retiam 2.0.2 → 2.0.3

package/dist/libs/tts/index.js

@@ -1,543 +1,12 @@
-import { D as M, c as D, p as F, g as L, s as V, f as P, a as N } from "../../msEdgeV2-BLEnvPJa.js";
-import { M as rt, b as ot } from "../../msEdgeV2-BLEnvPJa.js";
-import { Component as j, Color as y, hex as l, VGroup as H, SDFText as I, CENTER as $, Rect as X, Animation as K } from "retiam";
+import { TTS as p } from "./node/index.js";
+import { ElevenLabsTTSProvider as W, MsEdgeTTSProvider as A, TTSAudio as D, TTSCache as L, TTSClient as H, encodeWAV as P } from "./node/index.js";
+import { D as I, M as N, f as Y, a as X, g as Z, b as q, p as G, s as U } from "../../msEdgeV2-D3w4Xsn-.js";
+import { Component as x, Color as d, hex as a, VGroup as b, SDFText as m, CENTER as S, Rect as g, Animation as C } from "retiam";
 /*! Retiam: noncommercial use only unless separately licensed. See ../../LICENSE and ../../COMMERCIAL_LICENSE.md. Third-party notices: ../../THIRD_PARTY_NOTICES.md and ../../THIRD_PARTY_LICENSES.txt. */
-function Y(c, t = 24e3) {
-  const a = t * 2, r = c.length * 2, n = new ArrayBuffer(44 + r), o = new DataView(n);
-  B(o, 0, "RIFF"), o.setUint32(4, 36 + r, !0), B(o, 8, "WAVE"), B(o, 12, "fmt "), o.setUint32(16, 16, !0), o.setUint16(20, 1, !0), o.setUint16(22, 1, !0), o.setUint32(24, t, !0), o.setUint32(28, a, !0), o.setUint16(32, 2, !0), o.setUint16(34, 16, !0), B(o, 36, "data"), o.setUint32(40, r, !0);
-  let u = 44;
-  for (let h = 0; h < c.length; h++) {
-    const d = Math.max(-1, Math.min(1, c[h])), p = d < 0 ? d * 32768 : d * 32767;
-    o.setInt16(u, p, !0), u += 2;
-  }
-  return n;
-}
-function B(c, t, e) {
-  for (let i = 0; i < e.length; i++)
-    c.setUint8(t + i, e.charCodeAt(i));
-}
-class R {
-  text;
-  phonemes;
-  samples;
-  sampleRate;
-  duration;
-  durationMs;
-  voice;
-  speed;
-  wordBoundaries;
-  rawBuffer;
-  mimeType;
-  _wavBuffer = null;
-  _blob = null;
-  _objectUrl = null;
-  _dataUrl = null;
-  constructor(t) {
-    this.text = t.text, this.phonemes = t.phonemes || "", this.samples = t.samples ?? new Float32Array(0), this.sampleRate = t.sampleRate ?? 24e3, this.voice = t.voice ?? "af_heart", this.speed = t.speed ?? 1, this.wordBoundaries = t.wordBoundaries, this.rawBuffer = t.rawBuffer, this.mimeType = t.mimeType || (t.rawBuffer ? "audio/mpeg" : "audio/wav"), t.durationMs !== void 0 ? (this.durationMs = t.durationMs, this.duration = t.duration ?? t.durationMs / 1e3) : t.duration !== void 0 ? (this.duration = t.duration, this.durationMs = t.duration * 1e3) : (this.duration = this.samples.length > 0 ? this.samples.length / this.sampleRate : 0, this.durationMs = this.duration * 1e3);
-  }
-  /**
-   * Converts raw PCM samples or decodes underlying rawBuffer into a Web Audio [[AudioBuffer]].
-   */
-  async getOrDecodeAudioBuffer(t) {
-    if (this.samples && this.samples.length > 0)
-      return this.toAudioBuffer(t);
-    if (this.rawBuffer)
-      try {
-        const e = await t.decodeAudioData(this.rawBuffer.slice(0));
-        return this.samples = e.getChannelData(0), this.sampleRate = e.sampleRate, e;
-      } catch (e) {
-        console.warn("[TTSAudio] Failed to decode rawBuffer via context.decodeAudioData:", e);
-      }
-    return this.toAudioBuffer(t);
-  }
-  /**
-   * Converts raw PCM samples to a Web Audio [[AudioBuffer]].
-   */
-  toAudioBuffer(t) {
-    const e = t || (typeof window < "u" ? window.__retiam_audio_context : null);
-    if (e) {
-      const s = e.createBuffer(1, Math.max(1, this.samples.length), this.sampleRate);
-      return s.getChannelData(0).set(this.samples), s;
-    }
-    const i = typeof window < "u" ? window.AudioContext || window.webkitAudioContext : null;
-    if (i) {
-      const a = new i().createBuffer(1, Math.max(1, this.samples.length), this.sampleRate);
-      return a.getChannelData(0).set(this.samples), a;
-    }
-    throw new Error("[TTSAudio] Cannot create AudioBuffer: No AudioContext available in this environment.");
-  }
-  /**
-   * Returns an ArrayBuffer containing encoded 16-bit PCM WAV audio.
-   */
-  toWav() {
-    return this._wavBuffer || (this.rawBuffer && this.mimeType.includes("wav") ? this._wavBuffer = this.rawBuffer : this._wavBuffer = Y(this.samples, this.sampleRate)), this._wavBuffer;
-  }
-  /**
-   * Returns a standard WAV or MP3 [[Blob]].
-   */
-  toBlob() {
-    return this._blob || (this.rawBuffer ? this._blob = new Blob([this.rawBuffer], { type: this.mimeType }) : this._blob = new Blob([this.toWav()], { type: "audio/wav" })), this._blob;
-  }
-  /**
-   * Generates a base64 Data URL ("data:audio/...;base64,...").
-   */
-  toDataUrl() {
-    if (!this._dataUrl) {
-      const t = this.rawBuffer ? new Uint8Array(this.rawBuffer) : new Uint8Array(this.toWav());
-      let e = "";
-      const i = t.byteLength;
-      for (let a = 0; a < i; a++)
-        e += String.fromCharCode(t[a]);
-      const s = typeof btoa < "u" ? btoa(e) : Buffer.from(e, "binary").toString("base64");
-      this._dataUrl = `data:${this.mimeType};base64,${s}`;
-    }
-    return this._dataUrl;
-  }
-  /**
-   * Creates or returns a cached Object URL ("blob:...") for this audio.
-   */
-  toObjectUrl() {
-    if (!this._objectUrl)
-      if (typeof URL < "u" && typeof URL.createObjectURL == "function")
-        this._objectUrl = URL.createObjectURL(this.toBlob());
-      else
-        return this.toDataUrl();
-    return this._objectUrl;
-  }
-  /**
-   * Revokes any generated Object URL to release memory.
