Create a Project

import { Steps } from ‘@astrojs/starlight/components’;

A project in UltiVoice groups a source video with its settings, transcript, translations, and all rendered outputs. All processing state is saved in the project — you can close and reopen it at any time without losing work.

Create a new project

1. Click New Project in the toolbar (or press Ctrl+N).

2. Enter a project name — this becomes the folder name for outputs. Use descriptive names, for example product-demo-vi-en.

3. Choose an output folder where rendered videos and subtitle files will be saved.

4. Click Create.

The new project opens in the Workspace with the Pipeline tab active.

Import a source video

1. In the Pipeline tab, click Add Source Video (or drag a file into the panel).

2. Select your video file. Supported formats:

   Format	Notes
   MP4	H.264 or H.265, any audio codec
   MKV	Most codec combinations supported
   MOV	QuickTime container; H.264 recommended
   AVI	Legacy support
   WebM	VP8/VP9 video, Opus/Vorbis audio

3. UltiVoice reads the file metadata and displays duration, resolution, frame rate, and audio tracks.

4. If the video has multiple audio tracks, select the one to transcribe in the Audio track dropdown.

File size

There is no hard file-size limit, but videos longer than 2 hours or larger than 20 GB may be slow to import. For very long content, consider splitting into segments using a tool like FFmpeg before importing.

Configure language settings

Source language

Select the language spoken in the video from the Source language dropdown. If you are not sure, choose Auto-detect — Whisper will identify the language from the first 30 seconds of audio. Auto-detect adds a few seconds to transcription but is usually accurate.

Target language(s)

Select one or more languages to dub into. Each target language runs as a separate pipeline pass, producing its own output video. You can add up to 5 target languages per project in a single run.

Dialect / variant

Some languages offer regional variants — for example, Spanish (Latin America) vs Spanish (Spain) for TTS output voice. Select the appropriate variant to improve synthesis naturalness.

Save and continue

Project settings are saved automatically. When you are ready, proceed to Voice Selection to configure the voice for each target language, then start the dubbing pipeline.