@tensamin/audio 0.2.0 → 0.2.2

package/README.md

@@ -9,6 +9,9 @@ DeepFilterNet3-based noise suppression and realtime speaking detection for LiveK
 - Automatic mute/unmute for LiveKit tracks
 - Simple min/max dB speaking thresholds
 
+> [Noise suppression is provided via the `deepfilternet3-noise-filter` package.](https://www.npmjs.com/package/deepfilternet3-noise-filter)
+> [That package is based on DeepFilterNet by Rikorose.](https://github.com/Rikorose/DeepFilterNet)
+
 ## Installation
 
 ```bash

package/dist/chunk-7IKKNKM7.mjs

@@ -0,0 +1,92 @@
+import {
+  LevelBasedVAD
+} from "./chunk-AQ5RVY33.mjs";
+import {
+  getAudioContext,
+  registerPipeline,
+  unregisterPipeline
+} from "./chunk-OZ7KMC4S.mjs";
+import {
+  createLevelDetectorNode
+} from "./chunk-QNQK6QFB.mjs";
+
+// src/pipeline/remote-audio-monitor.ts
+import mitt from "mitt";
+async function createRemoteAudioMonitor(sourceTrack, config = {}) {
+  const context = getAudioContext();
+  registerPipeline();
+  const fullConfig = {
+    speaking: {
+      minDb: config.speaking?.minDb ?? -55,
+      maxDb: config.speaking?.maxDb ?? -20,
+      speakOnRatio: config.speaking?.speakOnRatio ?? 0.5,
+      speakOffRatio: config.speaking?.speakOffRatio ?? 0.3,
+      hangoverMs: config.speaking?.hangoverMs ?? 500,
+      attackMs: config.speaking?.attackMs ?? 100,
+      releaseMs: config.speaking?.releaseMs ?? 120
+    }
+  };
+  if (!sourceTrack || sourceTrack.kind !== "audio") {
+    throw new Error(
+      "createRemoteAudioMonitor requires a valid audio MediaStreamTrack"
+    );
+  }
+  if (sourceTrack.readyState === "ended") {
+    throw new Error("Cannot create monitor from an ended MediaStreamTrack");
+  }
+  const sourceStream = new MediaStream([sourceTrack]);
+  const sourceNode = context.createMediaStreamSource(sourceStream);
+  const emitter = mitt();
+  const vad = new LevelBasedVAD(fullConfig.speaking);
+  let lastState = { speaking: false, levelDb: -Infinity };
+  const levelHandle = await createLevelDetectorNode(context, (levelDb) => {
+    try {
+      const timestamp = context.currentTime * 1e3;
+      const nextState = vad.process(levelDb, timestamp);
+      const speakingChanged = nextState.speaking !== lastState.speaking;
+      const levelChanged = Math.abs(nextState.levelDb - lastState.levelDb) > 0.5;
+      if (speakingChanged || levelChanged) {
+        lastState = nextState;
+        emitter.emit("speakingChange", nextState);
+      }
+    } catch (error) {
+      const err = error instanceof Error ? error : new Error(String(error));
+      emitter.emit("error", err);
+    }
+  });
+  sourceNode.connect(levelHandle.node);
+  function dispose() {
+    try {
+      sourceNode.disconnect();
+      levelHandle.node.disconnect();
+      levelHandle.dispose();
+    } catch (error) {
+      console.error("Error during remote monitor disposal", error);
+    } finally {
+      unregisterPipeline();
+    }
+  }
+  const handle = {
+    events: emitter,
+    get state() {
+      return lastState;
+    },
+    setConfig: (next) => {
+      try {
+        if (next.speaking) {
+          vad.updateConfig(next.speaking);
+          fullConfig.speaking = { ...fullConfig.speaking, ...next.speaking };
+        }
+      } catch (error) {
+        const err = error instanceof Error ? error : new Error(String(error));
+        emitter.emit("error", err);
+      }
+    },
+    dispose
+  };
+  return handle;
+}
+
+export {
+  createRemoteAudioMonitor
+};

package/dist/{chunk-K4J3UUOR.mjs → chunk-BAUJY4Q2.mjs}

@@ -1,17 +1,17 @@
+import {
+  createDeepFilterNet3Node
+} from "./chunk-IS37FHDN.mjs";
+import {
+  LevelBasedVAD
+} from "./chunk-AQ5RVY33.mjs";
 import {
   getAudioContext,
   registerPipeline,
   unregisterPipeline
 } from "./chunk-OZ7KMC4S.mjs";
-import {
-  createDeepFilterNet3Node
-} from "./chunk-IS37FHDN.mjs";
 import {
   createLevelDetectorNode
 } from "./chunk-QNQK6QFB.mjs";
-import {
-  LevelBasedVAD
-} from "./chunk-AQ5RVY33.mjs";
 
 // src/pipeline/audio-pipeline.ts
 import mitt from "mitt";
@@ -28,12 +28,12 @@ async function createAudioPipeline(sourceTrack, config = {}) {
   const fullConfig = {
     noiseSuppression: nsConfig,
     speaking: {
-      minDb: config.speaking?.minDb ?? -60,
+      minDb: config.speaking?.minDb ?? -55,
       maxDb: config.speaking?.maxDb ?? -20,
-      speakOnRatio: config.speaking?.speakOnRatio ?? 0.6,
+      speakOnRatio: config.speaking?.speakOnRatio ?? 0.5,
       speakOffRatio: config.speaking?.speakOffRatio ?? 0.3,
-      hangoverMs: config.speaking?.hangoverMs ?? 350,
