Voice & Audio

Last updated: 3 June 2026

When you practise, the app listens to your voice only to measure pitch (svara) in real time — like a tuner. All of this happens inside your browser. Your audio is never recorded, never uploaded, and never sent to us or anyone else.

What actually happens

• When you start a practice or exercise, your browser asks for microphone permission. Nothing is captured until you grant it.
• The microphone signal is analysed locally using the Web Audio API to estimate the fundamental frequency (pitch) of your voice and compare it to the reference svara.
• The pitch curve you see on screen is computed on your device, in the moment.
• When you stop, the live audio is discarded. Nothing is kept.

What does not happen

• No recording is saved on your device or ours.
• No audio is transmitted to our server or to any third party.
• No speech recognition. We do not transcribe or interpret the words you say.
• No voiceprint and no biometric identification. We do not create a unique acoustic signature of your voice and cannot identify you from it. Pitch is a single acoustic measurement (a frequency), comparable to what a guitar tuner reads — not a biometric identifier.

Your control

• You can decline the microphone prompt and still browse, listen, and read everything on the site.
• You can revoke microphone permission at any time in your browser’s site settings.
• The same applies inside the embedded training exercises.

Why we tell you this

Some privacy laws (for example the Illinois Biometric Information Privacy Act, and the EU GDPR’s rules on biometric data) are specifically concerned with voice identification. Because we perform only local pitch detection and create no voiceprint, those biometric provisions do not apply — but we want to be explicit about it. For the broader picture, see our Privacy Policy.