Direct headphone monitoring, full-band WAV recording, high-frequency translation, low-frequency translation,
waveform analysis, recording, browser speech recognition, word-bank matching, and audible word repeat.
Feedback warning: Use headphones or earbuds. A live microphone routed into room speakers can create a sudden, very loud squeal.
Physical limit: This page can translate frequencies that your microphone and sound hardware actually capture.
A normal computer or phone microphone usually cannot record true ultrasound far above its hardware range or true infrasound below its hardware range.
Frequency translation makes captured high or low signals audible; it does not prove that a sound contains speech or reconstruct words that were never captured.
Raw Microphone WaveformUnprocessed input before gate or translation
WAV Recording WaveformShows the PCM signal currently being written to WAV
Massive Word Bank + Live Matcher
Loading word bank…
Loading vocabulary…
Words Heard Right Now
Turn on Word Matcher, then start the microphone. Final recognized phrases will appear here.
The matcher compares browser-recognized speech with the local vocabulary. It cannot prove that random noise, ultrasound, or infrasound contains genuine words.
Matched Word History
No matched words yet.
Mic Off
Monitor Off
Not Recording
Gate Closed
Word Matcher Off
Microphone names normally appear after microphone permission is granted.
Use “Test Ear Speakers” before recording. Output selection requires browser support.
Direct Voice bypasses the recording gate, fixing quiet verbal audio in headphones.
The recommended Full Natural Microphone mode records continuously. Choose translated signal only when you specifically want the processed scanner sound.
0% keeps recording continuously. Raise it only when you intentionally want to reject quiet room noise. This gate never silences Direct Voice monitoring.
A safety limiter is placed before the output, but begin at a low volume.
Raises quiet headphone audio before the safety limiter.
Boosts or lowers the saved natural microphone signal after the gate.
Near-ultrasonic energy around the upper microphone range is mixed downward into an audible band.
Very low captured rumble is mixed upward around this audible frequency.
Higher values reject more uncertain speech-recognition results. A reported confidence of zero is treated as unknown, not automatically rejected.
Matched final phrases are spoken through the selected system output when Audible Repeat is on.
Raw Microphone Input Level0%
Recording Time
00:00
Raw Input
0%
Saved Signal
0%
Low Band 1–200 Hz
0%
High Band 16 kHz+
0%
Recordings
0
Status: Ready. Press Test Ear Speakers, choose Direct Voice, then start the WAV recorder.