-      attackMs: config.speaking?.attackMs ?? 50,
+      hangoverMs: config.speaking?.hangoverMs ?? 500,
+      attackMs: config.speaking?.attackMs ?? 100,
       releaseMs: config.speaking?.releaseMs ?? 120
     },
     output: {

package/dist/{chunk-6BJ4XGSA.mjs → chunk-YQPL2O7D.mjs}

@@ -1,9 +1,11 @@
 import {
   createAudioPipeline
-} from "./chunk-K4J3UUOR.mjs";
+} from "./chunk-BAUJY4Q2.mjs";
+import {
+  createRemoteAudioMonitor
+} from "./chunk-7IKKNKM7.mjs";
 
 // src/livekit/integration.ts
-import "mitt";
 async function attachSpeakingDetectionToTrack(track, options = {}) {
   if (!track) {
     throw new Error(
@@ -74,7 +76,54 @@ async function attachSpeakingDetectionToTrack(track, options = {}) {
   };
   return controller;
 }
+async function attachSpeakingDetectionToRemoteTrack(track, options = {}) {
+  if (!track) {
+    throw new Error(
+      "attachSpeakingDetectionToRemoteTrack requires a valid RemoteAudioTrack"
+    );
+  }
+  const mediaTrack = track.mediaStreamTrack;
+  if (!mediaTrack || mediaTrack.readyState === "ended") {
+    throw new Error("RemoteAudioTrack has no live MediaStreamTrack to monitor");
+  }
+  const monitor = await createRemoteAudioMonitor(mediaTrack, options);
+  const listeners = /* @__PURE__ */ new Set();
+  let currentState = monitor.state;
+  const speakingHandler = (state) => {
+    currentState = state;
+    listeners.forEach((listener) => listener(state));
+  };
+  monitor.events.on("speakingChange", speakingHandler);
+  const errorHandler = (error) => {
+    console.error("Remote audio monitor error", error);
+  };
+  monitor.events.on("error", errorHandler);
+  const controller = {
+    get speaking() {
+      return currentState.speaking;
+    },
+    get levelDb() {
+      return currentState.levelDb;
+    },
+    onChange: (listener) => {
+      listeners.add(listener);
+      listener(currentState);
+      return () => listeners.delete(listener);
+    },
+    setConfig: (config) => {
+      monitor.setConfig(config);
+    },
+    dispose: () => {
+      monitor.events.off("speakingChange", speakingHandler);
+      monitor.events.off("error", errorHandler);
+      listeners.clear();
+      monitor.dispose();
+    }
+  };
+  return controller;
+}
 
 export {
-  attachSpeakingDetectionToTrack
+  attachSpeakingDetectionToTrack,
+  attachSpeakingDetectionToRemoteTrack
 };

package/dist/index.d.mts

@@ -1,4 +1,4 @@
-export { AudioPipelineHandle, LivekitSpeakingOptions, NoiseSuppressionConfig, OutputGainConfig, SpeakingController, SpeakingDetectionConfig, SpeakingEvents, SpeakingState } from './types.mjs';
-export { attachSpeakingDetectionToTrack } from './livekit/integration.mjs';
+export { AudioPipelineHandle, LivekitSpeakingOptions, NoiseSuppressionConfig, OutputGainConfig, RemoteSpeakingOptions, SpeakingController, SpeakingDetectionConfig, SpeakingEvents, SpeakingState } from './types.mjs';
+export { attachSpeakingDetectionToRemoteTrack, attachSpeakingDetectionToTrack } from './livekit/integration.mjs';
 import 'mitt';
 import 'livekit-client';

package/dist/index.d.ts

@@ -1,4 +1,4 @@
-export { AudioPipelineHandle, LivekitSpeakingOptions, NoiseSuppressionConfig, OutputGainConfig, SpeakingController, SpeakingDetectionConfig, SpeakingEvents, SpeakingState } from './types.js';
-export { attachSpeakingDetectionToTrack } from './livekit/integration.js';
+export { AudioPipelineHandle, LivekitSpeakingOptions, NoiseSuppressionConfig, OutputGainConfig, RemoteSpeakingOptions, SpeakingController, SpeakingDetectionConfig, SpeakingEvents, SpeakingState } from './types.js';
+export { attachSpeakingDetectionToRemoteTrack, attachSpeakingDetectionToTrack } from './livekit/integration.js';
 import 'mitt';
 import 'livekit-client';

package/dist/index.js

@@ -30,13 +30,11 @@ var __toCommonJS = (mod) => __copyProps(__defProp({}, "__esModule", { value: tru
 // src/index.ts
 var index_exports = {};
 __export(index_exports, {
+  attachSpeakingDetectionToRemoteTrack: () => attachSpeakingDetectionToRemoteTrack,
   attachSpeakingDetectionToTrack: () => attachSpeakingDetectionToTrack
 });
 module.exports = __toCommonJS(index_exports);
 
-// src/livekit/integration.ts
-var import_mitt2 = require("mitt");
-
 // src/pipeline/audio-pipeline.ts
 var import_mitt = __toESM(require("mitt"));
 
@@ -244,12 +242,12 @@ async function createAudioPipeline(sourceTrack, config = {}) {
   const fullConfig = {
     noiseSuppression: nsConfig,
     speaking: {
-      minDb: config.speaking?.minDb ?? -60,
+      minDb: config.speaking?.minDb ?? -55,
       maxDb: config.speaking?.maxDb ?? -20,
-      speakOnRatio: config.speaking?.speakOnRatio ?? 0.6,
