value) : adopt(result.value).then(fulfilled, rejected); }
+        step((generator = generator.apply(thisArg, _arguments || [])).next());
+    });
+};
+Object.defineProperty(exports, "__esModule", { value: true });
+exports.createStopSayHandler = createStopSayHandler;
+const zod_1 = require("zod");
+const withErrorHandling_1 = require("./utils/withErrorHandling");
+const requestSchema = zod_1.z.object({
+    sessionRef: zod_1.z.string()
+});
+function createStopSayHandler(voiceClient) {
+    return (0, withErrorHandling_1.withErrorHandling)((stopSayReq) => __awaiter(this, void 0, void 0, function* () {
+        requestSchema.parse(stopSayReq);
+        const { sessionRef } = stopSayReq;
+        try {
+            voiceClient.stopSynthesis();
+        }
+        catch (err) {
+            // We can only try
+        }
+        voiceClient.sendResponse({
+            stopSayResponse: {
+                sessionRef
+            }
+        });
+    }));
+}

package/dist/voice/handlers/index.d.ts

@@ -29,3 +29,4 @@ export * from "./createStreamGatherHandler";
 export * from "./createUnmuteHandler";
 export * from "./dial/createDialHandler";
 export * from "./gather/createGatherHandler";
+export * from "./createStopSayHandler";

package/dist/voice/handlers/index.js

@@ -45,3 +45,4 @@ __exportStar(require("./createStreamGatherHandler"), exports);
 __exportStar(require("./createUnmuteHandler"), exports);
 __exportStar(require("./dial/createDialHandler"), exports);
 __exportStar(require("./gather/createGatherHandler"), exports);
+__exportStar(require("./createStopSayHandler"), exports);

package/dist/voice/tts/Azure.js

@@ -67,7 +67,7 @@ class Azure extends AbstractTextToSpeech_1.AbstractTextToSpeech {
         const speechConfig = sdk.SpeechConfig.fromSubscription(subscriptionKey, serviceRegion);
         speechConfig.speechSynthesisVoiceName = voice;
         speechConfig.speechSynthesisOutputFormat =
-            sdk.SpeechSynthesisOutputFormat.Riff16Khz16BitMonoPcm;
+            sdk.SpeechSynthesisOutputFormat.Raw8Khz16BitMonoPcm;
         const stream = (0, createChunkedSynthesisStream_1.createChunkedSynthesisStream)(text, (chunkText) => __awaiter(this, void 0, void 0, function* () {
             const synthesizer = new sdk.SpeechSynthesizer(speechConfig);
             const isSSML = (0, isSsml_1.isSsml)(chunkText);

package/dist/voice/tts/Deepgram.d.ts

@@ -29,7 +29,7 @@ declare class Deepgram extends AbstractTextToSpeech<typeof ENGINE_NAME> {
     protected readonly OUTPUT_FORMAT = "sln16";
     protected readonly CACHING_FIELDS: string[];
     protected readonly AUDIO_ENCODING: "linear16";
-    protected readonly SAMPLE_RATE_HERTZ = 16000;
+    protected readonly SAMPLE_RATE_HERTZ = 8000;
     constructor(config: DeepgramTtsConfig);
     synthesize(text: string, options: SynthOptions): {
         ref: string;

package/dist/voice/tts/Deepgram.js

@@ -58,7 +58,7 @@ class Deepgram extends AbstractTextToSpeech_1.AbstractTextToSpeech {
         this.OUTPUT_FORMAT = "sln16";
         this.CACHING_FIELDS = ["voice"];
         this.AUDIO_ENCODING = "linear16";
-        this.SAMPLE_RATE_HERTZ = 16000;
+        this.SAMPLE_RATE_HERTZ = 8000;
         this.client = (0, sdk_1.createClient)(config.credentials.apiKey);
         this.engineConfig = config;
     }

