A Windows silence remover that batch-trims dead air and pauses out of your audio and video, right on your own PC.
Drag in a single clip, a stack of episodes, or point it at a folder and it walks every audio and video file inside.
Start with Gentle, Balanced, or Aggressive. The player marks every pause it found, so you see what would be cut before you commit. Need more control, switch to Custom and set the sensitivity and the shortest pause worth cutting.
The output lands next to the original with a _nosilence suffix, or in a folder you choose. Your source files are never touched. Each file gets a short report: length before and after, seconds and percent removed.
The powerful free option here is a Python script or a raw ffmpeg command, and the quick paid one is a plugin bolted onto an editor you may not even own. Voice Silence Remover is neither. You drop your recordings onto the window, press start, and pick them up trimmed. Nothing to install into another program, nothing to script.
Trimming dead air is the dull stretch of editing, and on a hour-long interview it runs sixty to ninety minutes by hand. Voice Silence Remover does the same work in about ten, so the tidying that used to swallow most of an evening is done before your coffee goes cold.
Cloud editors ask you to upload the whole recording, and their terms often let staff listen to it to train their models. Here nothing is uploaded and no account is needed. Unreleased episodes, client interviews, and lecture recordings stay on the machine they were recorded on.
A neural voice-activity model reads the recording in 30 ms frames at 16 kHz and marks where speech actually is. It tells a spoken word apart from room hiss, keyboard noise, or a passing car, which a decibel gate simply cannot do.
Feed it MP3, WAV, M4A, AAC, or FLAC, and MP4, MKV, MOV, AVI, or WMV. Video is trimmed by cutting whole segments, so picture and sound stay lined up instead of drifting out of sync.
Point it at a folder and it works through every file inside, subfolders and all. A hundred short clips or a stack of long takes go through in one unattended run.
Gentle, Balanced, and Aggressive cover most jobs. Custom opens up the parts that matter. You set how sensitive the detection is and the shortest pause worth cutting. You also choose how much breathing room to leave so speech does not sound clipped, and whether to trim the whole file or only its head and tail.
Cuts land on exact frames and the audio follows them, so lips still match the sound after the trim. Variable-framerate clips are handled too, which is where most one-line scripts fall apart.
After each file you get a plain before-and-after: 12:04 became 9:18, twenty-three percent gone. Batch a folder and the numbers add up across the whole run.
Strip the dead air out of an episode before you even open your editor, so the timeline you import is already tight.
Pull the long pauses out of a talking-head take and keep the pace up, without scrubbing the clip by hand frame by frame.
Shrink a two-hour lecture, webinar, or recorded interview down to the parts where someone is actually speaking, which also makes a transcript quicker to read.
Voice Silence Remover
60 Mb
1.0
08/07/26