Join My Newsletter
[sibwp_form id=2]
Want to create consistent AI characters and master AI image storytelling with full camera control? This Krea AI & WAN 2.1 tutorial will show you how to take full control of your AI-generated images and videos, from character consistency to advanced camera angles.
In this video, you’ll learn how to:
✔️ Build a consistent AI character sheet using Krea Chat
✔️ Train a LoRA model in Krea AI for precision control
✔️ Position a 3D model in any pose and angle
✔️ Use Krea Realtime to generate images with dynamic framing
✔️ Turn AI images into stunning videos with the WAN 2.1 model
We start by using Krea Chat, a powerful tool powered by DeepSeek, to generate a consistent AI character sheet. Uploading a 9-face image grid significantly improves results. You can download my pre-made grids from my blog post (linked below).
To generate your own, simply:
Watch the tutorial (at start of the Blog post) for detailed instructions at: 00:43
Krea AI: https://www.krea.ai/
Prompt for Krea AI Chat:
This photograph is a composite of nine photographs in 1:1 aspect ratio arranged in a 3×3 grid, featuring the same European woman in various angles and expressions against a plain, muted gray background.
The woman appears to be in her mid-20s, with fair skin, a fit physique, and a square face. She has long blonde hair and blue eyes. The images should include a mix of frontal, profile, slight side angle and three-quarter views, capturing a range of expressions such as neutral, happiness, sadness, and anger to add diversity.
The lighting should be soft, casting gentle shadows to highlight the contours of her face and body. The overall mood of the composite is intimate and reflective, emphasizing her natural beauty and the versatility of her expressions. The photographs should be crisp and well-defined, focusing on a realistic aesthetic.
And use the uploaded image as a reference for the grid and face positions

Once you have the perfect character sheet:
Watch the tutorial (at start of the Blog post) for detailed instructions at: 01:43
Split Images: https://splitter.imageonline.co/
Now, we train our LoRA model:
Watch the tutorial (at start of the Blog post) for detailed instructions at: 03:25

To control the camera angle and character placement, we use PoseMyArt, a free tool for creating 3D references:
Watch the tutorial (at start of the Blog post) for detailed instructions at: 04:35
Pose my art for creating 3D model: https://posemy.art/
Female standing
Male standing
Models for Low Angle – Full Body
Models for Low Angle – Looking down
Models for Low Angle – Medium Shot
Models for Low Angle – Close-up Shot
Female High Angle – Full Body Shot
Male High Angle – Full Body Shot
High Angle – Close-up Shot
With our trained LoRA model, we now move to Krea Realtime to control shot types:
Watch the tutorial (at start of the Blog post) for detailed instructions at: 05:44
To animate your AI-generated images:
For smoother motion, apply frame interpolation using Topaz Video AI or Flowframes (free alternative, linked below).
Watch the tutorial (at start of the Blog post) for detailed instructions at: 10:42
Flowframes for frame interpolation: https://nmkd.itch.io/flowframes
Resources:
✔️ Invite Link to Our Discord Group: https://discord.gg/gdDnxTFUDf
✔️ Krea AI: https://www.krea.ai/
✔️ Split Images: https://splitter.imageonline.co/
✔️ Flowframes for frame interpolation: https://nmkd.itch.io/flowframes
✔️ Topaz Gigapixel 8: https://www.topazlabs.com/gigapixel
✔️ Pose my art for creating 3D model: https://posemy.art/