package/dist/voice/tts/ElevenLabs.js

@@ -43,18 +43,40 @@ var __awaiter = (this && this.__awaiter) || function (thisArg, _arguments, P, ge
 };
 Object.defineProperty(exports, "__esModule", { value: true });
 exports.ElevenLabs = exports.ENGINE_NAME = void 0;
+/**
+ * Copyright (C) 2025 by Fonoster Inc (https://fonoster.com)
+ * http://github.com/fonoster/fonoster
+ *
+ * This file is part of Fonoster
+ *
+ * Licensed under the MIT License (the "License");
+ * you may not use this file except in compliance with
+ * the License. You may obtain a copy of the License at
+ *
+ *    https://opensource.org/licenses/MIT
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+const stream_1 = require("stream");
+const logger_1 = require("@fonoster/logger");
 const elevenlabs_1 = require("elevenlabs");
 const z = __importStar(require("zod"));
 const AbstractTextToSpeech_1 = require("./AbstractTextToSpeech");
+const convertUlawToPCM16_1 = require("./utils/convertUlawToPCM16");
 const createChunkedSynthesisStream_1 = require("./utils/createChunkedSynthesisStream");
 const streamToBuffer_1 = require("./utils/streamToBuffer");
+const logger = (0, logger_1.getLogger)({ service: "apiserver", filePath: __filename });
 const ENGINE_NAME = "tts.elevenlabs";
 exports.ENGINE_NAME = ENGINE_NAME;
 class ElevenLabs extends AbstractTextToSpeech_1.AbstractTextToSpeech {
     constructor(config) {
         super();
         this.engineName = ENGINE_NAME;
-        this.OUTPUT_FORMAT = "sln16";
+        this.OUTPUT_FORMAT = "sln16"; // TODO: Ask the team at ElevenLabs to provde PCM 16-bit at 8kHz
         this.CACHING_FIELDS = ["voice", "text"];
         this.client = new elevenlabs_1.ElevenLabsClient(config.credentials);
         this.engineConfig = config;
@@ -64,18 +86,28 @@ class ElevenLabs extends AbstractTextToSpeech_1.AbstractTextToSpeech {
         const { voice, model } = this.engineConfig.config;
         const ref = this.createMediaReference();
         const stream = (0, createChunkedSynthesisStream_1.createChunkedSynthesisStream)(text, (chunkText) => __awaiter(this, void 0, void 0, function* () {
-            const response = yield this.client.generate({
-                stream: true,
-                voice,
-                text: chunkText,
-                model_id: model !== null && model !== void 0 ? model : "eleven_flash_v2_5",
-                output_format: "pcm_16000",
-                // TODO: Make this configurable
-                optimize_streaming_latency: 2
-            }, {
-                maxRetries: 3
-            });
-            return (yield (0, streamToBuffer_1.streamToBuffer)(response));
+            try {
+                const response = yield this.client.generate({
+                    stream: true,
+                    voice,
+                    text: chunkText,
+                    model_id: model !== null && model !== void 0 ? model : "eleven_flash_v2_5",
+                    output_format: "ulaw_8000",
+                    // TODO: Make this configurable
+                    optimize_streaming_latency: 2
+                }, {
+                    maxRetries: 3
+                });
+                const ulawBuffer = yield (0, streamToBuffer_1.streamToBuffer)(response);
+                const pcmBuffer = yield (0, convertUlawToPCM16_1.convertUlawToPCM16)(stream_1.Readable.from(ulawBuffer));
+                return pcmBuffer;
+            }
+            catch (error) {
+                logger.error(`error in ElevenLabs synthesis: ${error.message}`, {
+                    stack: error.stack
+                });
+                throw error;
+            }
         }));
         return { ref, stream };
     }

package/dist/voice/tts/Google.d.ts

@@ -29,7 +29,7 @@ declare class Google extends AbstractTextToSpeech<typeof ENGINE_NAME> {
     protected readonly OUTPUT_FORMAT = "sln16";
     protected readonly CACHING_FIELDS: string[];
     protected readonly AUDIO_ENCODING: "LINEAR16";
-    protected readonly SAMPLE_RATE_HERTZ = 16000;
+    protected readonly SAMPLE_RATE_HERTZ = 8000;
     constructor(config: GoogleTtsConfig);
     synthesize(text: string, options: SynthOptions): {
         ref: string;

