Audio Cutter

Cutting audio without losing quality

Trimming audio sounds like it should be the simplest possible edit — you are removing, not changing. In practice there are a couple of things going on underneath that decide whether your cut sounds clean, and whether it costs you anything in quality.

Why a waveform matters more than a timeline

Plenty of cutting tools give you two number boxes and nothing else. That works if you already know the exact timestamps, and is miserable if you do not — you end up guessing, exporting, listening, and going back.

A waveform turns the problem into a visual one. Speech looks like clusters of spikes with obvious gaps between sentences. Music has a visible structure: the quiet intro, the point where the drums come in, the fade at the end. Silence is flat. Once you can see the shape of the sound, finding the moment you want takes seconds rather than several attempts.

The waveform here is drawn from the decoded audio itself, one peak per pixel column, so what you drag against is the recording rather than a decorative graphic.

The re-encoding question

Here is the part most tools do not explain. When you cut a compressed file such as an MP3, there are two ways to produce the result:

• Copy the compressed data and keep only the frames inside your selection. Nothing is re-compressed, so there is zero quality loss, but you can only cut on frame boundaries — roughly every 26 milliseconds for MP3.
• Decode and re-encode the selection. The cut lands exactly where you put it, at the cost of one more generation of lossy compression.

This tool re-encodes, at 192 kbps for MP3 and M4A. That is a deliberate choice: an exact cut is what people actually want, and one re-encode at a healthy bitrate is inaudible in normal listening. If you need a bit-perfect result — you are archiving, or the audio will be processed further — choose WAV and nothing is compressed at all.

Where cuts go wrong

Two things make a trimmed clip sound amateurish, and both are easy to avoid.

Cutting mid-word or mid-note. The ear is unforgiving about this. Zoom in on the waveform and place your cut in a gap — the flat bit between sentences, or the moment a note has decayed. A cut in silence is invisible; a cut through a vowel sounds broken.

Cutting on a loud sample. If the audio is at full volume at the exact instant the file ends, you get an audible click, because the sound goes from maximum to zero in one sample. Ending on a quieter moment avoids it entirely.

A note on ringtones

Cutting a song down to a ringtone is one of the most common reasons people look for a tool like this, and the cut is the easy part. Getting it onto the phone is what varies:

• Android — copy the file into the Ringtones folder on the device, then pick it in Settings. MP3 is fine.
• iPhone — Apple wants an M4R file placed through a computer, and the process changes with each iOS release. Cut to M4A here and follow whatever the current method is; the audio itself is the same.

Keep it under about thirty seconds. A ringtone that is longer than the time it takes someone to reach their phone is just a song playing in your pocket.

Nothing is uploaded

The file is read from your disk, decoded in your browser, drawn as a waveform, and cut in your device's memory. There is no server in the path at any point, which is why there is no account and no size limit imposed by us.

That matters more for audio than people expect. Voice memos, recorded meetings, interviews and consultations are among the most personal files most of us own, and they routinely get handed to free web tools whose privacy policy nobody reads.