Nysos spots the gaps
Using the audio and the transcript, Nysos finds the stretches where the sound stays very low for at least 0.35 seconds and no word is spoken.
Silence removal
The silences in your clips are cut automatically, and carefully
A podcast, an interview or a livestream is full of gaps: a thinking pause, a transition, dead air. On every rendered clip, Nysos automatically removes the real silences, where the sound is very low and nobody is speaking, then re-syncs the captions to the new cut. The result is a tighter clip, close to a jump cut, without you touching a single slider.
How it works
It keeps a margin
Each cut leaves 0.25 seconds of margin on both sides of the silence, and the total removed never exceeds 45% of the clip's length, so the voice comes back naturally.
Captions follow the cut
Once the silences are gone, the captions are re-synced to the new length so each word appears in step with the voice. It all happens automatically during rendering.
The exact rule applied to every clip
A stretch is removed only when two conditions are met at the same time: the sound stays very low for at least 0.35 seconds, and no word is spoken during it. Quiet speech, such as a whispered sentence, is therefore always kept. A pause shorter than 0.35 seconds is kept too.
Around each silence, Nysos keeps 0.25 seconds of margin on both sides, so cuts don't land right up against the words. A ceiling also protects the clip: no more than 45% of its length is ever removed. The captions are then re-synced to the new cut so they stay aligned with the voice.
Why this rule is cautious
Cutting everything that sounds quiet quickly produces a choppy clip where every sentence starts too abruptly. Nysos prefers caution: the double condition (low sound AND no word) protects passages where someone speaks softly, and the 0.25-second margin leaves the breath that comes before or after a sentence. The idea is a tighter clip that still sounds like a person talking.
Short pauses, the kind you use to stress a point or let an idea land, also stay as long as they last under 0.35 seconds. Only genuine dead time goes: a long thinking pause, the wait between two questions, the gap before a follow-up. And if a clip is almost silent, the 45% ceiling stops it from being reduced to almost nothing.
Where it applies
Silence removal is automatic on every rendered clip: the ones you start from the form with a YouTube, Twitch or Kick link or an imported file, the ones Live Watcher produces when a livestream ends, and the ones you regenerate from the clip page, for instance after changing a style.
In Smart Reframe, which works on a video you already have, the video keeps its original pauses. The same goes for a clip whose start and end you adjusted by hand: if you set them yourself in the editor, the cut you chose stays exactly as it is.
What it changes for a podcast, an interview or a livestream
In a video podcast or an interview, the gaps come from hesitations, transitions and moments when the guest searches for an answer. In a livestream, they are often stretches where the streamer reads, waits or watches the game. Removing them tightens the excerpt, without you having to spot each gap one by one.
The result is close to a jump cut: the end of one sentence connects straight to the next, and the animated captions follow the new rhythm. There's nothing to configure, and the clips still come out as vertical 9:16, as usual. It's most useful on long formats, such as podcasts, webinars or streams, where gaps pile up all along the source video.
See also
FAQ
Frequently asked questions
How do I remove silences from a video automatically?
With Nysos there's nothing to do: paste a YouTube, Twitch or Kick link, or import a file, and every rendered clip goes through silence removal. The captions are re-synced to the new cut, and the clip stays 9:16.
What exactly does Nysos cut from a clip?
The real silences, and only those: a stretch is cut only if the sound stays very low for at least 0.35 seconds and no word is spoken. The speech itself stays as it is.
How do I cut the dead air in a podcast without ruining the rhythm?
Nysos leaves 0.25 seconds of margin on both sides of every silence, keeps pauses shorter than 0.35 seconds, and caps the removal at 45% of the clip. Short pauses used for emphasis stay in place, and the captions follow the new cut.
How do I turn silence removal on?
It's already on: silence removal is automatic on every rendered clip, whether you start from a YouTube, Twitch or Kick link, an imported file or Live Watcher. Smart Reframe and clips whose start or end you adjusted by hand keep their original pauses.
Do the captions stay in sync after the cut?
Yes. After the silences are removed, Nysos re-syncs the captions to the new cut so they stay in step with the voice. You pick the style (Nysos Pro, Storytime, Punch, Anime, Casual or Clean) as usual.
Does it work on livestreams and Twitch VODs?
Yes. Clips from a Twitch, Kick or YouTube link, an imported file or Live Watcher all go through silence removal, because it applies to every rendered clip. Smart Reframe and hand-adjusted clips keep their original pauses.