-   */
-  revokeObjectUrl() {
-    this._objectUrl && typeof URL < "u" && typeof URL.revokeObjectURL == "function" && (URL.revokeObjectURL(this._objectUrl), this._objectUrl = null);
-  }
-  /**
-   * Triggers a browser download or writes out the audio file.
-   */
-  save(t) {
-    const e = this.mimeType.includes("mpeg") || this.mimeType.includes("mp3") ? ".mp3" : ".wav", i = `speech${e}`, s = t ? t.includes(".") ? t : `${t}${e}` : i;
-    if (typeof document < "u") {
-      const a = this.toObjectUrl(), r = document.createElement("a");
-      r.style.display = "none", r.href = a, r.download = s, document.body.appendChild(r), r.click(), setTimeout(() => {
-        document.body.removeChild(r);
-      }, 100);
-      return;
-    }
-    try {
-      const a = globalThis.require ? globalThis.require("fs") : null;
-      if (a) {
-        const r = this.rawBuffer ? Buffer.from(this.rawBuffer) : Buffer.from(this.toWav());
-        a.writeFileSync(s, r);
-      }
-    } catch (a) {
-      console.warn(`[TTSAudio] Unable to save file ${s}:`, a);
-    }
-  }
-  /**
-   * Plays the audio through Retiam's [[AudioManager]] on the dedicated voiceover bus
-   * with automatic BGM ducking and timeline synchronization.
-   */
-  async play(t, e = {}) {
-    let i;
-    return t && (typeof t.playVoiceover == "function" ? i = t : t.audio && typeof t.audio.playVoiceover == "function" && (i = t.audio)), i ? i.playVoiceover(this, e) : (console.warn("[TTSAudio] No active AudioManager provided to play(). Creating standalone audio playback."), this._playStandalone(e));
-  }
-  /**
-   * Standalone playback fallback if no Retiam scene/AudioManager is provided.
-   */
-  _playStandalone(t) {
-    return new Promise((e) => {
-      if (typeof Audio > "u") {
-        e(this.durationMs);
-        return;
-      }
-      const i = new Audio(this.toObjectUrl());
-      i.volume = Math.max(0, Math.min(1, t.volume ?? 1)), i.onended = () => {
-        const s = t.padEndMs ?? 100;
-        setTimeout(() => e(this.durationMs + s), s);
-      }, i.onerror = () => {
-        e(this.durationMs);
-      }, i.play().catch(() => {
-        e(this.durationMs);
-      });
-    });
-  }
-}
-class C {
-  maxSize;
-  cache = /* @__PURE__ */ new Map();
-  constructor(t = 128) {
-    this.maxSize = Math.max(1, t);
-  }
-  static createKey(t, e, i) {
-    return `${e}:${i.toFixed(2)}:${t.trim().toLowerCase()}`;
-  }
-  get(t, e, i) {
-    const s = C.createKey(t, e, i), a = this.cache.get(s);
-    if (a)
-      return this.cache.delete(s), this.cache.set(s, a), a;
-  }
-  set(t, e, i, s) {
-    const a = C.createKey(t, e, i);
-    if (this.cache.has(a))
-      this.cache.delete(a);
-    else if (this.cache.size >= this.maxSize) {
-      const r = this.cache.keys().next().value;
-      r && (this.cache.get(r)?.revokeObjectUrl(), this.cache.delete(r));
-    }
-    this.cache.set(a, s);
-  }
-  has(t, e, i) {
-    return this.cache.has(C.createKey(t, e, i));
-  }
-  clear() {
-    for (const t of this.cache.values())
-      t.revokeObjectUrl();
-    this.cache.clear();
-  }
-  get size() {
-    return this.cache.size;
-  }
-}
-class Z {
-  name = "msedge-tts";
-  _config;
-  _cache;
-  _isBrowser;
-  _nodeClients = [];
-  constructor(t = {}) {
-    this._isBrowser = typeof window < "u" && typeof window.document < "u", this._config = {
-      cacheAudio: t.cacheAudio !== !1,
-      maxCacheSize: t.maxCacheSize ?? 128,
-      msedgeEndpoint: t.msedgeEndpoint || "/api/tts/msedge"
-    }, this._cache = new C(this._config.maxCacheSize);
-  }
-  /**
-   * Microsoft Edge TTS is cloud-hosted and always ready immediately without local weight loading.
-   */
-  get isReady() {
-    return !0;
-  }
-  /**
-   * Initializes provider configuration. No weight downloading required.
-   */
-  async init(t) {
-    t?.msedgeEndpoint && (this._config.msedgeEndpoint = t.msedgeEndpoint), t?.cacheAudio !== void 0 && (this._config.cacheAudio = t.cacheAudio);
-  }
-  /**
-   * Synthesizes text into high-fidelity speech audio using Microsoft Edge neural voices.
-   */
-  async synthesize(t, e = {}) {
-    const i = t.trim(), s = e.voice || M, a = Math.max(0.25, Math.min(4, e.speed ?? 1));
-    if (!i)
-      return new R({
-        text: "",
-        samples: new Float32Array(0),
-        sampleRate: 24e3,
-        voice: s,
-        speed: a
-      });
-    if (this._config.cacheAudio) {
-      const n = this._cache.get(i, s, a);
-      if (n)
-        return n;
-    }
-    const r = this._isBrowser ? await this._synthesizeBrowser(i, s, a, e) : await this._synthesizeNode(i, s, a, e);
-    return this._config.cacheAudio && this._cache.set(i, s, a, r), r;
-  }
-  /**
-   * Node.js execution: directly calls the msedge-tts package in-process.
