Audio Editor
Record, edit, mix audio with waveform visualization. Cut, trim, apply effects to audio. Free online audio editor and recorder in browser
Audacity has been the open-source audio editor for 25 years; the modern lightweight version of "trim a podcast clip, normalize loudness, apply noise reduction" can run in your browser without installing anything. This editor handles waveform display, trim/cut/paste, fade in/out, normalization, gain, basic effects (EQ, compression, reverb), and exports to WAV (uncompressed) or MP3/Opus (compressed). For podcast clips, voice notes, and quick edits, browser-based works well; for full multi-track production, install a DAW.
Operations and what each does
- Trim / cut — remove sections. Use generously; tight cuts make recordings sound more professional.
- Fade in / fade out — gradually ramp volume. Standard at the start and end of clips. Without fades, transitions click.
- Normalize — set peak amplitude to a target (typically -0.1 to -1 dBFS to leave headroom). Used after editing for consistent volume across clips.
- Loudness normalization (LUFS) — match perceptual loudness. Different from peak. Streaming services target -14 LUFS for music, -16 to -19 LUFS for podcasts.
- Gain (boost / cut) — change overall level. For increasing quiet sections, watch headroom — clipping is unrecoverable.
- Compression — reduce dynamic range. Makes voice more consistent in level; can be overdone (sounds "flat").
- EQ (equalization) — boost or cut frequency ranges. High-pass at 80-100 Hz removes room rumble from voice recordings.
- Noise reduction — sample a "noise only" section, then subtract that pattern from the rest. Effective for fan / AC noise; less so for variable noise (people talking in background).
Working example: cleaning a voice memo
Input
A 5-minute voice recording with background AC hum
Output
Workflow:
1. High-pass filter at 100 Hz — removes hum and low rumble.
2. Noise reduction:
- Select 1-2 seconds of "silence" (no voice, just background) → "Get Noise Profile"
- Select entire recording → apply noise reduction (default settings).
3. Compression: 3:1 ratio, threshold -20 dB, attack 5 ms, release 100 ms — evens out level.
4. EQ: gentle bump at 2-4 kHz for vocal presence (+1-2 dB).
5. Normalize to -1 dBFS peak.
6. Loudness target -16 LUFS for podcast use.
7. Fade in 100 ms, fade out 200 ms.
8. Export as MP3 192 kbps (typical for spoken word; 320 kbps is overkill).
Result: 5 min, ~7 MB MP3, sounds substantially more professional than raw recording.For podcast-quality recording, the room and microphone matter more than post-processing. A treated room + decent USB microphone (Shure MV7, Rode NT-USB) produces audio that needs minimal cleanup. A bad room + great mic still has reverb and noise that no plugin fully removes.
File format choices
- WAV / AIFF — uncompressed. Largest file size. Use for editing masters and archival; final delivery if file size does not matter.
- FLAC — lossless compression. About 60% the size of WAV; bit-perfect on decode. Use for archival when storage matters.
- MP3 — universal compatibility. 128 kbps OK for voice, 192-256 kbps for music. Older format but still everywhere.
- AAC — newer, ~10-30% smaller than MP3 at equivalent quality. Used on iTunes, YouTube. Less universal than MP3 but supported in every modern player.
- Opus — newer still, even better compression. Default for WebRTC, Discord. Use for streaming where bandwidth matters.
- OGG Vorbis — older free alternative to MP3. Largely replaced by Opus.
When to reach for this tool
- You have a voice memo and want to trim and clean it before sharing.
- You are starting a podcast and need to learn the basic post-production workflow.
- You have an audio recording with background noise you want to reduce.
- You need to convert / re-encode an audio file for compatibility (e.g., WAV → MP3 for email).
What this tool will not do
- It will not handle multi-track recording / mixing. For multi-track work (separate mic + music + sound effects layers), use a DAW (Audacity desktop, Reaper, Logic, Ableton).
- It will not perform forensic audio analysis. Pitch extraction, speech recognition, speaker identification require specialized tools.
- It will not fix every recording. Severe issues (clipped voice, lost frequencies, room echo) are partially fixable but cannot be perfectly reversed.
- It will not match commercial mastering. Mastered tracks for streaming go through specialized engineers with dedicated tools; browser editors get you 80% of the way.
All editing happens in your browser. Audio files stay local; useful for confidential recordings (interview drafts, internal meeting notes, voice memos) that should not pass through a cloud service.
Frequently asked questions
What is LUFS and why does it matter?
LUFS (Loudness Units relative to Full Scale) measures perceptual loudness — what the ear hears, not what the meter shows. Streaming services normalize audio to a LUFS target (Spotify -14, Apple -16, YouTube -14, podcasts -16 to -19). Audio mastered to its own peak level will be turned down by the platform if "louder" than target; the perceived dynamics suffer. Master to platform LUFS targets.
How do I remove someone else's voice from a recording with mine?
You generally cannot reliably. Source separation (separating mixed voices into individual tracks) is a research problem; some AI tools (iZotope RX, Adobe Enhance Speech, Spleeter) attempt it. Results are mixed; for clean recording, capture each voice separately at recording time.
Why does noise reduction make my voice sound underwater?
Too aggressive. Noise reduction subtracts a noise pattern from the whole audio; if the pattern matches frequencies in the voice (or you apply too much reduction), the voice becomes thin / phasey / artifact-ridden. Use minimal reduction — just enough to hear the difference; aim for "noticeable but acceptable".
Should I export as MP3 or WAV?
For final delivery to listeners: MP3 or AAC (smaller, universal). For working master / archival: WAV or FLAC (lossless). For sharing with another producer: WAV preserves edits; MP3 cannot be re-edited losslessly.
How loud should my podcast be?
Target -16 to -19 LUFS integrated. Below that, listeners turn it up and notice background noise. Above that, your audio is loud relative to streaming-normalized music — sounds harsh in playlists. Apple Podcasts and Spotify normalize to -16 / -14 LUFS.
What is "true peak" vs "peak"?
"Peak" is the highest sample value in the digital file. "True peak" considers what happens after digital-to-analog conversion — interpolation can produce values slightly above the original peaks, causing clipping in playback even when the digital peaks were within limits. Master to -1 dB true peak for safety.
Related tools
Sleep sound machine with white, pink, brown noise, rain, ocean waves & nature sounds. Sleep timer with fade-out, binaural beats for deep sleep & focus. Free ambient noise generator
Create sounds with oscillators, harmonics and ADSR envelopes. Generate tones, waves, sound effects. Free online audio synthesizer and tone generator
Chromatic tuner using your microphone. Detect pitch, see nearest note, tune guitar, bass, ukulele, violin. Free online instrument tuner with cent deviation
Professional metronome with tap tempo, time signatures (4/4, 3/4, 6/8), accent patterns, and visual beat indicator. Free online metronome for musicians
Last updated · E-Utils editorial team