Silent video with title and subtitle | Generated by AI
Committed and pushed at abe19a5. Here’s what changed:
ww gen-video <file> now produces:
- 15 seconds total — 5 slides × 3 seconds each
- No audio — silent video
- Image centered in the frame (middle 60% of the vertical canvas)
- Title at the top (bold, white on semi-transparent black bar)
- Subtitle at the bottom (word-wrapped, same style)
- 1080×1920 vertical, optimized for Douyin / WeChat Video Account
Pipeline changes:
- LLM now generates
title+subtitle+image_promptper scene (no narration script) - Removed all TTS code (
saycommand, audio conversion, duration calculation) - Simplified video assembly — always fixed 3s per slide, no audio track
Generated by AI. Curating and sharing still takes effort.
If you find it useful, feel free to
donate.
WeChat: @lzwjava
·
X: @lzwjava ·
Say hi 👋