-   */
-  async _synthesizeNode(t, e, i, s) {
-    const { MsEdgeTTS: a, OUTPUT_FORMAT: r } = await import("msedge-tts");
-    D(a);
-    const n = new a(), o = Math.round((i - 1) * 100), u = o >= 0 ? `+${o}%` : `${o}%`;
-    await n.setMetadata(e, r.AUDIO_24KHZ_96KBITRATE_MONO_MP3, {
-      wordBoundaryEnabled: !0,
-      sentenceBoundaryEnabled: !0
-    });
-    const h = {
-      rate: u
-    };
-    if (s.pitch !== void 0 && (h.pitch = s.pitch), s.volume !== void 0 && s.volume !== null) {
-      if (typeof s.volume == "number")
-        h.volume = s.volume <= 1 && s.volume > 0 ? Math.round(s.volume * 100) : Math.round(s.volume);
-      else if (typeof s.volume == "string") {
-        const b = parseFloat(s.volume);
-        !isNaN(b) && !s.volume.includes("%") && !s.volume.startsWith("+") && !s.volume.startsWith("-") ? h.volume = b <= 1 && b > 0 ? Math.round(b * 100) : Math.round(b) : h.volume = s.volume;
-      }
-    }
-    const { audioStream: d, metadataStream: p } = n.toStream(t, h), f = [], x = [];
-    p && p.on("data", (b) => {
-      try {
-        const A = b.toString(), S = JSON.parse(A);
-        if (Array.isArray(S?.Metadata)) {
-          for (const v of S.Metadata)
-            if (v.Type === "WordBoundary" && v.Data) {
-              const U = (v.Data.Offset || 0) / 1e4, z = (v.Data.Duration || 0) / 1e4, W = v.Data.text?.Text || "";
-              x.push({
-                text: W,
-                offsetMs: U,
-                durationMs: z,
-                boundaryType: "WordBoundary"
-              });
-            }
-        }
-      } catch {
-      }
-    }), await new Promise((b, A) => {
-      d.on("data", (S) => {
-        f.push(S);
-      }), d.on("end", () => {
-        b();
-      }), d.on("error", (S) => {
-        A(S);
-      });
-    });
-    const k = Buffer.concat(f), g = k.buffer.slice(
-      k.byteOffset,
-      k.byteOffset + k.byteLength
-    ), m = F(new Uint8Array(g));
-    return new R({
-      text: t,
-      rawBuffer: g,
-      mimeType: "audio/mpeg",
-      samples: new Float32Array(m.totalSamples),
-      sampleRate: m.sampleRate,
-      duration: m.durationSec,
-      durationMs: m.durationMs,
-      voice: e,
-      speed: i,
-      wordBoundaries: x.length > 0 ? x : void 0
-    });
-  }
-  /**
-   * Browser execution: invokes the Vite development server proxy endpoint.
-   */
-  async _synthesizeBrowser(t, e, i, s) {
-    const a = `${this._config.msedgeEndpoint}/synthesize`, r = await fetch(a, {
-      method: "POST",
-      headers: {
-        "Content-Type": "application/json"
-      },
-      body: JSON.stringify({
-        text: t,
-        voice: e,
-        speed: i,
-        pitch: s.pitch,
-        volume: s.volume
-      })
-    });
-    if (!r.ok)
-      throw new Error(`[MsEdgeTTSProvider] Synthesis failed: HTTP ${r.status} ${r.statusText}`);
-    const n = r.headers.get("X-Audio-Duration-Ms"), o = r.headers.get("X-Word-Boundaries");
-    let u;
-    if (o)
-      try {
-        u = JSON.parse(decodeURIComponent(o));
-      } catch {
-      }
-    const h = await r.arrayBuffer();
-    let d = new Float32Array(0), p = 24e3, f = n ? parseFloat(n) : 0;
-    const x = window.AudioContext || window.webkitAudioContext;
-    if (x) {
-      const k = !!window.__retiam_audio_context, g = window.__retiam_audio_context || new x();
-      try {
-        const m = await g.decodeAudioData(h.slice(0));
-        d = m.getChannelData(0), p = m.sampleRate, f || (f = m.duration * 1e3);
-      } catch (m) {
-        console.warn("[MsEdgeTTSProvider] AudioContext decode deferred, using raw MP3 buffer:", m);
-      } finally {
-        !k && g && typeof g.close == "function" && g.close().catch(() => {
-        });
-      }
-    }
-    if (d.length === 0 || !f) {
-      const k = F(new Uint8Array(h));
-      p = k.sampleRate, f || (f = k.durationMs);
-    }
-    return new R({
-      text: t,
-      rawBuffer: h,
-      mimeType: "audio/mpeg",
-      samples: d,
-      sampleRate: p,
-      duration: f / 1e3,
-      durationMs: f,
-      voice: e,
-      speed: i,
-      wordBoundaries: u
-    });
-  }
-  /**
-   * Progressively streams synthesized speech sentence by sentence.
-   */
-  async *stream(t, e = {}) {
-    const i = e.voice || M, s = Math.max(0.25, Math.min(4, e.speed ?? 1)), a = e.splitPattern || /([.!?;]+(?:\s+|$))/g, r = [], n = t.split(a);
-    for (let h = 0; h < n.length; h += 2) {
-      const d = (n[h] + (n[h + 1] || "")).trim();
-      d.length > 0 && r.push(d);
-    }
-    r.length === 0 && r.push(t.trim());
-    let o = 0, u = 0;
-    for (const h of r) {
-      const d = await this.synthesize(h, {
-        ...e,
-        voice: i,
-        speed: s
-      });
-      yield {
-        text: h,
-        phonemes: "",
-        audio: d,
-        chunkIndex: o++,
-        startTimeMs: u
-      }, u += d.durationMs;
-    }
-  }
-  /**
-   * Returns curated Microsoft Edge neural voices immediately.
-   */
-  getVoices() {
-    return L();
-  }
-  /**
-   * Fetches the complete live catalog of all 320+ Microsoft Edge neural voices.
-   */
-  async fetchVoices() {
-    if (this._isBrowser)
-      try {
-        const t = await fetch(`${this._config.msedgeEndpoint}/voices`);
-        if (t.ok) {
-          const e = await t.json();
-          if (Array.isArray(e))
-            return V(e), e;
-        }
-      } catch {
-      }
-    return P();
-  }
-  /**
-   * Clears cached utterances.
-   */
-  clearCache() {
-    this._cache.clear();
-  }
-  /**
-   * Disposes resources and releases cache.
