About the incompatibility between the screen size and the characters

I am a beginner. I tried to generate a 9:16 picture in the gpt4 interface, but it always looks like the following picture. I want to know why? I set this up in “Customize ChatGPT”: Make sure to use accurate prompts to describe the people I’m mentioning, and always produce two images. The content must be displayed vertically and the image ratio is 9:16. But still The inability to change characters is a lateral problem that really bothers me. please help me.

1 Like

Hi @tyokenyi
I’ve just seen your question, long time past.

You may use these prompts:

For first and second images:

Create a 9:16 vertical image in an anime or cartoon style of a young boy with curly dark hair and bright blue eyes, joyfully running through a vibrant, lush garden filled with colorful flowers. The boy is wearing a blue shirt and jeans. A colorful parrot with red, blue, and green feathers is flying above the boy. The scene is brightly lit with sunlight filtering through the trees, casting a warm glow and enhancing the vivid colors. Ensure the composition is arranged vertically, capturing the flow from the ground to the sky. --ar 9:16, --chaos 10, --style anime


For third and fourth images:

Create a 9:16 vertical image of a young boy with curly dark hair and bright blue eyes, joyfully running through a vibrant, lush garden filled with colorful flowers. The boy is wearing a blue shirt and jeans. A colorful parrot with red, blue, and green feathers is flying above the boy. The scene is brightly lit with sunlight filtering through the trees, casting a warm glow and enhancing the vivid colors. Ensure the composition is arranged vertically, capturing the flow from the ground to the sky. --ar 9:16, --chaos 10


1 Like