Upload a long recording and get back short clips cut around its loudest moments — every one exactly fifteen seconds, faded at both ends, ready for TikTok, Reels and Shorts. Clips, a single mashup and a metadata file arrive in one ZIP.
15.0 s each
All in one file
JSON timestamps
10 s minimum
Energy is the only signal a machine can read reliably without understanding what you said. It finds laughter, emphasis, interruptions and the moment somebody raises their voice — which is where most good clips live.
The recording is measured one second at a time, so every moment gets its own level rather than one average for the whole file.
The loudest moments are taken in order, with at least ten seconds between them so two clips never cover the same passage.
Each clip runs five seconds before the peak and ten after, so the listener hears what led up to the moment, not just the moment itself.
Every clip is the same length, verified after cutting. Near an edge the window slides inward rather than getting shorter.
A short fade at each end, so a clip never starts or stops with a click halfway through a word.
Every clip carries an energy score from 0 to 100 relative to your own recording, so you know which to listen to first.
MP3, WAV, M4A, OGG or FLAC, up to 200 MB or three hours.
Five, ten or fifteen clips. You are told how many qualifying moments were actually found before anything is cut.
Clips, a mashup and a metadata file are produced together, usually in seconds rather than minutes.
Play each clip in the page, then take the ZIP. Your upload is deleted as soon as the clips exist.
The individual clips at 192 kbps, in chronological order.
Every clip back to back — useful for a quick listen or a trailer.
Start, end, peak time and energy score for each clip, ready for scripting.
What each file is, and what the energy score does and does not mean.
Limits. 200 MB or 3 hours per file, ten files per visitor per day, two jobs at once. A 60-minute episode is well inside all of them.
Ten candidate clips in seconds gives you a shortlist to skim instead of scrubbing an hour-long timeline.
Old episodes nobody has time to mine. Run each one, keep what is worth keeping.
Energy peaks usually land on the moment a guest gets animated, which is often the quotable part.
Even if you never publish the clips, the timestamps tell you where to look.
This is the honest limit of the whole idea. Energy finds where something happened; it cannot tell whether that something was interesting. A cough, a laugh at a bad joke and a brilliant one-liner all look identical to a level meter. Treat the output as a shortlist that took seconds instead of an afternoon, and listen before you publish.
The sensitivity setting decides how much of the recording counts as a candidate. On a lively episode with plenty of loud moments, the top ten are the same whichever setting you pick — the strictest and the loosest both find them. It earns its place on quiet or uneven recordings, where a strict setting would come back with only two or three clips.
This tool works on audio only. It does not produce video, subtitles or captions; pair the clips with the subtitle generator if you need burned-in text.
Yes, fifteen seconds each, checked after cutting. Where a peak sits near the start or end of the recording the window slides inward so the length never changes.
It is long enough to carry a complete thought and short enough for every vertical video platform. Nothing on the page depends on it, so it can be changed if you need a different length.
No. There is a minimum of ten seconds between the moments that get selected, so you never end up with two clips of the same sentence.
It is relative to your own recording: 100 is its loudest moment, not an absolute measure. A score of 70 on a calm interview may be more interesting than a 95 on a shouty one.
You are told the real number before anything is cut. Choosing a looser sensitivity widens the pool; if a recording genuinely has no loud moments there is nothing to find.
No, the clips are MP3 audio. You still add the visual layer in your editor or in whichever app you publish from.
Upload an episode, pick how many clips you want and download the ZIP.
Open the Clip Cutter →This page belongs to Bubixo Voice and Speech Tools, where every tool in the category is listed. These are the ones people usually open next.