Archiving a course at higher speed
Export a slow lecturer at 1.5x so re-watching later is already fast.
Guide
Speed Change makes a video or an audio file faster or slower as a whole: watch a recorded course at 1.5x and keep it that way, replay sports in slow motion, squeeze a 30-second voiceover into 15 seconds. It is the “Speed” tab of the audio/video workbench, with a range from 0.25x to 4x, a slider step of 0.05 and shortcut buttons for 0.25 / 0.5 / 0.75 / 1.25 / 1.5 / 2 / 3 / 4.
Updated 2026-09-093 min read
Speed Change makes a video or an audio file faster or slower as a whole: watch a recorded course at 1.5x and keep it that way, replay sports in slow motion, squeeze a 30-second voiceover into 15 seconds. It is the “Speed” tab of the audio/video workbench, with a range from 0.25x to 4x, a slider step of 0.05 and shortcut buttons for 0.25 / 0.5 / 0.75 / 1.25 / 1.5 / 2 / 3 / 4.
Sound goes through ffmpeg's atempo filter, which changes speed without touching pitch — 2x speech keeps its voice instead of turning into a chipmunk. Since one atempo stage only allows 0.5x–2x, the tool cascades them automatically: 4x becomes atempo=2,atempo=2. The picture uses setpts=PTS divided by the rate to rewrite timestamps, staying aligned with the audio.
<original-name>_2x.mp4 (audio files keep their extension, for example _1.5x.mp3).| Input | Output | Notes |
|---|---|---|
a 60-minute course MP4 at 2x |
a 30-minute MP4 with the voice unchanged |
one atempo=2 stage is enough |
a 10-second sports shot at 0.25x |
40 seconds of slow motion; the sound is stretched but keeps its pitch |
0.25x cascades to atempo=0.5,atempo=0.5 |
a 3-minute podcast MP3 at 1.25x |
an MP3 of 2 minutes 24 seconds |
audio keeps its format, re-encoded with libmp3lame -q:a 2 |
Export a slow lecturer at 1.5x so re-watching later is already fast.
Take a 5-second basketball or golf shot down to 0.25x and capture the key pose with “Video frame capture”.
A 30-second ad line recorded at 34 seconds gets a 1.15x lift that is almost inaudible.
Render a 3-hour MP3 at 1.75x and save an hour of the commute.
No. atempo is a time-stretching algorithm: it keeps the pitch and only changes duration. The chipmunk effect comes from resampling, which is not used here.
Yes. When no audio track is found, the command automatically switches to video only (-an) instead of failing on the missing stream.
A source normally has about 30 frames per second, and at 4x longer each frame stays on screen four times as long; the tool does not interpolate with optical flow, so it advances frame by frame.
Speed changes run locally in the browser through ffmpeg.wasm and files are never uploaded. On first use the engine, about 31 MB, is downloaded from a CDN (the site asset domain or jsDelivr); that request contains neither your audio nor your video.
Updated 2026-09-09
Speed up or slow down 0.25x–4x with audio kept in sync and pitch preserved
Everything runs locally in your browser; files are never uploaded to a server