Midjourney’s first generation video model is doing some incredible things! It's easy to use and surprisingly does a great job with dynamic motion, similar quality as kling 1.6. It also has a built-in video extension feature.
Here's what that does:
- Allows you to extend a video by 5s
- You can choose high or low motion
- You can choose auto or manually write a prompt
- You can extend 4 times for a full 20s clip
The process is super easy:
01. Create initial image + auto animate with high motion
02. Use “Auto/Manual” extend to add 5 seconds to your video
03. Repeat
My Original Image Prompt:
- Cat: Wide shot hyperrealistic style of an athletic tabby cat wearing a red tracksuit, leaping over a hurdle on a vibrant Olympic stadium track, with a cheering crowd in the background under bright daylight
- Koala: front shot of a koala with sunglasses riding on a red motorcycle on an empty highway with a helmet on, bright blue sunny sky.
Exact Animation Steps:
Initial Video: Auto prompt (motion high)
Extension 1: Auto prompt (motion high)
Extension 2: Auto prompt (motion high)
Extension 3: Manual prompt (motion high) Extension 4: Auto prompt (motion high)
Manual Prompts I used:
- “cat does a dramatic landing and rolls forward, sliding to a stop. The cat does a cool pose at the camera, laying on the ground making a finger gun with its paw”
- “koala does a wheelie on the motorcycle, camera cinematic tracking shot”
I love how seamless the transitions are with these scene extensions! I feel like the clips keep momentum and style consistency really well and using auto prompt (motion high) makes the entire process extremely easy to do. Give this a try and share your results in the comments. Even as a complete beginner it will take you less than 15 minutes to create an image, animate it, and extend it at least once.