gitwtfhub

wtf is videosubtitler?

goldyubrain/videosubtitler — explained in plain English

Analysis updated 2026-05-18

70PythonAudience · generalComplexity · 2/5LicenseSetup · moderate

TL;DR

A Windows tool that transcribes a video, translates it to Traditional Chinese, and burns the subtitles into the file.

Mindmap

mindmap
  root((VideoSubtitler))
    What it does
      Speech to text
      Translate to Traditional Chinese
      Burn subtitles into video
    Tech stack
      Python
      faster-whisper
      ffmpeg
    Use cases
      Subtitle videos
      Fix translations
      Subtitle audio files
    Audience
      Chinese content creators
    Limits
      Windows only
      One file at a time

Code map

Detail Auto

An interactive map of this repo's files and how they connect — its source is parsed live in your browser. Click Visualize to build it.

filefunction / class

Why would anyone build with this?

REASON 1

Add burned in Traditional Chinese subtitles to a video or audio recording.

REASON 2

Fix a mistranslated or mis-heard subtitle line and only re-run the burn step.

REASON 3

Turn a podcast style audio file into a subtitled video automatically.

What's in the stack?

Pythonfaster-whisperOpenCCffmpeg

How it stacks up

goldyubrain/videosubtitlerastrovela/vaneclayboxerreproduce/meccha-chameleon-tool
Stars707070
LanguagePythonPythonPython
Setup difficultymoderatemoderateeasy
Complexity2/54/52/5
Audiencegeneraldevelopergeneral

Figures from each repo's GitHub metadata at analysis time.

How do you spin it up?

Difficulty · moderate Time to first run · 30min

Windows only, and ffmpeg must be downloaded separately since it is not bundled with the project.

MIT: free to use, modify, and share, including commercially, as long as you keep the copyright notice. Note ffmpeg itself has its own separate license.

Wtf does this do

VideoSubtitler is a Windows command line tool that takes a local video or audio file, such as mp4, mp3, wav, or m4a, and produces subtitles in three steps: it transcribes the speech, translates the result into Traditional Chinese, and then burns the translated subtitles directly into a copy of the video. The transcription step uses an open source speech recognition tool called faster whisper and outputs both a timed subtitle file and a plain text file. The translation step converts the text to Traditional Chinese and additionally runs it through a conversion library called OpenCC to make the wording match Taiwan usage rather than mainland Chinese phrasing. The final step uses ffmpeg to hard code the translated subtitles into the video permanently, choosing a Windows system font by default. If you only provide an audio file, the tool automatically creates a plain dark background video first so the subtitles have something to sit on. You can run just one or two of the three steps instead of all of them, which is useful if you want to manually fix a mistake in the recognized text or the translation without redoing the whole process from scratch, since the subtitle files are ordinary text you can edit directly. There are options to control which whisper model size is used, which translation engine to use among a free Google translate option, a paid Claude API option for higher quality, or a plain Chinese conversion only option, plus subtitle styling choices like font, size, color, and position. To set it up you need Windows 10 or 11, Python 3.10 or newer, and ffmpeg, which is not bundled in the project because of its size and must be downloaded separately and placed in a tools folder or added to your system path. The project ships two small sample clips with no copyright concerns so you can try it immediately without your own footage. The tool currently only processes one file at a time with no batch mode, keeps Whisper's original line breaks which can make some subtitle lines long, only supports permanently burned in subtitles rather than a toggleable subtitle track, and has only been tested on Windows. The project is released under the MIT license, while noting that ffmpeg itself carries its own separate license terms.

Yoink these prompts

Prompt 1
Run VideoSubtitler on demo.mp4 using the large-v3 whisper model for the most accurate transcript.
Prompt 2
Show me how to manually fix a translation error in the generated .zh-TW.srt file and re-burn it.
Prompt 3
Explain the difference between the google, claude, and opencc translation engine options.
Prompt 4
Set VideoSubtitler to output yellow bold subtitles positioned at the top of the video.

Frequently asked questions

wtf is videosubtitler?

A Windows tool that transcribes a video, translates it to Traditional Chinese, and burns the subtitles into the file.

What language is videosubtitler written in?

Mainly Python. The stack also includes Python, faster-whisper, OpenCC.

What license does videosubtitler use?

MIT: free to use, modify, and share, including commercially, as long as you keep the copyright notice. Note ffmpeg itself has its own separate license.

How hard is videosubtitler to set up?

Setup difficulty is rated moderate, with roughly 30min to a first successful run.

Who is videosubtitler for?

Mainly general.

View the repo → Decode another repo

This repo across BitVibe Labs

Don't trust strangers blindly. Verify against the repo.