IN-BROWSER AUDIO SIMULATOR

Voice Analysis Console

Speak into your microphone and watch the pipeline score each 333 ms segment of audio for synthetic-voice indicators. Use Simulate cloned voice to see how the console escalates to a spoken challenge.

  • Uses your microphone after you press Start
  • Audio stays in this browser tab; nothing is uploaded or stored
  • Lightweight in-browser detector, not the production model
Link: IdlePress Start live analysis and allow microphone access to begin.

Synthetic-voice score

DEMO ONLY · SIMULATION

—

No result yet
Analysis time
—
Segments
0
Link
Idle
Ring buffer
0 segments
Drops
0
Resumes
0

Waveform · 16 kHz

333 ms segments

Waveform appears when analysis starts

Spectrogram

Spectrogram appears when analysis starts
Score history appears after the first analysed segment

Explainability

  • Awaiting active speech

Phonemic challenge

Triggers when the smoothed score stays amber or red for 4 consecutive segments. The caller must read a random phrase aloud; a cloned stream cannot respond naturally. Use Simulate cloned voice to trigger it.

Connection details
Session
—
Last segment
—
Reconnect backoff
1000 ms × 2
Max attempts
10 (±20% jitter)

Need to connect high-volume production SIP trunks or carrier media forks?

Launch Enterprise SOC Workspace