package/dist/voice/tts/Google.js

@@ -58,7 +58,7 @@ class Google extends AbstractTextToSpeech_1.AbstractTextToSpeech {
         this.OUTPUT_FORMAT = "sln16";
         this.CACHING_FIELDS = ["voice"];
         this.AUDIO_ENCODING = "LINEAR16";
-        this.SAMPLE_RATE_HERTZ = 16000;
+        this.SAMPLE_RATE_HERTZ = 8000;
         this.client = new text_to_speech_1.TextToSpeechClient(config);
         this.engineConfig = config;
     }

package/dist/voice/tts/utils/convertUlawToPCM16.d.ts

@@ -0,0 +1,26 @@
+/**
+ * Copyright (C) 2025 by Fonoster Inc (https://fonoster.com)
+ * http://github.com/fonoster/fonoster
+ *
+ * This file is part of Fonoster
+ *
+ * Licensed under the MIT License (the "License");
+ * you may not use this file except in compliance with
+ * the License. You may obtain a copy of the License at
+ *
+ *    https://opensource.org/licenses/MIT
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+import { Readable } from "stream";
+/**
+ * Converts a ulaw stream to PCM 16-bit at 8kHz
+ *
+ * @param readableStream - The input ulaw audio stream to convert
+ * @returns A Promise that resolves to a Buffer containing PCM 16-bit audio data at 8kHz
+ */
+export declare function convertUlawToPCM16(readableStream: Readable): Promise<Buffer>;

package/dist/voice/tts/utils/convertUlawToPCM16.js

@@ -0,0 +1,40 @@
+"use strict";
+var __awaiter = (this && this.__awaiter) || function (thisArg, _arguments, P, generator) {
+    function adopt(value) { return value instanceof P ? value : new P(function (resolve) { resolve(value); }); }
+    return new (P || (P = Promise))(function (resolve, reject) {
+        function fulfilled(value) { try { step(generator.next(value)); } catch (e) { reject(e); } }
+        function rejected(value) { try { step(generator["throw"](value)); } catch (e) { reject(e); } }
+        function step(result) { result.done ? resolve(result.value) : adopt(result.value).then(fulfilled, rejected); }
+        step((generator = generator.apply(thisArg, _arguments || [])).next());
+    });
+};
+Object.defineProperty(exports, "__esModule", { value: true });
+exports.convertUlawToPCM16 = convertUlawToPCM16;
+const logger_1 = require("@fonoster/logger");
+const wavefile_1 = require("wavefile");
+const streamToBuffer_1 = require("./streamToBuffer");
+const logger = (0, logger_1.getLogger)({ service: "apiserver", filePath: __filename });
+/**
+ * Converts a ulaw stream to PCM 16-bit at 8kHz
+ *
+ * @param readableStream - The input ulaw audio stream to convert
+ * @returns A Promise that resolves to a Buffer containing PCM 16-bit audio data at 8kHz
+ */
+function convertUlawToPCM16(readableStream) {
+    return __awaiter(this, void 0, void 0, function* () {
+        try {
+            const inputBuffer = yield (0, streamToBuffer_1.streamToBuffer)(readableStream);
+            const wav = new wavefile_1.WaveFile();
+            wav.fromScratch(1, 8000, "8m", inputBuffer);
+            wav.fromMuLaw();
+            const waveBuffer = wav.toBuffer();
+            return Buffer.from(waveBuffer.buffer, waveBuffer.byteOffset + 44, waveBuffer.byteLength - 44);
+        }
+        catch (error) {
+            logger.error(`error converting audio format: ${error.message}`, {
+                stack: error.stack
+            });
+            throw new Error(`Audio conversion failed: ${error.message}`);
+        }
+    });
+}

package/dist/voice/tts/utils/index.d.ts

@@ -0,0 +1,23 @@
+/**
+ * Copyright (C) 2025 by Fonoster Inc (https://fonoster.com)
+ * http://github.com/fonoster/fonoster
+ *
+ * This file is part of Fonoster
+ *
+ * Licensed under the MIT License (the "License");
+ * you may not use this file except in compliance with
+ * the License. You may obtain a copy of the License at
+ *
+ *    https://opensource.org/licenses/MIT
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+export * from "./createChunkedSynthesisStream";
+export * from "./createErrorStream";
+export * from "./isSsml";
+export * from "./streamToBuffer";
+export * from "./convertUlawToPCM16";