

Fourier — see your sound, then fix it.
Fourier is a native macOS audio-restoration and editing app, built like a professional spectral-repair studio. A precise waveform sits alongside a full-colour spectrogram, so you can both hear and see every problem before you touch it.
Circle a cough, a hum or a click right on the spectrogram and rebuild it from the audio around it. Beneath the surface sit 47 modules — de-noise, de-hum, de-click, EQ, dynamics, mastering, a Melodyne-style notes editor, and a full on-device AI toolkit — every one non-destructive, previewable, and A/B-comparable.
Everything runs locally on your Mac's GPU and Neural Engine. Nothing is uploaded, big files stream with bounded memory, and your undo history survives across sessions.
Error: No FastSpring product path found for this product.
Apple Silicon Mac
macOS 14 (Sonoma) or later
~2 GB free (plus space for optional AI models)
Not compatible with Intel Macs
On-device · one-time purchase · 2 machine seats · free updates.
If you can see it, you can fix it. Fourier turns invisible problems — hum, hiss, clicks, sibilance, room reverb — into shapes you can point at on the spectrogram, then heals exactly what you circled. Podcast dialogue, music, restoration: one surface, every problem, nothing uploaded.
Paint problems out of the picture
When you can see a hum, a click or a sibilant cloud on the spectrogram, you can select it and repair it directly — box, frequency-band and freehand lasso tools carve out any region, then Fourier rebuilds it from the surrounding audio.
Before
Healed
A 60 Hz hum line, a click transient and a noise cloud — selected on the left, gone on the right. Drag the slider.
Hear Fourier — before & after
Placeholder players — replace the Sonaar albums="" IDs with your own before/after demo tracks.
See your sound. Then fix it.
A precise waveform, a full-colour spectrogram, 47 modules and on-device AI — one app, one engine, zero uploads.
Forty-seven modules. One window.
Every repair and effect lives in a searchable launcher and opens as a floating panel with the same anatomy — presets, controls, a live feedback line and a commit footer. Tap a category to filter.
Spectral Repair
Interpolate
Spectral De-noise
De-reverb
De-hum
De-click
De-crackle
De-clip
Breath Control
De-plosive
De-ess
Phase
Azimuth
Distance
Spectral Recovery
AI Speech De-noise
EQ Filters
Dynamics
Reverb
Time & Pitch
Harmonic Clarity
Sound Enhance
Voice Enhance
Match
Center Extract
Stereo Width
Plugin Host
Gain
Loudness Control
Fade
Trim Silence
Resample
Channels
Dither
Signal Generator
Music Rebalance
Voice Isolate
Dialogue Rebalance
Prompt Isolate
Transcribe
Voice Rebuild
Text to Speech
AI Retouch
Generate
Repair Assistant
Music Tailor
Loop Creator


Any tool, scoped to what you circled
The spectral selection is a universal scope. Draw a time-frequency region once and confine any module to just that box, band or lasso — feather the edges, snap to a harmonic, and preview against the original before you commit.


Studio-grade AI, on your Mac
Transcription, speech de-noise, source separation, note detection and generative repair — all accelerated by your Mac's GPU and Neural Engine through MLX. Split a finished mix into stems, run a broadcast voice chain in one pass, edit speech like a document, or generate audio from a text prompt. Models download once, then run offline — nothing is ever uploaded.
Retune a performance on a piano roll
The Notes view turns audio into editable notes you can retune, retime, split and reshape, Melodyne-style. Mono, Poly and Piano detection handle a solo vocal or a dense chord — with snap-to-scale correction, vibrato, drift and per-note intensity, plus MIDI and per-note WAV export.

App features
- See & hear: synced waveform + full-colour spectrogram
- Spectral editing — box / band / lasso, heal · attenuate · boost
- 47 modules: Repair · Utility · Process · AI
- Module Chain + Batch processing across folders
- Melodyne-style Notes editor (mono & poly) — MIDI export
- On-device AI — transcription, de-noise, separation, generate
- Persistent undo history + big-file streaming
- Non-destructive: preview, A/B and undo every step
Technical specs
- macOS 14+ · Apple Silicon · direct (not App Store)
- 8/16/24/32-bit & float · 44.1–192 kHz · mono/stereo/surround
- Import: WAV · AIFF · CAF · MP3 · AAC · FLAC · Opus · +ffmpeg
- Export: WAV · AIFF · CAF · FLAC · ALAC · AAC · MP3 · Opus
- Peak · true-peak · RMS · LUFS metering · TPDF dither
- AI: MLX · whisper.cpp · Demucs/RoFormer · Stable Audio 3 · ACE-Step
- Plugin Host (Audio Unit) · headless CLI · SRT subtitles
- One-time purchase · 2 machine seats · free updates
Assemble a workflow. Run it once. Run it everywhere.
Stack de-hum, de-noise, EQ and loudness into an ordered Module Chain and apply it as one undoable pass — then point it at an entire folder and let Batch process every file unattended.


Videos
System requirements
Apple Silicon Mac (M1 or later) running macOS 14 (Sonoma) or later. Not compatible with Intel Macs. About 2 GB free for the app, plus additional space for optional on-device AI models (downloaded on demand). An internet connection is needed once to download models and for licence activation; all processing then runs offline on your machine.
Error: No FastSpring product path found for this product.
Secure checkout by FastSpring — tax & VAT handled. Instant delivery: serial + download.
