What problem does it solve?
This Skill automates accurate transcription, timing, and subtitle formatting for video and audio files so creators and teams no longer need to manually transcribe, time, or hardcode captions for platforms that strip soft subtitles.
Core Features & Use Cases
- Hebrew and English transcription: Auto-detects language and uses ivrit.ai for Hebrew and Whisper large-v3 for English with fallback handling.
- Translation and burn-in: Translate Hebrew to English and either produce SRT files or burn hardcoded subtitles into MP4 for sharing on platforms like WhatsApp.
- Readable, movie-style subtitles: Natural line breaks, 42 characters per line, 1–7s durations, and merging logic to create viewer-friendly captions.
- Use Case: Convert an interview or social clip into a translated, hardcoded MP4 for distribution and create an SRT for accessibility and archive.
Quick Start
Run the script to transcribe video.mp4, produce an SRT, translate to English, and burn subtitles into a new MP4 using ./scripts/generate_srt.py video.mp4 --srt --burn --translate en.