Syncfusion react speech to text

Syncfusion React SpeechToText Component

Component Overview

The SpeechToText component enables users to convert spoken words into text using the Web Speech API. This skill helps you implement, customize, and troubleshoot speech recognition in React applications. The main component that captures audio from the user's microphone and converts speech to text in real-time using browser APIs.

Key Capabilities

• Real-time speech recognition
• Multiple language support
• Customizable button and tooltip
• Event-driven architecture
• Programmatic control via methods
• Accessibility support (ARIA labels, keyboard navigation)
• Localization support
• Error handling and recovery

Documentation

Getting Started

📄 Read: references/getting-started.md

• Installation via npm
• Package installation and setup
• Basic component implementation
• CSS imports and theme selection
• First working example
• TypeScript configuration
• Disabling the component (disabled property)

Speech Recognition Features

📄 Read: references/speech-recognition-features.md

• Retrieving transcripts in real-time
• Setting language for recognition
• Managing interim results
• Listening state management with SpeechToTextState enum (Inactive, Listening, Stopped)
• Reading listeningState from event args and component ref
• Handling speech-to-text conversion
• Real-time vs final results

Button and Tooltip Customization

📄 Read: references/button-and-tooltip-customization.md

• Customizing button content and icons
• Icon positioning and styling
• Controlling tooltip visibility (showTooltip property)
• Tooltip configuration and placement (all 12 TooltipPosition values)
• CSS class styling (e-primary, e-success, etc.)
• Button appearance modes
• Responsive button design

Events and Methods

📄 Read: references/events-and-methods.md

• Event handling (created, onStart, onStop, onError, transcriptChanged)
• Correct event argument interfaces (StartListeningEventArgs, StopListeningEventArgs, ErrorEventArgs, TranscriptChangedEventArgs)
• cancel property to prevent listening start
• isInteracted to distinguish user vs programmatic triggers
• errorMessage for human-readable error details
• isInterimResult for interim vs final transcript results
• startListening(), stopListening(), and destroy() methods
• Ref-based component control
• Programmatic listening management

Globalization and Localization

📄 Read: references/globalization-and-localization.md

• Localization with L10n.load()
• Available locale strings and translations
• Language-specific error messages
• RTL support for right-to-left languages
• Accessibility labels and ARIA attributes
• htmlAttributes for custom HTML/ARIA attributes on the button element
• Multi-language interface support

Troubleshooting and Security

📄 Read: references/troubleshooting-and-security.md

• Common issues and solutions
• Browser compatibility matrix
• Microphone permission handling
• Security considerations and best practices
• Privacy and data transmission
• Performance optimization
• Offline fallback strategies

Quick Start Example

import { SpeechToTextComponent, TextAreaComponent, TranscriptChangedEventArgs } from '@syncfusion/ej2-react-inputs';
import { useState } from 'react';
import "@syncfusion/ej2-tailwind3-theme/styles/speech-to-text/index.css";

function VoiceNoteApp() {
  const [transcript, setTranscript] = useState('');

  const handleTranscriptChanged = (args: TranscriptChangedEventArgs) => {
    setTranscript(args.transcript);
  };

  return (
    <div style={{ padding: '20px' }}>
      <h2>Voice Note Recorder</h2>
      
      {/* SpeechToText component with microphone button */}
      <SpeechToTextComponent 
        id="speechToText"
        transcriptChanged={handleTranscriptChanged}
      />
      
      {/* Display transcribed text */}
      <TextAreaComponent
        id="noteArea"
        value={transcript}
        resizeMode="None"
        rows={5}
        cols={50}
        placeholder="Your voice will appear here..."
      />
    </div>
  );
}

export default VoiceNoteApp;

Common Patterns

Pattern 1: Voice Form Input

import { SpeechToTextComponent, TextBoxComponent } from '@syncfusion/ej2-react-inputs';
import { useState } from 'react';

function VoiceForm() {
  const [formData, setFormData] = useState({
    name: '',
    message: ''
  });

  const handleNameTranscript = (args: any) => {
    setFormData(prev => ({ ...prev, name: args.transcript }));
  };

  const handleMessageTranscript = (args: any) => {
    setFormData(prev => ({ ...prev, message: args.transcript }));
  };

  return (
    <div>
      <label>Name (speak):</label>
      <SpeechToTextComponent transcriptChanged={handleNameTranscript} />
      <TextBoxComponent value={formData.name} />
      