+      speakOnRatio: config.speaking?.speakOnRatio ?? 0.5,
       speakOffRatio: config.speaking?.speakOffRatio ?? 0.3,
-      hangoverMs: config.speaking?.hangoverMs ?? 350,
-      attackMs: config.speaking?.attackMs ?? 50,
+      hangoverMs: config.speaking?.hangoverMs ?? 500,
+      attackMs: config.speaking?.attackMs ?? 100,
       releaseMs: config.speaking?.releaseMs ?? 120
     },
     output: {
@@ -389,6 +387,83 @@ async function createAudioPipeline(sourceTrack, config = {}) {
   return handle;
 }
 
+// src/pipeline/remote-audio-monitor.ts
+var import_mitt2 = __toESM(require("mitt"));
+async function createRemoteAudioMonitor(sourceTrack, config = {}) {
+  const context = getAudioContext();
+  registerPipeline();
+  const fullConfig = {
+    speaking: {
+      minDb: config.speaking?.minDb ?? -55,
+      maxDb: config.speaking?.maxDb ?? -20,
+      speakOnRatio: config.speaking?.speakOnRatio ?? 0.5,
+      speakOffRatio: config.speaking?.speakOffRatio ?? 0.3,
+      hangoverMs: config.speaking?.hangoverMs ?? 500,
+      attackMs: config.speaking?.attackMs ?? 100,
+      releaseMs: config.speaking?.releaseMs ?? 120
+    }
+  };
+  if (!sourceTrack || sourceTrack.kind !== "audio") {
+    throw new Error(
+      "createRemoteAudioMonitor requires a valid audio MediaStreamTrack"
+    );
+  }
+  if (sourceTrack.readyState === "ended") {
+    throw new Error("Cannot create monitor from an ended MediaStreamTrack");
+  }
+  const sourceStream = new MediaStream([sourceTrack]);
+  const sourceNode = context.createMediaStreamSource(sourceStream);
+  const emitter = (0, import_mitt2.default)();
+  const vad = new LevelBasedVAD(fullConfig.speaking);
+  let lastState = { speaking: false, levelDb: -Infinity };
+  const levelHandle = await createLevelDetectorNode(context, (levelDb) => {
+    try {
+      const timestamp = context.currentTime * 1e3;
+      const nextState = vad.process(levelDb, timestamp);
+      const speakingChanged = nextState.speaking !== lastState.speaking;
+      const levelChanged = Math.abs(nextState.levelDb - lastState.levelDb) > 0.5;
+      if (speakingChanged || levelChanged) {
+        lastState = nextState;
+        emitter.emit("speakingChange", nextState);
+      }
+    } catch (error) {
+      const err = error instanceof Error ? error : new Error(String(error));
+      emitter.emit("error", err);
+    }
+  });
+  sourceNode.connect(levelHandle.node);
+  function dispose() {
+    try {
+      sourceNode.disconnect();
+      levelHandle.node.disconnect();
+      levelHandle.dispose();
+    } catch (error) {
+      console.error("Error during remote monitor disposal", error);
+    } finally {
+      unregisterPipeline();
+    }
+  }
+  const handle = {
+    events: emitter,
+    get state() {
+      return lastState;
+    },
+    setConfig: (next) => {
+      try {
+        if (next.speaking) {
+          vad.updateConfig(next.speaking);
+          fullConfig.speaking = { ...fullConfig.speaking, ...next.speaking };
+        }
+      } catch (error) {
+        const err = error instanceof Error ? error : new Error(String(error));
+        emitter.emit("error", err);
+      }
+    },
+    dispose
+  };
+  return handle;
+}
+
 // src/livekit/integration.ts
 async function attachSpeakingDetectionToTrack(track, options = {}) {
   if (!track) {
@@ -460,7 +535,54 @@ async function attachSpeakingDetectionToTrack(track, options = {}) {
   };
   return controller;
 }
+async function attachSpeakingDetectionToRemoteTrack(track, options = {}) {
+  if (!track) {
+    throw new Error(
+      "attachSpeakingDetectionToRemoteTrack requires a valid RemoteAudioTrack"
+    );
+  }
+  const mediaTrack = track.mediaStreamTrack;
+  if (!mediaTrack || mediaTrack.readyState === "ended") {
+    throw new Error("RemoteAudioTrack has no live MediaStreamTrack to monitor");
+  }
+  const monitor = await createRemoteAudioMonitor(mediaTrack, options);
+  const listeners = /* @__PURE__ */ new Set();
+  let currentState = monitor.state;
+  const speakingHandler = (state) => {
+    currentState = state;
+    listeners.forEach((listener) => listener(state));
+  };
+  monitor.events.on("speakingChange", speakingHandler);
+  const errorHandler = (error) => {
+    console.error("Remote audio monitor error", error);
+  };
+  monitor.events.on("error", errorHandler);
+  const controller = {
+    get speaking() {
+      return currentState.speaking;
+    },
+    get levelDb() {
+      return currentState.levelDb;
+    },
+    onChange: (listener) => {
+      listeners.add(listener);
+      listener(currentState);