-   */
-  async dispose() {
-    this.clearCache();
-    for (const t of this._nodeClients)
-      try {
-        t.close();
-      } catch {
-      }
-    this._nodeClients = [];
-  }
-}
-class O {
-  _provider;
-  _config;
-  constructor(t = {}, e) {
-    this._config = t, this._provider = e ?? new Z(t);
-  }
-  /** The active speech provider. */
-  get provider() {
-    return this._provider;
-  }
-  /** Whether the active provider is ready. */
-  get isReady() {
-    return this._provider.isReady;
-  }
-  /** Replaces the active provider instance. */
-  setProvider(t) {
-    this._provider = t;
-  }
-  /** Initializes the active provider. */
-  async init(t) {
-    t && (this._config = t), await this._provider.init(t ?? this._config);
-  }
-  /** Synthesizes text with the active provider. */
-  async synthesize(t, e = {}) {
-    return this._provider.synthesize(t, e);
-  }
-  /** Streams synthesized audio sentence by sentence. */
-  async *stream(t, e = {}) {
-    yield* this._provider.stream(t, e);
-  }
-  /** Synthesizes and plays speech through a Scene or AudioManager. */
-  async speak(t, e, i = {}) {
-    const s = await this.synthesize(e, i), a = await s.play(t, i);
-    return { audio: s, durationMs: a };
-  }
-  /** Returns voices exposed by the active provider. */
-  getVoices() {
-    return this._provider.getVoices();
-  }
-  /** Finds a voice by ID, display name, or partial match. */
-  findVoice(t) {
-    const e = t.toLowerCase();
-    return this._provider.getVoices().find(
-      (i) => i.id.toLowerCase() === e || i.name.toLowerCase() === e || i.id.toLowerCase().includes(e) || i.name.toLowerCase().includes(e)
-    ) ?? N(t);
-  }
-  /** Releases provider resources and cached audio. */
-  async dispose() {
-    await this._provider.dispose();
-  }
-}
-let E = null;
-function w() {
-  return E || (E = new O()), E;
-}
-class T {
-  /** Replaces the global client configuration and restores the built-in provider. */
-  static configure(t) {
-    E = new O(t);
-  }
-  /** Initializes the global speech provider. */
-  static async init(t) {
-    t && this.configure(t), await w().init(t);
-  }
-  /** Replaces the global provider instance. */
-  static setProvider(t) {
-    w().setProvider(t);
-  }
-  /** Active global provider. */
-  static get provider() {
-    return w().provider;
-  }
-  /** Synthesizes text to speech. */
-  static async synthesize(t, e) {
-    return w().synthesize(t, e);
-  }
-  /** Explicit alias for Microsoft Edge neural synthesis. */
-  static async msedge(t, e) {
-    return w().synthesize(t, e);
-  }
-  /** Streams synthesized speech sentence by sentence. */
-  static async *stream(t, e) {
-    yield* w().stream(t, e);
-  }
-  /** Synthesizes and plays speech through a Scene or AudioManager. */
-  static async speak(t, e, i) {
-    return w().speak(t, e, i);
-  }
-  /** Returns the active provider's voices. */
-  static getVoices() {
-    return w().getVoices();
-  }
-  /** Finds a Microsoft Edge neural voice. */
-  static findVoice(t) {
-    return w().findVoice(t);
-  }
-  /** Recommended default voice. */
-  static DEFAULT_VOICE = M;
-  /** Explicit provider-specific alias for the recommended default voice. */
-  static DEFAULT_MSEDGE_VOICE = M;
-}
-class Q extends j {
+class F extends x {
   voice;
   speed;
+  elevenlabs;
   animateSpeaking;
   pulseAmplitude;
   pulseFrequency;
@@ -548,7 +17,7 @@ class Q extends j {
   _baseScaleY = 1;
   _baseScaleZ = 1;
   constructor(t = {}) {
-    super(), this.voice = t.voice ?? "en-US-JennyNeural", this.speed = t.speed ?? 1, this.animateSpeaking = t.animateSpeaking !== !1, this.pulseAmplitude = t.pulseAmplitude ?? 0.05, this.pulseFrequency = t.pulseFrequency ?? 4, this.onSpeechStart = t.onSpeechStart, this.onSpeechEnd = t.onSpeechEnd;
+    super(), this.voice = t.voice, this.speed = t.speed, this.elevenlabs = t.elevenlabs, this.animateSpeaking = t.animateSpeaking !== !1, this.pulseAmplitude = t.pulseAmplitude ?? 0.05, this.pulseFrequency = t.pulseFrequency ?? 4, this.onSpeechStart = t.onSpeechStart, this.onSpeechEnd = t.onSpeechEnd;
   }
   onInit() {
     if (this.entity?.transform) {
@@ -569,20 +38,29 @@ class Q extends j {
    * @param overrides Voice and playback option overrides.
    * @returns Synthesized [[TTSAudioResult]] and total speech duration in milliseconds.
    */
-  async speak(t, e = {}) {
-    const i = this.entity?.scene, s = e.voice ?? this.voice, a = e.speed ?? this.speed;
+  async speak(t, i = {}) {
+    const e = this.entity?.scene, s = i.voice ?? this.voice, c = i.speed ?? this.speed;
     if (this.entity?.transform) {
-      const o = this.entity.transform.getScale();
-      this._baseScaleX = o.x, this._baseScaleY = o.y, this._baseScaleZ = o.z;
+      const h = this.entity.transform.getScale();
+      this._baseScaleX = h.x, this._baseScaleY = h.y, this._baseScaleZ = h.z;
     }
-    const r = await T.synthesize(t, { voice: s, speed: a, ...e });
-    this._isSpeaking = !0, this.onSpeechStart?.(t, r);
-    const n = await r.play(i, e);
-    return this._isSpeaking = !1, this._restoreBaseScale(), this.onSpeechEnd?.(t), { audio: r, durationMs: n };
+    const l = await p.synthesize(t, {
+      voice: s,
+      speed: c,
+      ...i,
+      elevenlabs: {
+        ...this.elevenlabs,
+        ...i.elevenlabs,
+        voiceSettings: { ...this.elevenlabs?.voiceSettings, ...i.elevenlabs?.voiceSettings }
+      }
+    });
+    this._isSpeaking = !0, this.onSpeechStart?.(t, l);
+    const o = await l.play(e, i);
+    return this._isSpeaking = !1, this._restoreBaseScale(), this.onSpeechEnd?.(t), { audio: l, durationMs: o };
