DirectoryAudio cleanup
Audio cleanup
Cleanup tools reduce noise, even out levels, strip fillers, or split a mix into stems. They are the right buy when the recording is already done and the room was not. They will not invent a microphone you did not use.
9 tools
- Adobe Podcast
Adobe's speech enhancement tool. Enhance Speech is a one-pass cleanup for noisy voice recordings, on the web.
FreeApproximate
- Alitu
An opinionated podcast studio: record, clean, assemble, write a transcript, and publish, including hosting up to a download cap.
Subscription
- Async
Chat-based editor for video and podcast recordings, formerly Podcastle. It cuts fillers and pauses, cleans the voice, adds captions, and pulls short clips. Recording is still in the product.
FreemiumApproximate
- Auphonic
Batch audio post: leveling, noise and reverb reduction, loudness targets, chapters, and encoding. Paid plans add speech recognition and show notes.
Freemium
- Cleanvoice
Removes filler words, mouth clicks, stutters, and dead air from dialogue recordings, with an edit list you can export.
SubscriptionApproximate
- Descript
Record, transcribe, and edit audio or video by editing the transcript. Studio Sound, filler-word removal, and captions sit in the same project.
Freemium
- iZotope RX
Repair toolkit for dialogue: hum, clicks, clipping, mouth sounds, and spectral editing. AI modules assist. A person still decides.
Mixed
- Krisp
Noise cancellation for calls and recordings. It removes room noise, and on some plans the other people's voices, before the file is edited.
FreemiumApproximate
- Lalal.ai
Stem separation. Splits vocals, music, and other parts so you can remove or remix a bed under a voice.
Usage-based
Questions
What is the difference between enhancement and restoration?
Enhancement, such as Adobe Podcast Enhance, makes speech clearer in one pass. Restoration, such as iZotope RX, is for specific problems: a hum, a click, a clipped word. Use the specific tool when you can hear the problem.
Should I clean audio before transcription?
Light cleanup helps. Heavy noise reduction can smear consonants and make the transcript worse. Transcribe a copy if you are still deciding.
Can stem splitters remove background music?
Lalal.ai and similar tools separate vocals from music with varying success. A voice recorded on top of a loud track will keep artifacts. It is a rescue, not a mix.