+      return () => listeners.delete(listener);
+    },
+    setConfig: (config) => {
+      monitor.setConfig(config);
+    },
+    dispose: () => {
+      monitor.events.off("speakingChange", speakingHandler);
+      monitor.events.off("error", errorHandler);
+      listeners.clear();
+      monitor.dispose();
+    }
+  };
+  return controller;
+}
 // Annotate the CommonJS export names for ESM import in node:
 0 && (module.exports = {
+  attachSpeakingDetectionToRemoteTrack,
   attachSpeakingDetectionToTrack
 });

package/dist/index.mjs

@@ -1,12 +1,15 @@
 import "./chunk-WBQAMGXK.mjs";
 import {
+  attachSpeakingDetectionToRemoteTrack,
   attachSpeakingDetectionToTrack
-} from "./chunk-6BJ4XGSA.mjs";
-import "./chunk-K4J3UUOR.mjs";
-import "./chunk-OZ7KMC4S.mjs";
+} from "./chunk-YQPL2O7D.mjs";
+import "./chunk-BAUJY4Q2.mjs";
 import "./chunk-IS37FHDN.mjs";
-import "./chunk-QNQK6QFB.mjs";
+import "./chunk-7IKKNKM7.mjs";
 import "./chunk-AQ5RVY33.mjs";
+import "./chunk-OZ7KMC4S.mjs";
+import "./chunk-QNQK6QFB.mjs";
 export {
+  attachSpeakingDetectionToRemoteTrack,
   attachSpeakingDetectionToTrack
 };

package/dist/livekit/integration.d.mts

@@ -1,7 +1,8 @@
-import { LivekitSpeakingOptions, SpeakingController } from '../types.mjs';
-import { LocalAudioTrack } from 'livekit-client';
+import { LivekitSpeakingOptions, SpeakingController, RemoteSpeakingOptions } from '../types.mjs';
+import { LocalAudioTrack, RemoteAudioTrack } from 'livekit-client';
 import 'mitt';
 
 declare function attachSpeakingDetectionToTrack(track: LocalAudioTrack, options?: LivekitSpeakingOptions): Promise<SpeakingController>;
+declare function attachSpeakingDetectionToRemoteTrack(track: RemoteAudioTrack, options?: RemoteSpeakingOptions): Promise<SpeakingController>;
 
-export { attachSpeakingDetectionToTrack };
+export { attachSpeakingDetectionToRemoteTrack, attachSpeakingDetectionToTrack };

package/dist/livekit/integration.d.ts

@@ -1,7 +1,8 @@
-import { LivekitSpeakingOptions, SpeakingController } from '../types.js';
-import { LocalAudioTrack } from 'livekit-client';
+import { LivekitSpeakingOptions, SpeakingController, RemoteSpeakingOptions } from '../types.js';
+import { LocalAudioTrack, RemoteAudioTrack } from 'livekit-client';
 import 'mitt';
 
 declare function attachSpeakingDetectionToTrack(track: LocalAudioTrack, options?: LivekitSpeakingOptions): Promise<SpeakingController>;
+declare function attachSpeakingDetectionToRemoteTrack(track: RemoteAudioTrack, options?: RemoteSpeakingOptions): Promise<SpeakingController>;
 
-export { attachSpeakingDetectionToTrack };
+export { attachSpeakingDetectionToRemoteTrack, attachSpeakingDetectionToTrack };

package/dist/livekit/integration.js

@@ -30,10 +30,10 @@ var __toCommonJS = (mod) => __copyProps(__defProp({}, "__esModule", { value: tru
 // src/livekit/integration.ts
 var integration_exports = {};
 __export(integration_exports, {
+  attachSpeakingDetectionToRemoteTrack: () => attachSpeakingDetectionToRemoteTrack,
   attachSpeakingDetectionToTrack: () => attachSpeakingDetectionToTrack
 });
 module.exports = __toCommonJS(integration_exports);
-var import_mitt2 = require("mitt");
 
 // src/pipeline/audio-pipeline.ts
 var import_mitt = __toESM(require("mitt"));
@@ -242,12 +242,12 @@ async function createAudioPipeline(sourceTrack, config = {}) {
   const fullConfig = {
     noiseSuppression: nsConfig,
     speaking: {
-      minDb: config.speaking?.minDb ?? -60,
+      minDb: config.speaking?.minDb ?? -55,
       maxDb: config.speaking?.maxDb ?? -20,
-      speakOnRatio: config.speaking?.speakOnRatio ?? 0.6,
+      speakOnRatio: config.speaking?.speakOnRatio ?? 0.5,
       speakOffRatio: config.speaking?.speakOffRatio ?? 0.3,
-      hangoverMs: config.speaking?.hangoverMs ?? 350,
-      attackMs: config.speaking?.attackMs ?? 50,
+      hangoverMs: config.speaking?.hangoverMs ?? 500,
+      attackMs: config.speaking?.attackMs ?? 100,
       releaseMs: config.speaking?.releaseMs ?? 120
     },
     output: {
@@ -387,6 +387,83 @@ async function createAudioPipeline(sourceTrack, config = {}) {
   return handle;
 }
 
+// src/pipeline/remote-audio-monitor.ts
+var import_mitt2 = __toESM(require("mitt"));
+async function createRemoteAudioMonitor(sourceTrack, config = {}) {
+  const context = getAudioContext();
+  registerPipeline();
+  const fullConfig = {
+    speaking: {
+      minDb: config.speaking?.minDb ?? -55,
+      maxDb: config.speaking?.maxDb ?? -20,
+      speakOnRatio: config.speaking?.speakOnRatio ?? 0.5,
+      speakOffRatio: config.speaking?.speakOffRatio ?? 0.3,
