# Ai Foley Editor

- **Event:** [fal x Sequoia 72-Hour Video Hackathon](https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon)
- **When:** Jul 17 at 12:00 PM – Jul 19 at 8:00 PM (PDT)
- **Where:** Online
- **Team:** [Narayan S S](https://cerebralvalley.ai/u/nrynss)
- **GitHub:** https://github.com/nrynss/AI-foley-generator
- **Demo video:** https://www.youtube.com/watch?v=cdZWlSXaSWk
- **Gallery:** https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery
- **Page:** https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery/55

AI video generation is getting remarkably good.
The hard problem is not “no audio.” It’s the wrong audio — and regenerating the entire video to fix one beat is slow, expensive, and risks the picture you already love.

Cinema solved this about a hundred years ago: picture and sound are separate tracks. You lock the picture. Then you author the soundtrack against it.

AI Foley is a picture-lock post pipeline for silent AI video:
• A vision model reads the clip and builds a beat graph — an editable event timeline
• Each event is routed to a specialized sound model and generated as SFX
• You review, direct in plain language, and re-render audio only — seconds, not a new shoot
• Same frames. New soundtrack.

Logs are here: https://github.com/nrynss/AI-foley-generator/blob/main/ledger.jsonl

---

Markdown version of https://cerebralvalley.ai/e/fal-x-sequoia-virtual-hackathon/hackathon/gallery/55. Site index for agents: https://cerebralvalley.ai/llms.txt · full text: https://cerebralvalley.ai/llms-full.txt