   }
-  update(t, e) {
+  update(t, i) {
     if (!(!this.enabled || !this.entity?.transform) && this._isSpeaking && this.animateSpeaking) {
-      const i = t / 1e3 * this.pulseFrequency * Math.PI * 2, s = 1 + Math.sin(i) * this.pulseAmplitude;
+      const e = t / 1e3 * this.pulseFrequency * Math.PI * 2, s = 1 + Math.sin(e) * this.pulseAmplitude;
       this.entity.transform.setScale([
         this._baseScaleX * s,
         this._baseScaleY * s,
@@ -601,14 +79,14 @@ class Q extends j {
     this._isSpeaking = !1, this._restoreBaseScale();
   }
 }
-const q = {
+const v = {
   obsidian: {
     fontSize: 0.38,
-    textColor: l("#F8FAFC"),
+    textColor: a("#F8FAFC"),
     textOpacity: 1,
-    backdropColor: l("#060913"),
+    backdropColor: a("#060913"),
     backdropOpacity: 0.85,
-    strokeColor: new y(0.22, 0.74, 0.97, 0.35),
+    strokeColor: new d(0.22, 0.74, 0.97, 0.35),
     strokeWidth: 0.02,
     cornerRadius: 0.22,
     pillShape: !1,
@@ -619,11 +97,11 @@ const q = {
   },
   pill: {
     fontSize: 0.38,
-    textColor: l("#FFFFFF"),
+    textColor: a("#FFFFFF"),
     textOpacity: 1,
-    backdropColor: l("#080D1A"),
+    backdropColor: a("#080D1A"),
     backdropOpacity: 0.88,
-    strokeColor: new y(0.38, 0.52, 0.75, 0.4),
+    strokeColor: new d(0.38, 0.52, 0.75, 0.4),
     strokeWidth: 0.02,
     pillShape: !0,
     padding: [0.7, 0.28],
@@ -633,11 +111,11 @@ const q = {
   },
   minimal: {
     fontSize: 0.38,
-    textColor: l("#F1F5F9"),
+    textColor: a("#F1F5F9"),
     textOpacity: 1,
-    backdropColor: l("#090D16"),
+    backdropColor: a("#090D16"),
     backdropOpacity: 0.8,
-    strokeColor: y.NONE,
+    strokeColor: d.NONE,
     strokeWidth: 0,
     cornerRadius: 0.16,
     pillShape: !1,
@@ -648,11 +126,11 @@ const q = {
   },
   broadcast: {
     fontSize: 0.38,
-    textColor: l("#FFFFFF"),
+    textColor: a("#FFFFFF"),
     textOpacity: 1,
-    backdropColor: l("#000000"),
+    backdropColor: a("#000000"),
     backdropOpacity: 0.92,
-    strokeColor: y.NONE,
+    strokeColor: d.NONE,
     strokeWidth: 0,
     cornerRadius: 0.08,
     pillShape: !1,
@@ -663,11 +141,11 @@ const q = {
   },
   neon: {
     fontSize: 0.38,
-    textColor: l("#E0F2FE"),
+    textColor: a("#E0F2FE"),
     textOpacity: 1,
-    backdropColor: l("#030712"),
+    backdropColor: a("#030712"),
     backdropOpacity: 0.92,
-    strokeColor: l("#00F0FF"),
+    strokeColor: a("#00F0FF"),
     strokeWidth: 0.025,
     cornerRadius: 0.24,
     pillShape: !1,
@@ -678,11 +156,11 @@ const q = {
   },
   glass: {
     fontSize: 0.38,
-    textColor: l("#FFFFFF"),
+    textColor: a("#FFFFFF"),
     textOpacity: 1,
-    backdropColor: new y(0.05, 0.09, 0.15, 0.65),
+    backdropColor: new d(0.05, 0.09, 0.15, 0.65),
     backdropOpacity: 0.65,
-    strokeColor: new y(1, 1, 1, 0.25),
+    strokeColor: new d(1, 1, 1, 0.25),
     strokeWidth: 0.015,
     cornerRadius: 0.25,
     pillShape: !1,
@@ -693,11 +171,11 @@ const q = {
   },
   light: {
     fontSize: 0.38,
-    textColor: l("#0F172A"),
+    textColor: a("#0F172A"),
     textOpacity: 1,
-    backdropColor: l("#F8FAFC"),
+    backdropColor: a("#F8FAFC"),
     backdropOpacity: 0.92,
-    strokeColor: new y(0.39, 0.45, 0.55, 0.25),
+    strokeColor: new d(0.39, 0.45, 0.55, 0.25),
     strokeWidth: 0.015,
     cornerRadius: 0.2,
     pillShape: !1,
@@ -707,18 +185,18 @@ const q = {
     showBackdrop: !0
   }
 };
-function _(c, t) {
-  return c ? typeof c == "string" ? l(c) : c : t;
+function n(r, t) {
+  return r ? typeof r == "string" ? a(r) : r : t;
 }
-class tt extends H {
+class R extends b {
   _textEntity;
   _backdrop = null;
   _fontSize = 0.38;
-  _textColor = l("#F8FAFC");
+  _textColor = a("#F8FAFC");
   _textOpacity = 1;
-  _backdropColor = l("#060913");
+  _backdropColor = a("#060913");
   _backdropOpacity = 0.85;
-  _strokeColor = new y(0.22, 0.74, 0.97, 0.35);
+  _strokeColor = new d(0.22, 0.74, 0.97, 0.35);
   _strokeWidth = 0.02;
   _showBorder = !0;
   _cornerRadius = 0.22;
@@ -731,14 +209,14 @@ class tt extends H {
   _clearTimeout = null;
   constructor(t = {}) {
     super();
-    const e = t.zIndex ?? 999;
-    this._textEntity = new I("", {
+    const i = t.zIndex ?? 999;
+    this._textEntity = new m("", {
       fontSize: this._fontSize,
       fill: this._textColor,
-      anchor: $
-    }), this._textEntity.paint.z_index = e + 1, this._backdrop = new X(2, 1, { cornerRadius: this._cornerRadius }), this._backdrop.paint.fill = this._backdropColor, this._backdrop.paint.stroke = this._strokeColor, this._backdrop.paint.strokeWidth = this._strokeWidth, this._backdrop.paint.z_index = e, this._backdrop.paint.opacity = 0, this.add(this._backdrop), this.add(this._textEntity), this.paint.z_index = e, this.applyOptions(t);
-    const i = t.bottomOffset ?? -5;
-    this.transform.position.set([0, i, 0]);
+      anchor: S
+    }), this._textEntity.paint.z_index = i + 1, this._backdrop = new g(2, 1, { cornerRadius: this._cornerRadius }), this._backdrop.paint.fill = this._backdropColor, this._backdrop.paint.stroke = this._strokeColor, this._backdrop.paint.strokeWidth = this._strokeWidth, this._backdrop.paint.z_index = i, this._backdrop.paint.opacity = 0, this.add(this._backdrop), this.add(this._textEntity), this.paint.z_index = i, this.applyOptions(t);
+    const e = t.bottomOffset ?? -5;
+    this.transform.position.set([0, e, 0]);
   }
   /** Access the underlying SDFText entity for advanced low-level styling */
   get textEntity() {
@@ -752,8 +230,8 @@ class tt extends H {
    * Applies a complete configuration options object or partial updates.
