FluffyAudio FFT Spectral Audio Editor
Spectral audio editor with 50+ modules and on-device AI
Start new thread

FluffyAudio FFT Spectral Audio Editor - Spectral audio editor with 50+ modules and on-device AI

by•
Repair. Rebuild. Reimagine. FluffyAudio FFT is a spectral audio editor for Mac and Windows. It shows audio as a waveform and a spectrogram: you see a problem before you fix it. Draw a box around a cough or a click and heal it. More than 50 modules in one app — repair, mastering, stems, AI voices, audio to MIDI, speech edited as text. Every engine, AI models included, runs on your own machine. Nothing uploaded. We make sample libraries, so we fix audio all day. FFT is the editor we always wanted.

Add a comment

Replies

Best
Maker
📌
Hi Product Hunt, We're Paolo and Olmo, friends, musicians and co-founders of FluffyAudio, a company specialized in professional sample libraries for music and composer. We make sample libraries, which means we fix audio all day: a click in a take, a noisy room, an air conditioner humming behind a beautiful performance, a singer's slip nobody will re-record. The problem was never one tool, it was the pile. A restoration suite for the clicks, a second app for pitch and timing, a third for stems or a web service for the AI that politely asks you to upload your unreleased master. Three subscriptions and two uploads to fix thirty seconds of music. That's why we build FFT, the editor we always wanted. FluffyAudio FFT is built the way professional spectral editors are: your audio as a waveform and as a high-resolution spectrogram, so you can see the problem before you fix it. E.g. draw a box around a cough, a click, a siren, then heal it. Then everything else is in the same app: more than 50 modules across repair, processing and mastering, stem separation, and an AI suite. Speech de-noise, voice isolation, transcription, audio to MIDI, generative inpaint and extend, and speech you edit like a document and rebuild in the same voice. There's also a notes editor for pitch and timing, chord detection with a PDF chart for the band, batch processing... everything you need for your audio. The part we care about most: every engine runs on your machine, AI models included. Nothing is uploaded, no account, no credits, no per-minute meter. The trade-off is honest: yes, a datacenter can out-run your laptop, but your unreleased material never leaves the room, and the app still works the day a service shuts down. Free 15-day trial, no card. We'll be here all day, and we'd rather have the hard questions than the nice ones. So: what's the audio-file you gave up on? Tell us what's wrong with it and we'll tell you what we'd do, spectrogram and all.