Hi everyone,

I’ve been looking for a while for an image-editing method for Anima that doesn’t require a lot of processing power.

Since I use Civitai and they’ve recently released their own nodes for Comfy, I naturally tried those.

Well, it didn’t work out, but in one of my bounties, someone gave me a good solution that I’m now publishing here. I’m trying to adapt and expand it a bit for my own use, but let’s take it one step at a time.

I asked @uncinsane475 if he’d be willing to publish it, and he said I could go ahead.

It's similar to inpainting.

Sources:

Bounty: https://civarchive.com/bounties/12482/i-want-to-be-able-to-edit-images-well-using-civitai-nodes-and-anima

Provided by: uncinsane475

Thank you for the nice workflow and that I can post is.

Installation:

1. Install ComfyUI

2. I recommend using the Extension Manager as suggested on the website

3. Import the workflow, e.g., via drag-and-drop

4. Install all necessary/missing nodes using the Manager

5. Restart and you’re good to go.

Usage:

1. Connect to Civitai API using OAuth or APIKey

2. Load/insert an image

3. Create a mask

4. Preview the mask to see if it’s good (optional)

5. Provide a prompt describing what’s visible in the mask—not “Change it this way or that,” but “You can see …,” simply describing what should be there

6. Start the workflow

7. View the result

8. If you’re not satisfied or it’s not finished, try again with different settings, or take the result, insert it as the starting image, and continue editing.

Translated with DeepL.com (free version)