Drop in a voice. See what it needs. Open the right tool.
Measure a recording locally, see the strongest signal evidence and open the right repair, transcription, tuning or transformation workflow.
Quick answer: Load one recording into Voice Workbench for a local signal check and up to three useful recommendations, or jump directly to Noise Remover, Speech to Text, Auto-Tune, Vocal Remover or Voice Changer.
Find the next useful move in one listen
Measured signal evidence, not a diagnosis. The source stays on this device unless you explicitly choose an external-model tool.
Find the right tool without reading the whole catalog
Search by a problem, tool name or result. Everything below stays available as static, indexable content.
25 tools, grouped by job.
All Voice & AI tools
Browse all 25 voice toolsRepair, understand, transform and generate
Restore recordings & dialogue
Map a complete transfer, remove narrow impulses, correct reference-assisted pitch drift, build matching ambience and fix focused defects instead of applying one destructive pass.
Old Audio Restoration Studio
A complete local restoration desk that maps common vinyl, cassette and tape-transfer defects, keeps every repair decision visible and exports an aligned result with an open archival receipt.
Open Old Audio Restoration Studio → ⌁Click & Crackle Remover
A local restoration editor that finds very short impulsive discontinuities, exposes every candidate as an editable marker and reconstructs only enabled spans from their immediate clean context.
Open Click & Crackle Remover → ∿Wow & Flutter Corrector
A reference-assisted pitch restoration workspace that measures time-varying frequency error, separates slow wow from faster flutter and applies a bounded inverse timing correction.
Open Wow & Flutter Corrector → ≈Room Tone Maker
A dialogue-repair workspace that finds stable low-variation ambience in a recording and extends the selected texture into an exact-duration room-tone fill.
Open Room Tone Maker → ⌁De-Reverb
A local room-tail reducer that estimates persistent spectral decay after direct sound and suppresses that late energy while protecting stronger attacks and the original timeline.
Open De-Reverb → 💨Online Plosive Remover
A focused dialogue repair tool that detects short low-frequency microphone bursts and applies bounded low-band reduction only around those events.
Open Online Plosive Remover → 👄Mouth Click Remover
A voice de-clicker that finds isolated millisecond-scale impulses and repairs only the accepted spans with local interpolation and edge smoothing.
Open Mouth Click Remover → 🌬️Vocal Breath Remover
A full-duration vocal editor that detects noise-like breaths between voiced phrases and turns them down without deleting time.
Open Vocal Breath Remover → 🔇Noise Remover
A spectral noise-reduction tool that learns or estimates a noise floor and subtracts it from the recording.
Open Noise Remover → 🔌Online Hum Remover (50/60 Hz)
A local mains-hum remover that detects or accepts a 50 Hz or 60 Hz fundamental and attenuates selected harmonic multiples with narrow notch filters.
Open Online Hum Remover (50/60 Hz) → 🔉Online De-Esser
A local high-band dynamics processor that detects strong sibilant energy and reduces the high-frequency component while the detector is active.
Open Online De-Esser → 🚪Online Noise Gate
A local conventional-DSP gate that follows signal level and lowers audio below a closing threshold while opening above a separate threshold.
Open Online Noise Gate →Separate & understand
Extract parts, map a sung range, turn speech into text or move a clear monophonic melody into editable note data.
Vocal Remover
A two-mode vocal separator: instant phase-based DSP for quick karaoke results, plus an optional neural UVR model for a slower separation attempt.
Open Vocal Remover → 🎚️Stem Splitter
A fast heuristic stem splitter that estimates vocals, drums, bass and other content with stereo, frequency and transient cues; it is not an AI separation model.
Open Stem Splitter → 🎧Speech to Text
A transcription page with two modes: browser live dictation and optional local Whisper transcription for uploaded audio.
Open Speech to Text → 🎹Audio to MIDI & Voice to Instrument
A local monophonic audio-to-MIDI sketchpad that follows a clear hummed, sung, whistled or instrumental melody and turns it into editable note events.
Open Audio to MIDI & Voice to Instrument → ↕Vocal Range Finder
A monophonic vocal-pitch map that finds reliable voiced notes over time and summarizes the lowest, highest and median detected pitch without changing the recording.
Open Vocal Range Finder →Generate & transform
Build a vocoder carrier, synthesize speech, match a voice profile you have permission to use, shift formants, build harmony, redesign a voice, correct pitch or arrange spoken audio as music.
Online Vocoder
A real filter-bank vocoder that measures changing energy across a voice or other modulator and applies those band envelopes to a harmonic synth carrier.
Open Online Vocoder → ◌Vocal Formant Shifter
A vocal timbre processor that remaps a short-time spectral envelope upward or downward while aiming to keep the source fundamental pitch and decoded timeline stable.
Open Vocal Formant Shifter → 🗣️Text to Speech
A text reader with fast system voices and an optional Kokoro neural voice that runs in the browser after its model downloads.
Open Text to Speech → 🧬Voice Cloning
A voice-profile matching experiment that measures a short sample and selects speech settings that resemble its pitch and speaking style; it does not clone the speaker’s timbre.
Open Voice Cloning → 🎭Voice Changer
A collection of 14 preset Web Audio effect chains for making a recording sound deeper, brighter, robotic, distorted or stylized.
Open Voice Changer → ♬Vocal Harmonizer
A local three-voice harmonizer that renders independently pitched, leveled and panned copies around one source while keeping the decoded timeline length.