    */
   applyOptions(t = {}) {
-    const i = { ...t.preset ? q[t.preset] : {}, ...t };
-    return i.fontSize !== void 0 && (this._fontSize = i.fontSize, this._textEntity.fontSize = i.fontSize), i.textColor !== void 0 && (this._textColor = _(i.textColor, this._textColor), this._textEntity.paint.fill = this._textColor), i.textOpacity !== void 0 && (this._textOpacity = i.textOpacity), i.backdropColor !== void 0 && (this._backdropColor = _(i.backdropColor, this._backdropColor)), i.backdropOpacity !== void 0 && (this._backdropOpacity = i.backdropOpacity), i.strokeColor !== void 0 && (this._strokeColor = _(i.strokeColor, this._strokeColor)), i.strokeWidth !== void 0 && (this._strokeWidth = i.strokeWidth), i.showBorder !== void 0 && (this._showBorder = i.showBorder), i.cornerRadius !== void 0 && (this._cornerRadius = i.cornerRadius), i.pillShape !== void 0 && (this._pillShape = i.pillShape), i.showBackdrop !== void 0 && (this._showBackdrop = i.showBackdrop), i.padding !== void 0 && (typeof i.padding == "number" ? this._padding = [i.padding, i.padding] : Array.isArray(i.padding) && (this._padding = [i.padding[0], i.padding[1]])), i.maxWidth !== void 0 && (this._maxWidth = i.maxWidth), i.maxLineChars !== void 0 && (this._maxLineChars = Math.max(15, i.maxLineChars)), i.lineHeight !== void 0 && (this._lineHeight = i.lineHeight), i.bottomOffset !== void 0 && this.setBottomOffset(i.bottomOffset), this._textEntity.text && this._updateBackdropDimensions(this._textEntity.text), this.version++, this;
+    const e = { ...t.preset ? v[t.preset] : {}, ...t };
+    return e.fontSize !== void 0 && (this._fontSize = e.fontSize, this._textEntity.fontSize = e.fontSize), e.textColor !== void 0 && (this._textColor = n(e.textColor, this._textColor), this._textEntity.paint.fill = this._textColor), e.textOpacity !== void 0 && (this._textOpacity = e.textOpacity), e.backdropColor !== void 0 && (this._backdropColor = n(e.backdropColor, this._backdropColor)), e.backdropOpacity !== void 0 && (this._backdropOpacity = e.backdropOpacity), e.strokeColor !== void 0 && (this._strokeColor = n(e.strokeColor, this._strokeColor)), e.strokeWidth !== void 0 && (this._strokeWidth = e.strokeWidth), e.showBorder !== void 0 && (this._showBorder = e.showBorder), e.cornerRadius !== void 0 && (this._cornerRadius = e.cornerRadius), e.pillShape !== void 0 && (this._pillShape = e.pillShape), e.showBackdrop !== void 0 && (this._showBackdrop = e.showBackdrop), e.padding !== void 0 && (typeof e.padding == "number" ? this._padding = [e.padding, e.padding] : Array.isArray(e.padding) && (this._padding = [e.padding[0], e.padding[1]])), e.maxWidth !== void 0 && (this._maxWidth = e.maxWidth), e.maxLineChars !== void 0 && (this._maxLineChars = Math.max(15, e.maxLineChars)), e.lineHeight !== void 0 && (this._lineHeight = e.lineHeight), e.bottomOffset !== void 0 && this.setBottomOffset(e.bottomOffset), this._textEntity.text && this._updateBackdropDimensions(this._textEntity.text), this.version++, this;
   }
   /** Set an aesthetic preset theme */
   setPreset(t) {
@@ -768,39 +246,39 @@ class tt extends H {
     return this._pillShape = t, this._textEntity.text && this._updateBackdropDimensions(this._textEntity.text), this;
   }
   /** Set the inner horizontal and vertical padding around the subtitle text */
-  setPadding(t, e = t) {
-    return this._padding = [t, e], this._textEntity.text && this._updateBackdropDimensions(this._textEntity.text), this;
+  setPadding(t, i = t) {
+    return this._padding = [t, i], this._textEntity.text && this._updateBackdropDimensions(this._textEntity.text), this;
   }
   /** Configure backdrop fill color and opacity */
-  setBackdropColor(t, e) {
-    return this._backdropColor = _(t, this._backdropColor), e !== void 0 && (this._backdropOpacity = Math.max(0, Math.min(1, e))), this._backdrop && (this._backdrop.paint.fill = this._backdropColor, this._textEntity.text && (this._backdrop.paint.opacity = this._backdropOpacity), this._backdrop.invalidate()), this;
+  setBackdropColor(t, i) {
+    return this._backdropColor = n(t, this._backdropColor), i !== void 0 && (this._backdropOpacity = Math.max(0, Math.min(1, i))), this._backdrop && (this._backdrop.paint.fill = this._backdropColor, this._textEntity.text && (this._backdrop.paint.opacity = this._backdropOpacity), this._backdrop.invalidate()), this;
   }
-  setStroke(t, e) {
-    return typeof t == "number" ? (this._strokeWidth = t, this._showBorder = t > 0, e !== void 0 && (typeof e == "string" || e instanceof y) && (this._strokeColor = _(e, this._strokeColor))) : (this._strokeColor = _(t, this._strokeColor), typeof e == "number" && (this._strokeWidth = e, this._showBorder = e > 0)), this._backdrop && (this._backdrop.paint.stroke = this._strokeColor, this._backdrop.paint.strokeWidth = this._strokeWidth, this._backdrop.invalidate()), this;
+  setStroke(t, i) {
+    return typeof t == "number" ? (this._strokeWidth = t, this._showBorder = t > 0, i !== void 0 && (typeof i == "string" || i instanceof d) && (this._strokeColor = n(i, this._strokeColor))) : (this._strokeColor = n(t, this._strokeColor), typeof i == "number" && (this._strokeWidth = i, this._showBorder = i > 0)), this._backdrop && (this._backdrop.paint.stroke = this._strokeColor, this._backdrop.paint.strokeWidth = this._strokeWidth, this._backdrop.invalidate()), this;
   }
   /** Set backdrop border stroke color */
   setStrokeColor(t) {
-    return this._strokeColor = _(t, this._strokeColor), this._backdrop && (this._backdrop.paint.stroke = this._strokeColor, this._backdrop.invalidate()), this;
+    return this._strokeColor = n(t, this._strokeColor), this._backdrop && (this._backdrop.paint.stroke = this._strokeColor, this._backdrop.invalidate()), this;
   }
   /** Set backdrop border stroke width */
   setStrokeWidth(t) {
     return this._strokeWidth = t, this._showBorder = t > 0, this._backdrop && (this._backdrop.paint.strokeWidth = this._strokeWidth, this._backdrop.invalidate()), this;
   }
   /** Configure subtitle typography (font color and world unit font size) */
-  setTextColor(t, e = 1) {
-    return this._textColor = _(t, this._textColor), this._textOpacity = e, this._textEntity.paint.fill = this._textColor, this._textEntity.paint.opacity = e, this.version++, this;
+  setTextColor(t, i = 1) {
+    return this._textColor = n(t, this._textColor), this._textOpacity = i, this._textEntity.paint.fill = this._textColor, this._textEntity.paint.opacity = i, this.version++, this;
   }
   /** Set subtitle font size in world units */
   setFontSize(t) {
     return t > 0 && (this._fontSize = t, this._textEntity.fontSize = t, this._textEntity.text && this._updateBackdropDimensions(this._textEntity.text)), this;
   }
   /** Configure backdrop visual styling (colors, stroke border, corner radius) */
-  setBackdropStyle(t, e, i, s) {
-    return this._backdropColor = _(t, this._backdropColor), e !== void 0 && (this._strokeColor = _(e, this._strokeColor), this._showBorder = !0), i !== void 0 && (this._strokeWidth = i, this._showBorder = i > 0), s !== void 0 && (this._cornerRadius = s, this._pillShape = !1), this._textEntity.text && this._updateBackdropDimensions(this._textEntity.text), this.version++, this;
+  setBackdropStyle(t, i, e, s) {
+    return this._backdropColor = n(t, this._backdropColor), i !== void 0 && (this._strokeColor = n(i, this._strokeColor), this._showBorder = !0), e !== void 0 && (this._strokeWidth = e, this._showBorder = e > 0), s !== void 0 && (this._cornerRadius = s, this._pillShape = !1), this._textEntity.text && this._updateBackdropDimensions(this._textEntity.text), this.version++, this;
   }
   /** Configure subtitle typography */
-  setTextStyle(t, e) {
-    return this._textColor = _(t, this._textColor), this._textEntity.paint.fill = this._textColor, e !== void 0 && e > 0 && (this._fontSize = e, this._textEntity.fontSize = e), this._textEntity.text && this._updateBackdropDimensions(this._textEntity.text), this.version++, this;
+  setTextStyle(t, i) {
+    return this._textColor = n(t, this._textColor), this._textEntity.paint.fill = this._textColor, i !== void 0 && i > 0 && (this._fontSize = i, this._textEntity.fontSize = i), this._textEntity.text && this._updateBackdropDimensions(this._textEntity.text), this.version++, this;
   }
   /** Set maximum characters per line before word-wrapping occurs */
   setMaxLineChars(t) {
@@ -818,14 +296,14 @@ class tt extends H {
    * Computes safe corner radius: clamped so circular arcs never self-intersect,
    * or computes stadium capsule radius when pillShape is enabled.
