How to Edit Audio Online Without Installing Any Software

What You Can Do in a Browser (and What You Can't)

The Web Audio API, introduced in modern browsers around 2011, gives JavaScript direct access to audio processing on the device's CPU and DSP. By 2025, this enables a substantial subset of what desktop audio software can do.

Fully supported in-browser:

• Trim and cut to precise timestamps
• Merge multiple files sequentially
• Adjust playback speed while preserving pitch (via phase vocoder algorithms)
• Normalize volume (peak normalization and LUFS loudness normalization)
• Remove silence automatically
• Add fade-in and fade-out with adjustable curves
• Apply equalization (multi-band EQ)
• Dynamic range compression
• Pitch shifting (semitone adjustments)
• Reverb and delay effects
• Format conversion (MP3, WAV, FLAC, AAC, OGG)
• Waveform visualization

Requires AI/server processing:

• Background noise removal (requires ML inference — handled by server with privacy)
• Vocal separation (stem separation requires significant compute)
• Transcription-based editing (audio editing via text transcript)
• Real-time collaboration editing

For the vast majority of audio editing tasks — everything in the first list — browser-based tools are sufficient.

Step 1: Trim and Cut Audio

Trimming is the most common audio editing operation: removing unwanted material from the beginning, end, or middle of a recording.

How to trim:

1. Upload your audio file to the Trim & Cut tool

2. The waveform visualization loads — use this to identify where silence or unwanted sections begin and end

3. Drag the start handle to the point where your content begins (trimming the head)

4. Drag the end handle to where your content ends (trimming the tail)

5. For cuts in the middle: use the split function to mark a region, then delete that region

6. Preview the trimmed result before downloading

Reading the waveform: Tall peaks are loud sections; flat areas near the center line are silence or very quiet audio. Most recordings start with a few seconds of room noise before the actual content — trim this away. Podcasts and interviews often have long pauses you can identify and remove.

Precision tips:

• Zoom in on the waveform to set precise cut points — a 100ms cut error is audible
• Cut at zero-crossings (where the waveform crosses the center line) to avoid clicks at cut points
• For clean music cuts, cut at beat or phrase boundaries to avoid rhythmic jarring

Step 2: Remove Silence Automatically

For long recordings with many pauses — interviews, lectures, voice memos, podcast recordings — manually cutting every silent section is impractical. Automatic silence removal detects silent regions below a threshold and removes them in one pass.

Settings that matter:

• Silence threshold: Audio below this level is considered silence. Set too high, and the tool cuts into quiet speech. Set too low, and background noise is not considered silence. For typical voice recordings in a quiet room, −40 dB to −35 dB is a good starting point.
• Minimum silence duration: Only remove silent sections longer than this length. Setting this to 0.5–1.0 seconds preserves natural speech pauses while removing dead air. Setting it too short creates an unnatural, rushed rhythm.
• Padding: Leave a small amount of silence on each side of a cut (50–100ms). This prevents the audio from feeling abrupt.

Typical results for podcasts:

• 45-minute interview: removes 5–12 minutes of dead air
• Resulting audio sounds more professional without feeling rushed
• Processing time: seconds in-browser

When not to use automatic silence removal: Music with quiet passages (the threshold will cut through pianissimo sections), audio where pauses are meaningful (comedy timing, dramatic speeches), and voice recordings with inconsistent background noise (the noise floor may exceed your silence threshold).

Step 3: Adjust Speed Without Changing Pitch

Playing audio faster saves time; playing it slower aids comprehension or transcription. The challenge is that simply changing the playback speed also changes pitch — speed up a recording 1.5× and voices sound like chipmunks.

Time-stretching separates tempo from pitch. The algorithm (phase vocoder or WSOLA — Waveform Similarity Overlap and Add) analyzes overlapping windows of audio, stretches or compresses them, and reconstructs audio at a different speed with pitch preserved.

