Video Dubbing

Video Dubbing Overview

This page helps you understand the real NativeVid dubbing workflow first, then decide whether to continue with Quick Start, the FAQ, or plugin docs.

On this page

  • Build a clear overall understanding of the full dubbing workflow.
  • Focuses on asset preparation, role voices, review, and final export.
  • Best read together with Quick Start for a faster ramp-up.

What is video dubbing in NativeVid?

Video dubbing in NativeVid is not just “reading subtitles aloud”. It is a full localization production chain: first match videos and subtitles, then complete terminology review and subtitle translation when needed, and continue into role setup, voice selection, batch voice generation, and line-by-line review, and finally re-compose the new voices with the video to export a new final video file and subtitle file.

🔒 Privacy boundary: Video files, subtitle files, intermediate dubbing assets, and final exports all remain local; only the translation or TTS requests you actively use go over the network, and original media assets are not uploaded to NativeVid servers.

What real features does this section cover?

📁 Asset matching

After you choose a media folder, the system automatically matches videos and subtitles by filename. Unmatched files can be assigned manually, and duplicate content is moved into the _duplicates folder to reduce mistakes in batch jobs.

🌍 Translation branch

You can choose “translate before dubbing”, or skip translation directly when subtitles are already in the target language; if terminology review is enabled, the workflow first enters a terminology confirmation step.

🎭 Role assignment

Supports a task-level role library, line-by-line role assignment, drag-to-label actions, and shortcut-based switching, which works well for narration, interviews, courses, and other multi-speaker scenarios.

🎙️ Voice selection

Each role that has subtitle lines assigned can choose its own target-language voice, with keyword search, gender and age filters, and voice preview, while also distinguishing free and paid voice sources.

🎵 Background music separation

When creating a job, you can enable background music separation and choose to keep drums, bass, other accompaniment, guitar, piano, and more; the new dub is layered back during composition, which is useful when you want to preserve the original atmosphere.

🔁 Batch generation and retry

The system generates dubbed lines in batch by subtitle row, and the review page shows progress plus success or failure status for each file, with support for batch retry of failed items and single-line regeneration.

🎬 Audio-video alignment and video composition

After dubbing is complete, the system aligns generated audio against subtitle timing and clip duration, then outputs the final video and aligned subtitles; the preview page supports subtitle toggles, playback speed adjustment, and subtitle-position dragging for final checks.

Typical use cases

Courses and knowledge videos

You already have the video and subtitles and want to create multilingual teaching content quickly while preserving background music or ambient sound.

Interview and podcast clips

A single video contains multiple speakers, so you need to identify roles first and then assign voices for the target-language version.

Batch processing for a series

Put multiple episodes from the same series or course into one folder and process them consistently through auto-matching and batch flow.

Actual workflow

In the current NativeVid implementation, video dubbing usually goes through these pages and steps:

  1. Asset matching: Choose the folder that contains videos and subtitles. The system prioritizes exact filename matches and lets you complete unmatched items manually.
  2. Translation settings: Choose the source and target languages. If subtitles are already in the target language, you can skip translation directly; you can also enable terminology review and background music separation when needed.
  3. Role setup: Create roles and assign subtitle lines to them. You can move on only after all lines are assigned.
  4. Role voices: Choose and save a voice for every participating role, then start the batch dubbing job.
  5. Dubbing review: Check generation progress for each file, preview individual lines, and revise or regenerate a single line when needed.
  6. Alignment and video composition: After all dubbing is complete, the system aligns audio length to subtitle timing, composes the video, and lets you inspect the final video, subtitle file, and output directory from the preview page.
💡 One key branch: If you enable “Skip Translation” when creating the task, NativeVid uses the uploaded subtitles directly as target-language text, so the workflow goes straight to Role Assignment instead of entering translation first.

Output results and folder rules

By default, the output directory is automatically set to the dubbed subfolder inside the selected media folder. After dubbing and composition finish, you usually get:

  • A final video file, such as video.zh.mp4
  • An aligned subtitle file, such as video.zh.srt
  • Intermediate audio and state data used for retries, resume points, and line-by-line review

Both the preview page and the review page let you open the output folder directly to inspect the final result.

What should you read next?

If you are about to run your first task, start with Quick Start. If you run into workflow or quality issues, continue to the FAQ.