+      hangoverMs: config.speaking?.hangoverMs ?? 500,
+      attackMs: config.speaking?.attackMs ?? 100,
+      releaseMs: config.speaking?.releaseMs ?? 120
+    }
+  };
+  if (!sourceTrack || sourceTrack.kind !== "audio") {
+    throw new Error(
+      "createRemoteAudioMonitor requires a valid audio MediaStreamTrack"
+    );
+  }
+  if (sourceTrack.readyState === "ended") {
+    throw new Error("Cannot create monitor from an ended MediaStreamTrack");
+  }
+  const sourceStream = new MediaStream([sourceTrack]);
+  const sourceNode = context.createMediaStreamSource(sourceStream);
+  const emitter = (0, import_mitt2.default)();
+  const vad = new LevelBasedVAD(fullConfig.speaking);
+  let lastState = { speaking: false, levelDb: -Infinity };
+  const levelHandle = await createLevelDetectorNode(context, (levelDb) => {
+    try {
+      const timestamp = context.currentTime * 1e3;
+      const nextState = vad.process(levelDb, timestamp);
+      const speakingChanged = nextState.speaking !== lastState.speaking;
+      const levelChanged = Math.abs(nextState.levelDb - lastState.levelDb) > 0.5;
+      if (speakingChanged || levelChanged) {
+        lastState = nextState;
+        emitter.emit("speakingChange", nextState);
+      }
+    } catch (error) {
+      const err = error instanceof Error ? error : new Error(String(error));
+      emitter.emit("error", err);
+    }
+  });
+  sourceNode.connect(levelHandle.node);
+  function dispose() {
+    try {
+      sourceNode.disconnect();
+      levelHandle.node.disconnect();
+      levelHandle.dispose();
+    } catch (error) {
+      console.error("Error during remote monitor disposal", error);
+    } finally {
+      unregisterPipeline();
+    }
+  }
+  const handle = {
+    events: emitter,
+    get state() {
+      return lastState;
+    },
+    setConfig: (next) => {
+      try {
+        if (next.speaking) {
+          vad.updateConfig(next.speaking);
+          fullConfig.speaking = { ...fullConfig.speaking, ...next.speaking };
+        }
+      } catch (error) {
+        const err = error instanceof Error ? error : new Error(String(error));
+        emitter.emit("error", err);
+      }
+    },
+    dispose
+  };
+  return handle;
+}
+
 // src/livekit/integration.ts
 async function attachSpeakingDetectionToTrack(track, options = {}) {
   if (!track) {
@@ -458,7 +535,54 @@ async function attachSpeakingDetectionToTrack(track, options = {}) {
   };
   return controller;
 }
+async function attachSpeakingDetectionToRemoteTrack(track, options = {}) {
+  if (!track) {
+    throw new Error(
+      "attachSpeakingDetectionToRemoteTrack requires a valid RemoteAudioTrack"
+    );
+  }
+  const mediaTrack = track.mediaStreamTrack;
+  if (!mediaTrack || mediaTrack.readyState === "ended") {
+    throw new Error("RemoteAudioTrack has no live MediaStreamTrack to monitor");
+  }
+  const monitor = await createRemoteAudioMonitor(mediaTrack, options);
+  const listeners = /* @__PURE__ */ new Set();
+  let currentState = monitor.state;
+  const speakingHandler = (state) => {
+    currentState = state;
+    listeners.forEach((listener) => listener(state));
+  };
+  monitor.events.on("speakingChange", speakingHandler);
+  const errorHandler = (error) => {
+    console.error("Remote audio monitor error", error);
+  };
+  monitor.events.on("error", errorHandler);
+  const controller = {
+    get speaking() {
+      return currentState.speaking;
+    },
+    get levelDb() {
+      return currentState.levelDb;
+    },
+    onChange: (listener) => {
+      listeners.add(listener);
+      listener(currentState);
+      return () => listeners.delete(listener);
+    },
+    setConfig: (config) => {
+      monitor.setConfig(config);
+    },
+    dispose: () => {
+      monitor.events.off("speakingChange", speakingHandler);
+      monitor.events.off("error", errorHandler);
+      listeners.clear();
+      monitor.dispose();
+    }
+  };
+  return controller;
+}
 // Annotate the CommonJS export names for ESM import in node:
 0 && (module.exports = {
+  attachSpeakingDetectionToRemoteTrack,
   attachSpeakingDetectionToTrack
 });

package/dist/livekit/integration.mjs

@@ -1,11 +1,14 @@
 import {
+  attachSpeakingDetectionToRemoteTrack,
   attachSpeakingDetectionToTrack
-} from "../chunk-6BJ4XGSA.mjs";
-import "../chunk-K4J3UUOR.mjs";
-import "../chunk-OZ7KMC4S.mjs";
+} from "../chunk-YQPL2O7D.mjs";
+import "../chunk-BAUJY4Q2.mjs";
 import "../chunk-IS37FHDN.mjs";
-import "../chunk-QNQK6QFB.mjs";
+import "../chunk-7IKKNKM7.mjs";
 import "../chunk-AQ5RVY33.mjs";
+import "../chunk-OZ7KMC4S.mjs";
+import "../chunk-QNQK6QFB.mjs";
