Notes on AI lip sync and talking-photo videos with Fameo
| Filename | Latest commit message | Latest commit date |
|---|---|---|
| README.md | ||
Preparing stills for lip sync
Most of the quality in a talking video is already decided when you choose the photo.
Prefer
- A single, front-facing face
- Even light on the mouth
- A crop where the head is large
- JPG, PNG, or WebP stills, or a short MP4 / MOV if you upload video
Avoid
- Profiles and extreme angles
- Hands, microphones, or objects over the lips
- Heavy shadow across the mouth
- Crowded group shots when you only want one speaker
These notes apply on Fameo for portraits, pets, cartoons, and artwork. If the still has two people side by side and you want a conversation, use multi-talk instead of forcing the general tool. If the subject is a pet, start on the pet page with a photo you have the right to use.