All Collections
Using “seed” images
Using “seed” images

How to use “seed” images to create consistency across frames

Karen Mc Guinness avatar
Written by Karen Mc Guinness
Updated over a week ago

When you use AI to generate images, a seed image is a tool that can help guide the process. Here’s a simple explanation of what it is and how it can be useful 👇

What is a Seed Image?

A seed image is a starting image that you provide to an AI system before it generates a new image. The purpose of the seed image is to give the AI a starting point for its image generation. Only AI-generated images can be used as seeds.

Why Use a Seed Image?

Here are some reasons why a seed image can be helpful:

  • Direction: The seed image helps to guide the AI in creating an image that is closer to what you have in mind. It’s like giving the AI a hint about what you want to see in the final picture.

  • Consistency: If you’re creating a series of images and want them to have a similar look or feel, using the same seed image for each one can help achieve that consistency.

  • Control: Using a seed image gives you more control over the image generation process. You can influence the outcome, making sure the AI creates an image that aligns with your vision.

Seed Images in Boords

After generating an AI image, you’ll have an option to use that image as a seed.

From here, you can generate another version of the seed image by adjusting the prompt to better reflect what you’re looking for. This is useful if you're happy with the style of the image, but want to make some adjustments to the image.

This will result in a slightly different image with a similar style, depending on the new prompt that you’ve given.

Seed Images to create consistency across frames

To create consistency across frames, we suggest first duplicating the frame with the style that you'd like to continue.

In the duplicate image, use the frame image as a seed. Then, edit the prompt to suit what you’d like the image to show.

This will help to create an element of consistency across subsequent frames.

Did this answer your question?