 export {
+  attachSpeakingDetectionToRemoteTrack,
   attachSpeakingDetectionToTrack
 };

package/dist/pipeline/audio-pipeline.js

@@ -239,12 +239,12 @@ async function createAudioPipeline(sourceTrack, config = {}) {
   const fullConfig = {
     noiseSuppression: nsConfig,
     speaking: {
-      minDb: config.speaking?.minDb ?? -60,
+      minDb: config.speaking?.minDb ?? -55,
       maxDb: config.speaking?.maxDb ?? -20,
-      speakOnRatio: config.speaking?.speakOnRatio ?? 0.6,
+      speakOnRatio: config.speaking?.speakOnRatio ?? 0.5,
       speakOffRatio: config.speaking?.speakOffRatio ?? 0.3,
-      hangoverMs: config.speaking?.hangoverMs ?? 350,
-      attackMs: config.speaking?.attackMs ?? 50,
+      hangoverMs: config.speaking?.hangoverMs ?? 500,
+      attackMs: config.speaking?.attackMs ?? 100,
       releaseMs: config.speaking?.releaseMs ?? 120
     },
     output: {

package/dist/pipeline/audio-pipeline.mjs

@@ -1,10 +1,10 @@
 import {
   createAudioPipeline
-} from "../chunk-K4J3UUOR.mjs";
-import "../chunk-OZ7KMC4S.mjs";
+} from "../chunk-BAUJY4Q2.mjs";
 import "../chunk-IS37FHDN.mjs";
-import "../chunk-QNQK6QFB.mjs";
 import "../chunk-AQ5RVY33.mjs";
+import "../chunk-OZ7KMC4S.mjs";
+import "../chunk-QNQK6QFB.mjs";
 export {
   createAudioPipeline
 };

package/dist/pipeline/remote-audio-monitor.d.mts

@@ -0,0 +1,12 @@
+import { Emitter } from 'mitt';
+import { SpeakingEvents, SpeakingState, RemoteSpeakingOptions } from '../types.mjs';
+
+interface RemoteAudioMonitorHandle {
+    readonly events: Emitter<SpeakingEvents>;
+    readonly state: SpeakingState;
+    setConfig(config: Partial<RemoteSpeakingOptions>): void;
+    dispose(): void;
+}
+declare function createRemoteAudioMonitor(sourceTrack: MediaStreamTrack, config?: RemoteSpeakingOptions): Promise<RemoteAudioMonitorHandle>;
+
+export { type RemoteAudioMonitorHandle, createRemoteAudioMonitor };

package/dist/pipeline/remote-audio-monitor.d.ts

@@ -0,0 +1,12 @@
+import { Emitter } from 'mitt';
+import { SpeakingEvents, SpeakingState, RemoteSpeakingOptions } from '../types.js';
+
+interface RemoteAudioMonitorHandle {
+    readonly events: Emitter<SpeakingEvents>;
+    readonly state: SpeakingState;
+    setConfig(config: Partial<RemoteSpeakingOptions>): void;
+    dispose(): void;
+}
+declare function createRemoteAudioMonitor(sourceTrack: MediaStreamTrack, config?: RemoteSpeakingOptions): Promise<RemoteAudioMonitorHandle>;
+
+export { type RemoteAudioMonitorHandle, createRemoteAudioMonitor };

package/dist/pipeline/remote-audio-monitor.js

@@ -0,0 +1,276 @@
+"use strict";
+var __create = Object.create;
+var __defProp = Object.defineProperty;
+var __getOwnPropDesc = Object.getOwnPropertyDescriptor;
+var __getOwnPropNames = Object.getOwnPropertyNames;
+var __getProtoOf = Object.getPrototypeOf;
+var __hasOwnProp = Object.prototype.hasOwnProperty;
+var __export = (target, all) => {
+  for (var name in all)
+    __defProp(target, name, { get: all[name], enumerable: true });
+};
+var __copyProps = (to, from, except, desc) => {
+  if (from && typeof from === "object" || typeof from === "function") {
+    for (let key of __getOwnPropNames(from))
+      if (!__hasOwnProp.call(to, key) && key !== except)
+        __defProp(to, key, { get: () => from[key], enumerable: !(desc = __getOwnPropDesc(from, key)) || desc.enumerable });
+  }
+  return to;
+};
+var __toESM = (mod, isNodeMode, target) => (target = mod != null ? __create(__getProtoOf(mod)) : {}, __copyProps(
+  // If the importer is in node compatibility mode or this is not an ESM
+  // file that has been converted to a CommonJS file using a Babel-
+  // compatible transform (i.e. "__esModule" has not been set), then set
+  // "default" to the CommonJS "module.exports" for node compatibility.
+  isNodeMode || !mod || !mod.__esModule ? __defProp(target, "default", { value: mod, enumerable: true }) : target,
+  mod
+));
+var __toCommonJS = (mod) => __copyProps(__defProp({}, "__esModule", { value: true }), mod);
+
+// src/pipeline/remote-audio-monitor.ts
+var remote_audio_monitor_exports = {};
+__export(remote_audio_monitor_exports, {
+  createRemoteAudioMonitor: () => createRemoteAudioMonitor
+});
+module.exports = __toCommonJS(remote_audio_monitor_exports);
+var import_mitt = __toESM(require("mitt"));
+
+// src/context/audio-context.ts
+var sharedContext = null;
+var activePipelines = 0;
+function getAudioContext(options) {
+  if (typeof window === "undefined" || typeof AudioContext === "undefined") {
+    throw new Error(
+      "AudioContext is not supported in this environment (browser only)."