    */
-  _getEffectiveCornerRadius(t, e) {
-    const i = Math.min(t, e) / 2;
-    return this._pillShape ? i : typeof this._cornerRadius == "number" ? Math.max(0, Math.min(this._cornerRadius, i)) : Array.isArray(this._cornerRadius) ? [
-      Math.max(0, Math.min(this._cornerRadius[0], i)),
-      Math.max(0, Math.min(this._cornerRadius[1], i)),
-      Math.max(0, Math.min(this._cornerRadius[2], i)),
-      Math.max(0, Math.min(this._cornerRadius[3], i))
-    ] : Math.min(0.22, i);
+  _getEffectiveCornerRadius(t, i) {
+    const e = Math.min(t, i) / 2;
+    return this._pillShape ? e : typeof this._cornerRadius == "number" ? Math.max(0, Math.min(this._cornerRadius, e)) : Array.isArray(this._cornerRadius) ? [
+      Math.max(0, Math.min(this._cornerRadius[0], e)),
+      Math.max(0, Math.min(this._cornerRadius[1], e)),
+      Math.max(0, Math.min(this._cornerRadius[2], e)),
+      Math.max(0, Math.min(this._cornerRadius[3], e))
+    ] : Math.min(0.22, e);
   }
   /**
    * Internal geometry reflow: updates backdrop width & height using genuine Rect
@@ -837,35 +315,35 @@ class tt extends H {
       this._backdrop.paint.opacity = 0, this._backdrop.invalidate();
       return;
     }
-    const e = this._textEntity.getGroupLocalAABB(), i = e.isEmpty() ? 0 : e.size.x, s = e.isEmpty() ? 0 : e.size.y, [a, r] = this._padding;
-    let n = i, o = s;
-    if (n <= 0.1 || o <= 0.1) {
-      const p = t.split(`
+    const i = this._textEntity.getGroupLocalAABB(), e = i.isEmpty() ? 0 : i.size.x, s = i.isEmpty() ? 0 : i.size.y, [c, l] = this._padding;
+    let o = e, h = s;
+    if (o <= 0.1 || h <= 0.1) {
+      const f = t.split(`
 `);
-      n = Math.max(...p.map((x) => x.length)) * (this._fontSize * 0.44), o = p.length * (this._fontSize * this._lineHeight);
+      o = Math.max(...f.map((y) => y.length)) * (this._fontSize * 0.44), h = f.length * (this._fontSize * this._lineHeight);
     }
-    const u = Math.max(2.4, Math.min(this._maxWidth, n + a * 2)), h = o + r * 2, d = this._getEffectiveCornerRadius(u, h);
-    this._backdrop.transform.setScale([1, 1, 1]), this._backdrop.width = u, this._backdrop.height = h, this._backdrop.cornerRadius = d, this._backdrop.paint.fill = this._backdropColor, this._backdrop.paint.stroke = this._showBorder ? this._strokeColor : y.NONE, this._backdrop.paint.strokeWidth = this._showBorder ? this._strokeWidth : 0, this._backdrop.paint.opacity = this._backdropOpacity, this._backdrop.invalidate();
+    const _ = Math.max(2.4, Math.min(this._maxWidth, o + c * 2)), u = h + l * 2, k = this._getEffectiveCornerRadius(_, u);
+    this._backdrop.transform.setScale([1, 1, 1]), this._backdrop.width = _, this._backdrop.height = u, this._backdrop.cornerRadius = k, this._backdrop.paint.fill = this._backdropColor, this._backdrop.paint.stroke = this._showBorder ? this._strokeColor : d.NONE, this._backdrop.paint.strokeWidth = this._showBorder ? this._strokeWidth : 0, this._backdrop.paint.opacity = this._backdropOpacity, this._backdrop.invalidate();
   }
   /**
    * Displays a subtitle sentence, formatting lines with word-wrapping and
    * reflowing the backdrop pill with true circular corner arcs.
    */
-  showText(t, e) {
+  showText(t, i) {
     this._clearTimeout && (clearTimeout(this._clearTimeout), this._clearTimeout = null);
-    const i = t.trim();
-    let s = i;
-    if (i.length > this._maxLineChars) {
-      const a = i.split(" "), r = [];
-      let n = "";
-      for (const o of a)
-        (n + " " + o).trim().length > this._maxLineChars ? (n && r.push(n), n = o) : n = n ? n + " " + o : o;
-      n && r.push(n), s = r.join(`
+    const e = t.trim();
+    let s = e;
+    if (e.length > this._maxLineChars) {
+      const c = e.split(" "), l = [];
+      let o = "";
+      for (const h of c)
+        (o + " " + h).trim().length > this._maxLineChars ? (o && l.push(o), o = h) : o = o ? o + " " + h : h;
+      o && l.push(o), s = l.join(`
 `);
     }
-    return this._textEntity.text = s, this._textEntity.paint.opacity = this._textOpacity, this._updateBackdropDimensions(s), this.version++, e && e > 0 && (this._clearTimeout = setTimeout(() => {
+    return this._textEntity.text = s, this._textEntity.paint.opacity = this._textOpacity, this._updateBackdropDimensions(s), this.version++, i && i > 0 && (this._clearTimeout = setTimeout(() => {
       this.clear();
-    }, e)), this;
+    }, i)), this;
   }
   /**
    * Synchronizes this subtitle banner with a [[TTSAudioResult]], displaying the spoken text
@@ -881,7 +359,7 @@ class tt extends H {
     return this._clearTimeout && (clearTimeout(this._clearTimeout), this._clearTimeout = null), this._textEntity.text = "", this._textEntity.paint.opacity = 0, this._backdrop && (this._backdrop.paint.opacity = 0, this._backdrop.invalidate()), this.version++, this;
   }
 }
-class et extends K {
+class T extends C {
   text;
   voice;
   speed;
@@ -892,12 +370,21 @@ class et extends K {
   duckingRatio;
   volume;
   _audioResult = null;
+  _audioTask = null;
+  _synthesisOptions;
   _isStarted = !1;
   _baseScale = [1, 1, 1];
-  constructor(t, e = {}) {
-    super(e), this.text = t, this.voice = e.voice ?? "en-US-JennyNeural", this.speed = e.speed ?? 1, this.pitch = e.pitch, this.subtitles = e.subtitles, this.targetEntity = e.entity, this.pulseAmplitude = e.pulseAmplitude ?? 0.05, this.duckingRatio = e.duckingRatio ?? 0.25, this.volume = e.volume ?? 1;
-    const i = t.trim().split(/\s+/).length;
-    this.duration = Math.max(1e3, i * 420 * (1 / this.speed));
+  constructor(t, i = {}) {
+    super(i), this.text = t, this.voice = i.voice, this.speed = i.speed ?? 1, this.pitch = i.pitch, this.subtitles = i.subtitles, this.targetEntity = i.entity, this.pulseAmplitude = i.pulseAmplitude ?? 0.05, this.duckingRatio = i.duckingRatio ?? 0.25, this.volume = i.volume ?? 1, this._synthesisOptions = {
+      voice: i.voice,
+      speed: i.speed,
+      pitch: i.pitch,
+      volume: i.volume,
+      displayText: i.displayText,
+      elevenlabs: i.elevenlabs
+    };
+    const e = t.trim().split(/\s+/).length;
+    this.duration = Math.max(1e3, e * 420 * (1 / this.speed));
   }
   /**
    * Prepares animation state.
