Audio Trimmer

Voice memos rarely start where the useful part starts. There’s the fumbling to hit record, a few seconds of pocket noise, maybe “okay, testing, is this… okay” before whatever you actually meant to capture begins — and often a similar stretch of dead air or an abrupt stop at the end. Podcast recordings and field recordings have the same problem at a larger scale: a long raw take with the good ten minutes buried somewhere in the middle of forty. This tool exists for exactly that: cutting an MP3 or WAV file down to the start and end points you specify, in seconds, and discarding everything outside that range.

Why this one skips re-encoding

Most of the conversion tools on this site decode a file and re-encode it into something new, which is unavoidable when you’re changing formats. Trimming doesn’t need that. Instead, this tool uses what’s called stream copy: it locates your start and end points in the existing compressed audio and copies the bytes in between directly into a new file, without decoding and re-encoding the audio data at all. The upshot is twofold — it’s fast, since there’s no actual audio processing happening, and it’s lossless, since the audio itself is never touched, only the boundaries of what gets kept. Whatever quality your MP3 or WAV had going in, the trimmed section has exactly that same quality, because it’s literally the same encoded bytes, just fewer of them.

The one thing worth knowing about stream copy is that the cut points land on the nearest natural boundary in the compressed audio, so a trim requested at “12.0 seconds” might land a fraction of a second off from that exact instant. In practice this is inaudible for spoken word or almost anything else — you’d need a very deliberate, sample-accurate edit, like chopping a beat out of a musical loop, to notice it at all.

The situations this solves

Podcasters trimming a raw recording are a natural fit: cutting the pre-roll chatter and the “okay I think we’re done” tail off a full episode take before it goes anywhere near an editing timeline, so the file that gets shared or archived is just the actual content. Field recordists and journalists have a similar need — a phone or recorder left running captures a long stretch of ambient noise around the ten seconds of usable interview audio or environmental sound, and trimming down to just that section makes the file useful without opening a full audio editor.

Voice memos are probably the most everyday case. Someone dictates a note to themselves, and the file has fifteen seconds of throat-clearing before the actual thought starts. Cutting that down before forwarding it to someone else, or before it goes into a transcription tool, saves both people the awkward listen through the beginning. And for musicians, isolating a short loop or sample from a longer WAV recording — a drum hit, a vocal phrase, a field-recorded texture — is a quick way to pull just the useful seconds out without opening a full DAW.

There’s also the more administrative version of this: someone with a stack of old cassette or answering-machine transfers, digitized as one long continuous WAV file with several messages back to back, who just wants to split out one specific message to send to someone. Rather than opening a full editor to isolate thirty seconds out of a forty-minute transfer, entering the two timestamps here and downloading the result is the whole job.

Getting the timing right

Since the start and end points are entered as plain numbers in seconds rather than dragged on a visual timeline, the practical workflow is to play the file once first, in whatever player you already have it open in, and jot down roughly where the section you want begins and ends. A couple of seconds of margin on either side is usually safer than trying to nail the exact moment on the first attempt; because re-running the trim costs nothing beyond the time it takes to process (no re-encoding, remember, so it’s fast), it’s easy to adjust the numbers and try again if the cut lands slightly off.

Local processing, same as the rest of the toolkit

This runs through the same local, WebAssembly build of FFmpeg used by the other audio and video tools here, meaning your recording, whether it’s a voice memo, an interview, or something else entirely, is processed on your device and never uploaded anywhere. That’s worth knowing if the recording is something private, like a personal voice memo or an unreleased track. The only real time cost is on the very first conversion of a session, when the roughly 30MB engine needs to download before anything can run; trims after that, on this tool or any other audio/video tool on the site, start almost instantly since the engine is already cached in your browser.

It’s a narrow tool that does one specific thing well: given a file and two numbers, it hands back exactly the section you asked for, with none of the quality loss that comes from a full re-encode. For anything more involved than a quick trim — reordering sections, mixing multiple tracks, adding fades — you’ll want proper audio editing software. But for cutting dead air off the front and back of a recording, this is faster than opening one.