+    );
+  }
+  if (!sharedContext || sharedContext.state === "closed") {
+    sharedContext = new AudioContext(options);
+  }
+  return sharedContext;
+}
+function registerPipeline() {
+  activePipelines++;
+}
+function unregisterPipeline() {
+  activePipelines = Math.max(0, activePipelines - 1);
+}
+
+// src/vad/vad-node.ts
+function createLevelDetectorWorkletCode(smoothing) {
+  return `
+class LevelDetectorProcessor extends AudioWorkletProcessor {
+  constructor() {
+    super();
+    this.smoothed = 0;
+    this.smoothing = ${smoothing};
+  }
+
+  process(inputs) {
+    const input = inputs[0];
+    if (!input || input.length === 0) return true;
+    const channel = input[0];
+    if (!channel || channel.length === 0) return true;
+
+    let sum = 0;
+    for (let i = 0; i < channel.length; i++) {
+      const sample = channel[i];
+      sum += sample * sample;
+    }
+    const rms = Math.sqrt(sum / channel.length);
+    this.smoothed = this.smoothed * this.smoothing + rms * (1 - this.smoothing);
+    const levelDb = 20 * Math.log10(Math.max(1e-8, this.smoothed));
+    this.port.postMessage({ levelDb });
+    return true;
+  }
+}
+
+registerProcessor('level-detector-processor', LevelDetectorProcessor);
+`;
+}
+async function createLevelDetectorNode(context, onLevel, options) {
+  const smoothing = options?.smoothing ?? 0.9;
+  const workletCode = createLevelDetectorWorkletCode(smoothing);
+  const blob = new Blob([workletCode], { type: "application/javascript" });
+  const url = URL.createObjectURL(blob);
+  try {
+    await context.audioWorklet.addModule(url);
+  } finally {
+    URL.revokeObjectURL(url);
+  }
+  const node = new AudioWorkletNode(context, "level-detector-processor", {
+    numberOfInputs: 1,
+    numberOfOutputs: 0
+  });
+  node.port.onmessage = (event) => {
+    const { levelDb } = event.data ?? {};
+    if (typeof levelDb === "number" && !Number.isNaN(levelDb)) {
+      onLevel(levelDb);
+    }
+  };
+  node.port.onmessageerror = (event) => {
+    console.error("Level detector port error", event);
+  };
+  return {
+    node,
+    dispose: () => {
+      try {
+        node.port.onmessage = null;
+        node.port.close();
+      } catch (error) {
+        console.error("Failed to dispose level detector node", error);
+      }
+    }
+  };
+}
+
+// src/vad/vad-state.ts
+var LevelBasedVAD = class {
+  config;
+  speaking = false;
+  pendingSpeechSince = null;
+  pendingSilenceSince = null;
+  constructor(config) {
+    this.config = {
+      minDb: config.minDb,
+      maxDb: config.maxDb,
+      speakOnRatio: config.speakOnRatio ?? 0.6,
+      speakOffRatio: config.speakOffRatio ?? 0.3,
+      hangoverMs: config.hangoverMs ?? 350,
+      attackMs: config.attackMs ?? 50,
+      releaseMs: config.releaseMs ?? 120
+    };
+  }
+  updateConfig(config) {
+    this.config = {
+      ...this.config,
+      ...config,
+      speakOnRatio: config.speakOnRatio ?? this.config.speakOnRatio,
+      speakOffRatio: config.speakOffRatio ?? this.config.speakOffRatio,
+      hangoverMs: config.hangoverMs ?? this.config.hangoverMs,
+      attackMs: config.attackMs ?? this.config.attackMs,
+      releaseMs: config.releaseMs ?? this.config.releaseMs
+    };
+  }
+  process(levelDb, timestampMs) {
+    const {
+      minDb,
+      maxDb,
+      speakOnRatio,
+      speakOffRatio,
+      hangoverMs,
+      attackMs,
+      releaseMs
+    } = this.config;
+    const clamped = Math.min(maxDb, Math.max(minDb, levelDb));
+    const norm = (clamped - minDb) / Math.max(1, maxDb - minDb);
+    if (!this.speaking) {
+      if (norm >= speakOnRatio) {
+        this.pendingSpeechSince = this.pendingSpeechSince ?? timestampMs;
+        if (timestampMs - this.pendingSpeechSince >= attackMs) {
+          this.speaking = true;
+          this.pendingSpeechSince = null;
+          this.pendingSilenceSince = null;
+        }
+      } else {
+        this.pendingSpeechSince = null;
+      }
+    } else {
+      if (norm <= speakOffRatio) {
+        this.pendingSilenceSince = this.pendingSilenceSince ?? timestampMs;
+        const releaseWindow = Math.max(releaseMs, hangoverMs);
+        if (timestampMs - this.pendingSilenceSince >= releaseWindow) {
+          this.speaking = false;
+          this.pendingSilenceSince = null;
+          this.pendingSpeechSince = null;
+        }
+      } else {
+        this.pendingSilenceSince = null;
+      }
+    }
+    return {
+      speaking: this.speaking,
+      levelDb: clamped
+    };
+  }
+};
+
+// src/pipeline/remote-audio-monitor.ts
+async function createRemoteAudioMonitor(sourceTrack, config = {}) {
