To cut an audio recording, you decode the file in your browser, enter a start and an end time in seconds, and export the selected range as a PCM16 WAV — no upload, no account, no installation. The Audio Cutter works on a single continuous range, which is exactly the operation people mean when they say "trim the first 30 seconds off this interview" or "pull out the quote between minute 4 and minute 6 of this recording." The tool reads your file locally, lets Web Audio decode it, rounds the two timestamps to the nearest sample frame of the decoded buffer, and writes a new WAV that contains only the slice you asked for. Because every step happens inside your current tab, the original recording never leaves your computer. That matters when you are working with a confidential interview, a private voice memo, a podcast draft, or a meeting you only recorded for yourself — uploading it to a remote service would be unnecessary exposure for what is essentially a quick trim and download.
What "Cutting an Audio Recording" Actually Does
A "cut" in this context is the simplest edit you can make to a recording: pick one continuous slice and keep just that slice. You give the tool two numbers — the moment the slice should begin and the moment it should end — and everything outside that range is discarded. The exported file is a brand-new audio file, not a modified copy of the source.
For a voice recording, the typical tasks are easy to describe:
- Trim the silence at the start of an interview before you send it to a transcriber.
- Pull a 20-second quote out of a 45-minute meeting recording.
- Cut the false start or a long pause out of a voice memo.
- Isolate one song from a long recording of a live performance.
These are all single-range cuts. Audio Cutter is built for this exact shape of task, which is why the inputs are just two time fields rather than a multi-track timeline.
Why a Browser-Based Cutter Works Well for Recordings
Most recordings people want to trim are exactly the kind of file they would rather not hand to a random website. A browser-based tool sidesteps three problems that show up with installed editors:
- Privacy: the file is read by your browser, decoded locally, and the output is offered as a download from a temporary local URL. Nothing is sent to a server for processing.
- Friction: there is nothing to download, no codec pack to install, no admin password to enter, and no project file format to learn.
- Speed: for a quick slice on a small file, you can open the page, drop in the recording, type two times, and have a new WAV ready before a desktop editor would have finished launching.
For everyday recordings — voice memos, podcast drafts, meeting audio, lecture captures — this is enough. The next section walks through exactly how to use it.
How to Cut Your Recording Step by Step
The Audio Cutter follows three operating steps, and each one has a small but important detail.
- Choose a browser-decodable audio file. Click the file picker and select your recording. The picker recognizes common MP3, WAV, M4A, AAC, Ogg, WebM, and FLAC names and MIME types. Keep the file under 25 MB, since the tool checks the raw byte size before it tries to decode anything. If the file is mislabeled, encrypted, or only partly downloaded, you will see a clear decode error rather than a silent fallback.
- Wait for the decode and read the page. After the file loads, Web Audio decodes it (the underlying call is described on MDN's decodeAudioData reference). The page then shows the decoded duration, sample rate, channel count, and total frame count. These numbers describe the working buffer, not necessarily the original encoded file — your browser may resample the source to its own AudioContext rate during decode.
- Enter start and end times in moments. Type a non-negative start time and a later end time. End must be greater than start, start cannot sit at the very end of the audio, and end cannot exceed the exact decoded duration. The tool rounds both values to the nearest sample frame; the start frame is included and the end frame is excluded, so a selection that rounds to less than one frame is rejected.
- Cut and export WAV. Click the cut button. The tool takes the decoded samples in your selected range, clamps the floating-point values to the PCM range, and encodes them as interleaved little-endian signed 16-bit PCM inside a 44-byte RIFF/WAVE header.
- Preview the result. Play the exported WAV inside the page and confirm that the spoken words or musical transient line up at both edges. This is the only reliable way to catch a cut that started half a syllable too early or ended half a beat too late.
- Download the new WAV. Click the download link. The temporary URL is released when you cut again, choose a new file, or close the page, so the working data does not linger.
What the Output Looks Like
The file you download is always a freshly encoded uncompressed WAV, even if the source was an MP3 or AAC recording. The original compressed codec, bitrate, tags, album art, cue points, chapters, loudness metadata, and encoder settings are not preserved. Multi-channel recordings stay in their original channel order, up to eight channels.
Because PCM16 WAV is uncompressed, the file can easily be larger than the source MP3 it came from. The size follows a simple formula:
output bytes = (selected frames × channels × 2) + 44
For a worked example, take a 60-second stereo recording decoded at 44,100 Hz:
- Selected frames: 60 × 44,100 = 2,646,000
- Bytes for samples: 2,646,000 × 2 (channels) × 2 (bytes per sample) = 10,584,000
- Header: 44 bytes
- Total: 10,584,044 bytes, roughly 10.09 MB
The tool reports the output duration and the exact file size before you download, so you can confirm the size before committing.
Limits You May Hit on Longer Recordings
Audio Cutter rejects over-limit files with a message rather than silently shortening them. The limits are explicit and worth comparing:
| Limit | Maximum |
|---|---|
| Encoded input file size | 25 MB |
| Decoded duration | 15 minutes |
| Channels | 8 |
| Decoded sample rate | 192 kHz |
| Total channel-samples across all channels | 30,000,000 |
Two details matter here. First, Web Audio may resample the recording during decode, so the rate you see on screen is the working buffer's rate, not necessarily the encoded rate. Second, the decoded-sample limit means a high-rate or multichannel file can reach the boundary before the duration limit does — a 12-minute mono file at 48 kHz already passes the 30 million channel-sample limit and will be rejected, even though it is still under the 15-minute decoded duration limit.
For longer recordings, your only option is to open a dedicated editor offline, since Audio Cutter does not stretch, resample to a target rate, or transcode to fit.
Tips for Clean Cuts on Voice and Music
A few habits make a real difference in the final cut:
- Preview before you download. The page-level preview is the only check that catches a cut that lands mid-word or mid-beat. Do not skip it.
- Add a half-second cushion on each side. If you want the word that starts at 12.0 s, try 11.7 s as the start and let the next edit tighten it. Frame rounding only matters at the millisecond scale, but human timing rarely does.
- Keep the original file. The export is a brand-new WAV with none of the original metadata, so the source recording is your only archive of the original compressed audio.
- Use a dedicated editor when you need fades, level matching, or crossfades. Audio Cutter does not offer resampling, remixing, normalization, fade-in or fade-out, noise removal, pitch shifting, or compression. For voice cleanup or music mastering, a real DAW is the right tool.
- Combine segments in a second step. If you need to splice multiple cut segments into a single file, the Audio Joiner handles 2–10 files in a chosen order.
For routine trims on voice memos, interviews, podcast drafts, and meetings, Audio Cutter is fast, accurate to the sample frame, and keeps the recording on your own machine throughout the process.