README.md

aiOla TTS Synthesize Vanilla JS SDK

Version: 0.1.0

The aiOla TTS Synthesize Test Client provides a simple interface to convert text into speech using the aiOla Text-to-Speech (TTS) API and download the generated audio file.

Table of Contents

• Features
• Requirements
• Installation
• Usage
  • Initialization
  • Synthesis
  • Generate Downloadable Link
  • Streaming
  • Playback
• Configuration
  • Microphone Configuration
  • Events
• Example Callbacks

---

Requirements

• Browser:
  • Modern browsers (Chrome, Firefox, Edge).
  • Must support AudioWorklet and Web Audio API.
• Server:
  • Connection to Aiola backend with a valid API key.
• HTTPS:
  • Application must be served over https:// or localhost to access the microphone.

---

Installation

1. Download or clone the repository:

   git clone https://github.com/aiola-lab/aiola-js-sdk

2. Place aiola_tts_client.js in your project directory.

---

Code Highlights

Initialization

The TTS client is initialized with the base API URL:

import AiolaTTSClient from './aiola_tts_client.js';

const baseUrl = 'https://tesla.internal.aiola.ai/api/tts';
const ttsClient = new AiolaTTSClient(baseUrl);

Synthesis

The synthesize method is used to generate the audio:

const audioBlob = await ttsClient.synthesize(text, voice);

Generate Downloadable Link

The audio file is made downloadable using:

const audioUrl = URL.createObjectURL(audioBlob);
const link = document.createElement('a');
link.href = audioUrl;
link.download = 'synthesized_audio.wav';

Streaming

The synthesizeStream method is used for real-time streaming:

const audioBlob = await ttsClient.synthesizeStream(text, voice);

Playback

The streamed audio is played directly in the browser using:

audioPlayer.src = URL.createObjectURL(audioBlob);
audioPlayer.play();
### Customization

To change the list of voices, update the \<select\> element in index.html:

```html
<option value="af_bella">Bella</option>
<option value="af_nicole">Nicole</option>