Practical speed ranges:

• 0.75× — 25% slower. Useful for understanding fast speech, slowing down music for transcription, or creating a dramatic slow-down effect.
• 1.25× — 25% faster. The maximum speed increase that still sounds completely natural for speech.
• 1.5× — 50% faster. Obvious speed increase but still intelligible. Good for lecture recordings when you know the material.
• 2.0× — Double speed. Intelligible speech but clearly accelerated. Maximum practical speed for podcasts.
• Above 2.0× — Artifacts become significant; speech intelligibility drops.

For music: Speed changes beyond ±20% introduce audible artifacts — transients smear, pitched instruments develop a slightly blurred quality. This is inherent to time-stretching algorithms and is not a tool limitation.

Step 4: Merge Multiple Files

Merging combines multiple audio files into a single file, played sequentially. Common use cases: combining podcast segments, joining multiple interview recordings taken when a recorder was paused, assembling a playlist into a single file, and adding intro/outro music to a recording.

File order matters: The merge tool combines files in the order you arrange them. Drag to reorder before processing.

Cross-fade transitions: A hard cut between two audio files produces an audible click if the audio waveforms don't cross zero at the cut point. A brief cross-fade (50–200ms) prevents this by fading out the end of the first file while fading in the beginning of the second. Use cross-fades between speech segments for seamless joins; use them between music tracks to avoid clicks.

Format handling: When merging files of different formats or sample rates, the merge tool normalizes to a consistent sample rate and bit depth before combining. The output format is your choice — select the format appropriate for your intended use.

Common workflow — podcast assembly:

1. Record intro and outro separately

2. Record main content (possibly in multiple takes)

3. Merge: intro → segment 1 → segment 2 → outro

4. Apply silence removal to remove dead air within segments

5. Normalize loudness

6. Export as MP3 128 kbps

Step 5: Normalize Volume

Normalization adjusts the overall volume of a file so it meets a target level. For distribution, this ensures your audio is neither too quiet nor so loud it causes distortion on listeners' devices.

Peak normalization sets the loudest single sample to a target level (typically −0.1 dBFS — just below the maximum). It prevents clipping but does not reflect how loud the audio sounds perceptually, because a single loud peak can prevent the rest of the audio from being louder even if the average level is low.

LUFS loudness normalization measures integrated loudness across the entire file and matches it to a target. LUFS (Loudness Units Full Scale) is the industry standard for perceptual loudness matching. Streaming platforms apply LUFS normalization themselves:

• Spotify: −14 LUFS
• Apple Music: −16 LUFS
• YouTube: −14 LUFS
• Podcasts: −16 LUFS
• Netflix: −27 LUFS

The practical implication: If your audio is louder than the platform's target (say, −11 LUFS when YouTube targets −14), YouTube will turn it down automatically. If it is quieter (−18 LUFS), YouTube will turn it up — but if the audio is noisy, this also turns up the noise. Mixing and mastering to the target platform's LUFS value before upload gives you full control.

Recommended workflow: Use LUFS normalization with a target of −16 LUFS for podcasts (or your podcast host's specification). For music, target −14 LUFS for streaming-optimized delivery.

Step 6: Export in the Right Format

Format selection at export determines file size, compatibility, and whether quality was preserved. Revisit the format guide for full context; the short rules for common export scenarios:

Podcast: MP3 128 kbps constant bitrate, mono (stereo doubles the file size with no perceptible benefit for speech). This is the podcast industry standard and plays on every podcast app.

YouTube upload: MP3 320 kbps or AAC 256 kbps. YouTube transcodes everything, so uploading high-quality reduces transcoding artifacts.

Music distribution: WAV 16-bit 44.1 kHz (CD standard) for distributors like DistroKid, TuneCore, or CD Baby. They transcode to streaming formats. Submit lossless to give the platform's transcoder the best source material.

Voice memos / notes: M4A AAC 64–96 kbps. Dramatically smaller than WAV for the same speech quality.

Ringtones: iPhone: M4R format, max 40 seconds. Android: MP3, saved to /Ringtones folder.

Archiving raw recordings: WAV or FLAC. Never archive in a lossy format.