Skip to content
In this guide

Audio and captions

Remove filler words

Find the "um"s and the long "eeeh" sounds in your voice and cut them from the video in one go. You need captions first.

Remove filler words finds the hesitations in your voice and cuts them out of the whole video. It looks for two things:

  • Filler words in the transcript, like “um”, “uh”, “eh” or “hmm”.
  • Held sounds in the audio, like a long “eeeh” or “mmm”, which transcripts often leave out.

It only looks for pure hesitation sounds. Words that people also use as fillers, like “like” or “so”, aren’t treated as filler words.

Captions come first

OnScreen uses the captions to tell a word from a hesitation. With the text, a drawn-out vowel inside a real word, like “sooo”, stays, and words aren’t cut in half. So generate the captions before you start. Recordings without captions are left alone, and if none has them, OnScreen tells you and offers Go to Captions.

The search uses the microphone, or the camera’s sound if there’s no microphone. System audio isn’t checked.

Remove them

  1. Click Remove filler words in the timeline toolbar. It’s the speech bubble icon next to Auto zoom from clicks.
  2. OnScreen checks the voice in every clip and tells you what it found, for example We found 12 filler words, along with their total length in seconds.
  3. Click Remove to cut them, or Cancel to leave the video as it is.

Each clip is split around the cuts, so you’ll see more clips on the timeline. If there’s nothing to cut, OnScreen says We didn’t find filler words in this video.

Bring them back

All the cuts count as a single step, so Ctrl + Z (on Mac ⌘ + Z) brings them all back at once.

Tip: Read through the transcript first. If a real word was written down as “um”, correct it in the Transcript tab of the Captions panel so it isn’t cut.

Next step

Reduce background noise in your microphone.

Can't find what you need? Write to us