Add README.md
This commit is contained in:
@@ -0,0 +1,22 @@
|
||||
# Image-to-video prompt cheat sheet
|
||||
|
||||
Short reference for writing prompts that turn a still frame into a moving shot.
|
||||
|
||||
## Formula
|
||||
Subject + one action + one camera move + light.
|
||||
|
||||
## Examples
|
||||
- A ceramic mug on a wooden desk, steam rising, slow push-in, warm morning light.
|
||||
- A paper boat on a puddle, ripples spreading, static camera, overcast sky.
|
||||
|
||||
## Habits that help
|
||||
- Keep each prompt to one sentence per element.
|
||||
- Name the camera move explicitly; leave it out and the shot stays almost still.
|
||||
- Change a single word between two test runs so the difference is easy to see.
|
||||
|
||||
## Habits that hurt
|
||||
- Stacking several actions in the same shot.
|
||||
- Mixing styles such as photoreal and cartoon in one line.
|
||||
|
||||
## Try it online
|
||||
Prompts like these can be tested in a [text to video AI](https://wan3.io/) tool that runs in the browser.
|
||||
Reference in New Issue
Block a user