Skip to content

VoiceSetup.md

Latest commit

 

History

History
98 lines (76 loc) · 3.85 KB

VoiceSetup.md

File metadata and controls

98 lines (76 loc) · 3.85 KB
Raw

🎤 Web Speech API Setup & Troubleshooting

Browser Support

Browser Status Notes
Chrome/Chromium ✅ Full Support Best performance
Chrome (Android) ✅ Full Support Uses webkitSpeechRecognition
Edge ✅ Full Support Chromium-based
Opera ✅ Full Support Chromium-based
Safari ✅ Full Support macOS & iOS 14.5+
Mobile Safari (iOS) ⚠️ Limited iOS 14.5+ only
Firefox ❌ Not Supported Web Speech API removed in recent versions
Internet Explorer ❌ Not Supported Use a modern browser

The voice button is automatically hidden when the browser does not expose SpeechRecognition or webkitSpeechRecognition.

Requirements

  1. HTTPS Connection — Required for microphone access (http://localhost also works for local dev)
  2. Microphone Permission — Browser will prompt for permission on first use
  3. Stable Internet — Some browsers (Chrome) send audio to cloud services for transcription
  4. Modern Browser — See support table above

How It Works

  1. Tap/click the 🎤 microphone button in the chat input bar
  2. Browser requests microphone access (first time only)
  3. The button turns red and pulses, and a "🎤 Listening…" toast appears
  4. Speak clearly — interim results are streamed into the input box in real-time
  5. Recognition stops automatically when you finish speaking (continuous = false)
  6. Tap/click the button again at any time to stop early
  7. The final transcript is appended to whatever text is already in the input box

Troubleshooting

"Microphone permission denied"

  • Chrome/Edge: Settings → Privacy & security → Site settings → Microphone → Allow for this site
  • Safari (macOS): System Settings → Privacy & Security → Microphone → enable your browser
  • Safari (iOS): Settings → [Browser] → Microphone → Allow
  • Android (Chrome): tap the lock icon in the address bar → Microphone → Allow

"No speech detected"

  • Check your microphone volume in system settings
  • Speak clearly at a normal pace
  • Reduce background noise
  • Try again after a short pause

"Network error"

  • Check your internet connection (Chrome requires cloud access for transcription)
  • Try a different browser

Voice button not visible

  • Your browser doesn't support the Web Speech API — the button is hidden automatically
  • Switch to Chrome, Edge, or Safari

Transcription quality is poor

  • Speak more clearly and at a moderate pace
  • Use a better quality microphone
  • Reduce background noise

Language Support

Defaults to English (en-US). To change the language, edit the recognition.lang line in js/attachments.js:

recognition.lang = 'es-ES'; // Spanish
recognition.lang = 'fr-FR'; // French
recognition.lang = 'de-DE'; // German
recognition.lang = 'ja-JP'; // Japanese
recognition.lang = 'zh-CN'; // Chinese (Simplified)

Privacy & Security

  • Chrome sends audio to Google's cloud for transcription; this is handled entirely by the browser
  • Safari can process speech on-device on supported hardware
  • Quasar AI does not store, log, or transmit voice recordings
  • Check your browser's own privacy policy for details on how it processes audio

Tips for Best Results

✅ Use a quiet environment
✅ Speak at a natural pace
✅ Keep the microphone 6–12 inches from your mouth
✅ Ensure an HTTPS connection
✅ Check the microphone is not muted in system settings

Still Having Issues?

  1. Open browser DevTools (F12 / Cmd+Option+I)
  2. Check the Console tab for errors beginning with Speech recognition error
  3. Try a different browser
  4. File an issue on GitHub with:
    • Browser name & version
    • Error message from the console
    • Microphone type (built-in / external / headset)
    • Environment (quiet/noisy, HTTPS/localhost)