Remove Dead Air from Podcast Audio
Dead air is the enemy of a podcast. A guest pauses to think, a host looks for a note, a connection stalls, and listeners drift away. Cutting each gap by hand in an editor takes hours.
This tool does the first pass automatically. The Podcast preset removes pauses longer than 0.4 s and leaves short natural pauses, so the conversation still breathes.
Where it helps
- Interview podcasts: Remove thinking pauses and connection gaps from remote recordings.
- Solo shows: Clean the stops and restarts from a single-host script.
- Clips for social: Make short promotional clips punchy before you add captions.
Tips for the best result
Do a first pass with the Podcast preset, then listen to the result. If a joke or dramatic pause was removed, raise the minimum silence to 0.6 s and run it again.
Process each speaker track separately if you recorded multitrack. This avoids cutting one person's silence while the other is talking.
Which settings should I use?
- Default: -45 dB threshold, 0.1 s minimum silence. A balanced first try for most clean audio.
- Podcast: -40 dB, 0.4 s. Removes long pauses and keeps short natural ones.
- Voice note: -42 dB, 0.25 s. Good for phone recordings and WhatsApp messages.
- AI voice / TTS: -50 dB, 0.15 s. Generated speech is very clean, so a low threshold is enough.
- Lecture: -38 dB, 0.5 s. Handles room noise and long thinking pauses in classes and talks.
Frequently asked questions
What is dead air?
Dead air is unintended silence in a recording, such as long pauses, waiting time or gaps caused by technical problems.
How much shorter will my episode get?
It depends on the speakers. Many conversational recordings shrink by 5 to 15 percent, and heavily paused recordings by more.
Will it remove natural pauses that add drama?
Only gaps longer than your minimum silence. Raise it to keep dramatic pauses.
Should I run it before or after noise reduction?
Run noise reduction first. Cleaner background noise makes silence detection more accurate.
Can I use it with multitrack recordings?
Yes, but process each track on its own, then align them in your editor, or use the combined mixdown if you only need one file.