Open Vocal Harmonizer → 🎤✨Auto-Tune
A local vocal pitch-correction tool that uses bounded YIN/CMNDF tracking to move confident monophonic regions toward a selected key and scale.
Open Auto-Tune → 🗣️🎵Speech to Song
A playful speech effect that tracks spoken pitch and combines it with procedural drums, bass and chords.
Open Speech to Song →Choose by outcome
Every card is one direct link. The full explanations remain visible to people, search engines and answer engines.
- I want toRestore a complete old recordingOld Audio Restoration Studio maps clicks, clipping, dropouts, DC, hum and high-frequency noise, lets you keep or ignore repairs and exports Source, Restored, Difference and an open receipt.Open Old Audio Restoration Studio
- I want toRemove narrow clicks and crackleClick & Crackle Remover ranks millisecond-scale impulsive discontinuities, protects wider musical attacks and repairs only enabled markers from immediate clean context.Open Click & Crackle Remover
- I want toCorrect trackable wow and flutterWow & Flutter Corrector follows a stable reference frequency, separates slower and faster drift and applies a bounded inverse timing curve.Open Wow & Flutter Corrector
- I want toBuild a matching room-tone fillRoom Tone Maker ranks stable low-variation ambience, lets you choose the clean region and extends it into a deterministic exact-duration stereo fill.Open Room Tone Maker
- I want toReduce microphone popsPlosive Remover detects short low-frequency bursts and turns down only their low band instead of filtering the whole voice.Open Online Plosive Remover
- I want toRepair isolated mouth clicksMouth Click Remover marks millisecond-scale impulses and repairs accepted spans with bounded interpolation and crossfades.Open Mouth Click Remover
- I want toTurn down breaths, keep timingVocal Breath Remover finds noise-like inhalations between voiced phrases and attenuates them without cutting duration.Open Vocal Breath Remover
- I want toTranscribe speechUse instant live browser dictation or run a local Whisper model on an uploaded file or microphone recording.Open Speech to Text
- I want toSeparate a mixVocal Remover offers instant phase DSP and an optional local neural model; Stem Splitter provides a fast heuristic four-part estimate.Open Vocal Remover
- I want toChange a voiceAuto-Tune adjusts pitch while Voice Changer applies local effect graphs.Open Voice Changer
- I want toBuild a vocoder voiceVocoder transfers the changing frequency-band envelopes of speech onto a harmonic synth carrier for articulated robot and synth-vocal textures.Open Online Vocoder
- I want toControl harsh sibilanceDe-Esser detects strong high-band energy and dynamically reduces it; it does not identify phonemes or replace listening.Open Online De-Esser
- I want toRemove electrical humHum Remover targets a selected 50 Hz or 60 Hz harmonic series with measured narrow notches; use Noise Remover for broadband sound instead.Open Online Hum Remover (50/60 Hz)
- I want toLower sound between phrasesNoise Gate follows level with hysteresis and timed gain changes; it does not remove noise underneath wanted speech.Open Online Noise Gate
- I want toReduce room reflectionsDe-Reverb estimates persistent spectral decay after direct sound and suppresses that late energy while protecting stronger attacks.Open De-Reverb
- I want toBuild parallel vocal harmonyVocal Harmonizer creates three independently pitched, panned and leveled voices around one source while retaining the decoded timeline.Open Vocal Harmonizer
- I want toMap a sung vocal rangeVocal Range Finder tracks reliable monophonic notes over time and reports the lowest, highest and median accepted pitch without changing audio.Open Vocal Range Finder
- I want toShift vocal formants, keep the noteFormant Shifter moves the broad vocal resonance envelope while aiming to preserve the original fundamental pitch and complete timeline.Open Vocal Formant Shifter
Useful browser-audio workflows
- Repair close-mic dialogue
- Online Plosive Remover → Mouth Click Remover → Vocal Breath Remover
- Prepare clean narration text
- Mouth Click Remover → Vocal Breath Remover → Speech to Text
- Clean and transcribe
- Noise Remover → Speech to Text → Audio Trimmer
- Prepare karaoke
- Vocal Remover → Audio Trimmer → Mastering Studio
- Build a character vocal
- Voice Recorder → Voice Changer → Add Reverb
- Polish spoken dialogue
- Noise Remover → Online De-Esser → Loudness Normalizer
- Treat stable mains buzz
- Online Hum Remover (50/60 Hz) → Noise Remover → Audio File Analyzer
- Control room tone between phrases
- Noise Remover → Online Noise Gate → Audio Compressor
- Build and polish a harmony stack
- Voice Recorder → Vocal Harmonizer → Add Reverb
- Find a range before transposing
- Vocal Range Finder → Song Key Finder → Audio Key Changer & Pitch Shifter
- Reshape vocal timbre
- Vocal Formant Shifter → Vocal Range Finder → Loudness Normalizer
- Clean a roomy narration
- De-Reverb → Noise Remover → Loudness Normalizer
- Build a robot vocal texture
- Online Vocoder → Vocal Formant Shifter → Add Reverb
- Patch a dialogue edit with location tone
- Room Tone Maker → Silence Remover → De-Reverb
- Restore a mechanically damaged transfer
- Old Audio Restoration Studio → Wow & Flutter Corrector → BWF Metadata Editor
Limits and privacy
Core browser DSP keeps selected audio on the device and avoids an upload queue. Optional model-based or system-speech modes are disclosed on their own pages. Practical limits come from decoded duration, channel count, algorithm memory and the current device; compressed audio can expand substantially in memory.