KERO VOICE BENCH / HARD-TUNE WORKBENCH
Record a solo vocal or load WAV/MP3, snap up to 60 seconds of monophonic singing into hard-tuned steps, compare the original and processed sound, and export WAV. Free, no sign-up, no installation; audio processing stays in your browser.
Record a short solo vocal, or load an audio file with clear sung notes or sustained vowels.
Microphone recording is available over HTTPS or localhost. Your audio is not uploaded.
Thin lines show detected pitch; the bold stepped line shows the correction target, not a measurement of the processed output.
Load audio to display its pitch graph.
You can start without knowing the key. Set a key and scale only when you want the vocal to follow a song.
Available after pitch analysis.
Switch between equal-length versions and listen for the deliberately stepped pitch movement.
A short guide to the hard-tune effect and this browser workbench.
Record a short vocal or load WAV/MP3, choose the target mode and correction strength, then select PROCESS VOICE. Compare the original and processed audio, and export the result as WAV. No account or installation is required.
Hard tuning pulls detected notes sharply toward scale tones, creating stepped electronic pitch movement. This bench is designed for experimentation, not transparent studio correction.
Leave KEY UNKNOWN selected. The bench can snap pitch to all 12 chromatic notes. Choose a key and MAJOR or MINOR only when matching a song.
No. Audio decoding, pitch analysis, processing, playback, and WAV generation stay in your browser. Site traffic measurement is separate from audio processing.
Short, dry solo vocals with one note at a time work best. Sustained vowels and a clear melody are easier to analyze; accompaniment, harmony, overlapping voices, speech, and strong noise may reduce accuracy.
No. Kero Voice Bench uses offline, non-real-time processing after recording or loading a file. It processes up to the first 60 seconds, then lets you compare the original and processed audio.
LAB NOTE
This technical prototype is designed for solo, monophonic vocals. Accompaniment, multiple voices, speech, or heavy noise may produce unreliable correction.
Recording, loading, analysis, processing, and WAV export happen locally in this browser. The app does not upload your audio. Site traffic measurement is handled separately. The side ad spaces are layout placeholders; no ad-delivery tag is installed.
This is an independent browser prototype. It is not affiliated with or endorsed by any commercial pitch-correction product or brand.