      <label>Message (speak):</label>
      <SpeechToTextComponent transcriptChanged={handleMessageTranscript} />
      <TextBoxComponent value={formData.message} />
    </div>
  );
}

Pattern 2: Programmatic Control

import { SpeechToTextComponent } from '@syncfusion/ej2-react-inputs';
import { useRef } from 'react';

function VoiceControlApp() {
  const speechRef = useRef<SpeechToTextComponent>(null);

  const startVoiceInput = () => {
    speechRef.current?.startListening();
  };

  const stopVoiceInput = () => {
    speechRef.current?.stopListening();
  };

  return (
    <div>
      <SpeechToTextComponent ref={speechRef} />
      <button onClick={startVoiceInput}>Start Recording</button>
      <button onClick={stopVoiceInput}>Stop Recording</button>
    </div>
  );
}

Pattern 3: Error Handling

import { SpeechToTextComponent, ErrorEventArgs } from '@syncfusion/ej2-react-inputs';
import { useState } from 'react';

function VoiceWithErrorHandling() {
  const [error, setError] = useState('');
  const [isListening, setIsListening] = useState(false);

  const handleError = (args: ErrorEventArgs) => {
    // args.errorMessage is the human-readable description; args.error is the error code
    setError(args.errorMessage || `Error: ${args.error}`);
  };

  const handleStart = () => {
    setIsListening(true);
    setError('');
  };

  const handleStop = () => {
    setIsListening(false);
  };

  return (
    <div>
      <SpeechToTextComponent 
        onError={handleError}
        onStart={handleStart}
        onStop={handleStop}
      />
      {isListening && <p>🎤 Listening...</p>}
      {error && <p style={{ color: 'red' }}>{error}</p>}
    </div>
  );
}

Key Props

Prop	Type	Description
lang	string	Language for speech recognition (e.g., 'en-US', 'fr-FR')
transcript	string	Current transcribed text
allowInterimResults	boolean	Show real-time results (default: true)
listeningState	SpeechToTextState	Current listening state (Inactive, Listening, Stopped)
buttonSettings	ButtonSettingsModel	Customize button appearance and content
tooltipSettings	TooltipSettingsModel	Configure tooltip display
showTooltip	boolean	Whether to display the tooltip on hover (default: true)
cssClass	string	Apply CSS classes for styling
disabled	boolean	Disable all component interaction (default: false)
htmlAttributes	{ [key: string]: string }	Additional HTML attributes (ARIA, data-*, etc.) for the root button element
locale	string	Localization language code
enableRtl	boolean	Enable right-to-left layout
enablePersistence	boolean	Persist component state between page reloads via localStorage

Event Handlers

Event	Args	Description
created	-	Fired when component is initialized
onStart	StartListeningEventArgs	Fired when speech recognition begins. Args: cancel, event, isInteracted, listeningState, name
onStop	StopListeningEventArgs	Fired when speech recognition ends. Args: event, isInteracted, listeningState, name
onError	ErrorEventArgs	Fired when an error occurs. Args: error, errorMessage, event, name
transcriptChanged	TranscriptChangedEventArgs	Fired when transcription updates. Args: transcript, isInterimResult, event, name

Methods

Method	Description
startListening()	Begin speech recognition programmatically
stopListening()	Stop speech recognition programmatically
destroy()	Destroy the component instance and release all resources

Troubleshooting

Microphone permission denied

Solution: Check browser permissions settings and allow microphone access in security settings

Speech not recognized

Solution: Check microphone volume, speak clearly, verify correct language setting

Component not rendering

Solution: Ensure CSS imports are included and license key is registered

Browser not supported

Solution: Check if browser supports Web Speech API (Chrome, Edge, Safari support it)

Related Components

• TextArea: For displaying transcribed text
• TextBox: For input fields with voice capabilities
• Button: For custom voice control buttons
• Tooltip: For contextual help on voice features