1. Prepare a short compatible clip
MP4 with H.264 video and AAC audio is the most dependable starting point in current browsers. Choicer Voicer Space accepts a local video up to 200 MB and three minutes, then asks you to select a scene of no more than 60 seconds.
The source file remains on the device. Choosing a file gives the page temporary browser access; it does not upload the video to this site.
2. Mark every dialogue range
Play the video and pause at the beginning of a spoken line. Use the current video time for the start marker, move to the end of the line, and capture the end marker. Add the character name and words so the performer has a clear cue.
Leave enough time for natural breath and delivery, but avoid overlapping ranges unless two characters are intentionally speaking together.
3. Choose how the original audio behaves
Mute during your lines removes the complete original soundtrack only inside the marked dialogue ranges. It is the clearest simple option, but background music and effects inside those ranges will also disappear. Lower original audio keeps the full source track quietly underneath. Keep original audio overlays your take without reduction.
These modes are volume automation, not AI dialogue separation. They cannot isolate speech from music that is already mixed into one track.
4. Record and review each take
Listen to the original timing, start the countdown, and perform while the video plays silently. The waveform confirms that the microphone receives sound. After each line, play the local take and record again if the timing or delivery needs work.
Headphones reduce echo. A quiet room and consistent microphone distance usually improve the result more than aggressive browser processing.
5. Preview, create, and download
Watch the complete dub before export and replace any line that starts late or covers the next beat. Choose Create & Download Video when the timeline is ready. Export runs in real time, so keep the tab visible until the browser finishes and starts the download.
The browser selects WebM or MP4 based on its encoder support. The recovery feature stores timing and voice recordings locally, but not a second copy of the source video.
6. Check rights before sharing
A downloaded file is not automatically cleared for publication. Share only footage, dialogue, music, performances, and characters that you own, licensed, received permission to use, or can otherwise legally reuse.