@@ -912,12 +399,15 @@ class et extends K {
    * Synthesizes audio ahead of time and locks exact duration.
    */
   async prepareAudio() {
-    return this._audioResult || (this._audioResult = await T.synthesize(this.text, {
-      voice: this.voice,
-      speed: this.speed,
-      pitch: this.pitch,
-      volume: this.volume
-    }), this.duration = this._audioResult.durationMs), this._audioResult;
+    if (!this._audioResult) {
+      this._audioTask ??= p.synthesize(this.text, this._synthesisOptions);
+      try {
+        this._audioResult = await this._audioTask, this.duration = this._audioResult.durationMs;
+      } finally {
+        this._audioTask = null;
+      }
+    }
+    return this._audioResult;
   }
   update(t) {
     if (this.paused) return !1;
@@ -926,10 +416,10 @@ class et extends K {
       t = -this.delay, this.delay = 0;
     }
     if (this._isStarted || (this._isStarted = !0, this._beginPlayback()), this.elapsed += t, this.targetEntity?.transform) {
-      const e = this.elapsed / 1e3 * 4 * Math.PI * 2, i = 1 + Math.sin(e) * this.pulseAmplitude;
+      const i = this.elapsed / 1e3 * 4 * Math.PI * 2, e = 1 + Math.sin(i) * this.pulseAmplitude;
       this.targetEntity.transform.setScale([
-        this._baseScale[0] * i,
-        this._baseScale[1] * i,
+        this._baseScale[0] * e,
+        this._baseScale[1] * e,
         this._baseScale[2]
       ]), this.targetEntity.version++;
     }
@@ -937,13 +427,13 @@ class et extends K {
   }
   async _beginPlayback() {
     if (this.targetEntity?.transform) {
-      const i = this.targetEntity.transform.getScale();
-      this._baseScale = [i.x, i.y, i.z];
+      const e = this.targetEntity.transform.getScale();
+      this._baseScale = [e.x, e.y, e.z];
     }
     const t = await this.prepareAudio();
     this.subtitles && this.subtitles.syncWithAudio(t);
-    const e = this.targetEntity?.scene;
-    await t.play(e, {
+    const i = this.targetEntity?.scene;
+    await t.play(i, {
       duckingRatio: this.duckingRatio,
       volume: this.volume
     });
@@ -952,38 +442,41 @@ class et extends K {
     this.targetEntity?.transform && (this.targetEntity.transform.setScale(this._baseScale), this.targetEntity.version++);
   }
 }
-async function it(c, t, e = {}) {
-  const i = c?.audio || c;
-  i && typeof i.awaitUnlock == "function" && await i.awaitUnlock();
-  const s = await T.synthesize(t, {
-    voice: e.voice,
-    speed: e.speed ?? 1,
-    pitch: e.pitch,
-    volume: e.volume
+async function M(r, t, i = {}) {
+  const e = r?.audio || r;
+  e && typeof e.awaitUnlock == "function" && await e.awaitUnlock();
+  const s = await p.synthesize(t, {
+    voice: i.voice,
+    speed: i.speed,
+    pitch: i.pitch,
+    volume: i.volume,
+    displayText: i.displayText,
+    elevenlabs: i.elevenlabs
   });
-  return e.subtitles && e.subtitles.syncWithAudio(s), s.play(i, {
-    duckingRatio: e.duckingRatio ?? 0.25,
-    volume: e.volume ?? 1
-  }), c && typeof c.wait == "function" && await c.wait(s.durationMs), s;
+  return i.subtitles && i.subtitles.syncWithAudio(s), s.play(e, {
+    duckingRatio: i.duckingRatio ?? 0.25,
+    volume: i.volume ?? 1
+  }), r && typeof r.wait == "function" && await r.wait(s.durationMs), s;
 }
 export {
-  M as DEFAULT_MSEDGE_VOICE,
-  rt as MSEDGE_VOICES,
-  Z as MsEdgeTTSProvider,
-  q as SUBTITLE_PRESETS,
-  et as Speak,
-  tt as Subtitles,
-  T as TTS,
-  R as TTSAudio,
-  C as TTSCache,
-  O as TTSClient,
-  Q as TTSNarrationComponent,
-  Y as encodeWAV,
-  P as fetchRemoteMsEdgeVoices,
-  N as findMsEdgeVoice,
-  L as getAvailableMsEdgeVoices,
-  ot as getMsEdgeVoicesByLanguage,
-  F as parseMp3Metadata,
-  V as setDynamicMsEdgeVoices,
-  it as speak
+  I as DEFAULT_MSEDGE_VOICE,
+  W as ElevenLabsTTSProvider,
+  N as MSEDGE_VOICES,
+  A as MsEdgeTTSProvider,
+  v as SUBTITLE_PRESETS,
+  T as Speak,
+  R as Subtitles,
+  p as TTS,
+  D as TTSAudio,
+  L as TTSCache,
+  H as TTSClient,
+  F as TTSNarrationComponent,
+  P as encodeWAV,
+  Y as fetchRemoteMsEdgeVoices,
+  X as findMsEdgeVoice,
+  Z as getAvailableMsEdgeVoices,
+  q as getMsEdgeVoicesByLanguage,
+  G as parseMp3Metadata,
+  U as setDynamicMsEdgeVoices,
+  M as speak
 };