A Python script to generate and burn subtitles into videos using OpenAI Whisper, pysubs2, and FFmpeg.
It supports local transcription, .srt and .ass subtitle styling, Unicode languages (like Hindi), and fully customizable font, size, color, and position.

• Automatic Transcription: Uses OpenAI Whisper to transcribe audio from any video.
• Subtitle Generation: Converts Whisper output into .srt and .ass subtitle files.
• Custom Styling: Adjust font, font size, color, position, and vertical margin.
• Unicode & Multilingual Support: Supports non-Latin scripts (e.g., Hindi/Devanagari).
• Burn Subtitles into Video: Uses FFmpeg to embed subtitles directly into the output video.
• Clean Visuals: No black borders by default for sleek, modern subtitle appearance.
• Optional Pre-existing Transcript: Can skip transcription if you already have a .srt file.

The script works in four main stages:

• If a transcript (.srt) is not provided, the script uses Whisper to transcribe the audio from the video.
• Whisper returns segments, each containing:
  • start time
  • end time
  • text spoken in that interval

• Converts Whisper segments into an .srt file.
• Each subtitle block in .srt format contains: 1 00:00:05,000 --> 00:00:10,000 Hello, world!

• Converts .srt → .ass using pysubs2.
• Custom styling options include:
• font and font_hindi for multilingual support
• fontsize for text size
• color for subtitle text color
• position (bottom, top, center) and margin
• outline and shadow are removed for clean visuals (no black border)

• Uses FFmpeg to embed .ass subtitles into the video:

ffmpeg -i input.mp4 -vf "ass=styled.ass" -c:v libx264 -crf 18 -preset medium -c:a copy output.mp4

• The final video has the subtitles permanently burned in, compatible with any player.

• Python 3.8+
• FFmpeg
• Python packages:

pip install openai-whisper pysubs2 torch

python subtitle_burner.py \
  --video input.mp4 \
  --out output_subbed.mp4 \
  --font "/path/to/NotoSans-Regular.ttf" \
  --fontsize 36 \
  --color "#FFFFFF" \
  --position bottom

python subtitle_burner.py \
--video input.mp4 \
--out output_subbed.mp4 \
--model small \
--language hi \
--font "/path/to/NotoSans-Regular.ttf" \
--font_hindi "/path/to/NotoSansDevanagari-Regular.ttf" \
--fontsize 36 \
--color "#FFFFFF" \
--position bottom

python subtitle_burner.py \
  --video input.mp4 \
  --transcript subtitles.srt \
  --out output_subbed.mp4 \
  --font "/path/to/NotoSans-Regular.ttf"

This project aims to simplify the process of generating high-quality, styled subtitles for videos.
By combining OpenAI Whisper for transcription, pysubs2 for styling, and FFmpeg for burning, you can create professional-looking subtitles in multiple languages with minimal effort.

Whether you are creating content for YouTube, online courses, or personal projects, subtitle_burner.py allows you to automate the workflow while maintaining full control over font, color, size, and position.

Contributions, suggestions, and improvements are welcome! Feel free to open an issue or submit a pull request.