+  const context = getAudioContext();
+  registerPipeline();
+  const fullConfig = {
+    speaking: {
+      minDb: config.speaking?.minDb ?? -55,
+      maxDb: config.speaking?.maxDb ?? -20,
+      speakOnRatio: config.speaking?.speakOnRatio ?? 0.5,
+      speakOffRatio: config.speaking?.speakOffRatio ?? 0.3,
+      hangoverMs: config.speaking?.hangoverMs ?? 500,
+      attackMs: config.speaking?.attackMs ?? 100,
+      releaseMs: config.speaking?.releaseMs ?? 120
+    }
+  };
+  if (!sourceTrack || sourceTrack.kind !== "audio") {
+    throw new Error(
+      "createRemoteAudioMonitor requires a valid audio MediaStreamTrack"
+    );
+  }
+  if (sourceTrack.readyState === "ended") {
+    throw new Error("Cannot create monitor from an ended MediaStreamTrack");
+  }
+  const sourceStream = new MediaStream([sourceTrack]);
+  const sourceNode = context.createMediaStreamSource(sourceStream);
+  const emitter = (0, import_mitt.default)();
+  const vad = new LevelBasedVAD(fullConfig.speaking);
+  let lastState = { speaking: false, levelDb: -Infinity };
+  const levelHandle = await createLevelDetectorNode(context, (levelDb) => {
+    try {
+      const timestamp = context.currentTime * 1e3;
+      const nextState = vad.process(levelDb, timestamp);
+      const speakingChanged = nextState.speaking !== lastState.speaking;
+      const levelChanged = Math.abs(nextState.levelDb - lastState.levelDb) > 0.5;
+      if (speakingChanged || levelChanged) {
+        lastState = nextState;
+        emitter.emit("speakingChange", nextState);
+      }
+    } catch (error) {
+      const err = error instanceof Error ? error : new Error(String(error));
+      emitter.emit("error", err);
+    }
+  });
+  sourceNode.connect(levelHandle.node);
+  function dispose() {
+    try {
+      sourceNode.disconnect();
+      levelHandle.node.disconnect();
+      levelHandle.dispose();
+    } catch (error) {
+      console.error("Error during remote monitor disposal", error);
+    } finally {
+      unregisterPipeline();
+    }
+  }
+  const handle = {
+    events: emitter,
+    get state() {
+      return lastState;
+    },
+    setConfig: (next) => {
+      try {
+        if (next.speaking) {
+          vad.updateConfig(next.speaking);
+          fullConfig.speaking = { ...fullConfig.speaking, ...next.speaking };
+        }
+      } catch (error) {
+        const err = error instanceof Error ? error : new Error(String(error));
+        emitter.emit("error", err);
+      }
+    },
+    dispose
+  };
+  return handle;
+}
+// Annotate the CommonJS export names for ESM import in node:
+0 && (module.exports = {
+  createRemoteAudioMonitor
+});

package/dist/pipeline/remote-audio-monitor.mjs

@@ -0,0 +1,9 @@
+import {
+  createRemoteAudioMonitor
+} from "../chunk-7IKKNKM7.mjs";
+import "../chunk-AQ5RVY33.mjs";
+import "../chunk-OZ7KMC4S.mjs";
+import "../chunk-QNQK6QFB.mjs";
+export {
+  createRemoteAudioMonitor
+};

package/dist/types.d.mts

@@ -29,6 +29,9 @@ interface LivekitSpeakingOptions {
     output?: OutputGainConfig;
     muteWhenSilent?: boolean;
 }
+interface RemoteSpeakingOptions {
+    speaking?: SpeakingDetectionConfig;
+}
 interface SpeakingState {
     speaking: boolean;
     levelDb: number;
@@ -52,4 +55,4 @@ interface SpeakingController {
     dispose(): void;
 }
 
-export type { AudioPipelineHandle, LivekitSpeakingOptions, NoiseSuppressionConfig, OutputGainConfig, SpeakingController, SpeakingDetectionConfig, SpeakingEvents, SpeakingState };
+export type { AudioPipelineHandle, LivekitSpeakingOptions, NoiseSuppressionConfig, OutputGainConfig, RemoteSpeakingOptions, SpeakingController, SpeakingDetectionConfig, SpeakingEvents, SpeakingState };

package/dist/types.d.ts

@@ -29,6 +29,9 @@ interface LivekitSpeakingOptions {
     output?: OutputGainConfig;
     muteWhenSilent?: boolean;
 }
+interface RemoteSpeakingOptions {
+    speaking?: SpeakingDetectionConfig;
+}
 interface SpeakingState {
     speaking: boolean;
     levelDb: number;
@@ -52,4 +55,4 @@ interface SpeakingController {
     dispose(): void;
 }
 
-export type { AudioPipelineHandle, LivekitSpeakingOptions, NoiseSuppressionConfig, OutputGainConfig, SpeakingController, SpeakingDetectionConfig, SpeakingEvents, SpeakingState };
+export type { AudioPipelineHandle, LivekitSpeakingOptions, NoiseSuppressionConfig, OutputGainConfig, RemoteSpeakingOptions, SpeakingController, SpeakingDetectionConfig, SpeakingEvents, SpeakingState };

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@tensamin/audio",
-  "version": "0.2.0",
+  "version": "0.2.2",
   "author": {
     "email": "aloisianer@proton.me",
     "name": "Alois"