The AI Generation Workflow in Image Editor

Generating or editing images with AI in Fater's Image Editor follows a general workflow. Understanding these steps will help you effectively use the various AI models available.


Core Workflow Steps:

  1. Prepare Your Canvas & Context:

    • Layers: Ensure the correct layers are visible and arranged in the Layers Tab. The AI often considers the composite view of visible layers within the Generation Area as its input or context.

    • Generation Area Bounding Box: Adjust this dashed rectangle to define the precise region where you want the AI to operate. This is crucial for focus and detail.

    • Mask (for Editing/Inpainting): If you're modifying an existing image (e.g., inpainting or object removal), create a precise Mask over the area you want the AI to change.

    • Control Layers (for Controlled Generation): If using a "Controlled" model, prepare and designate your Edges or Depth map layer(s) in the Control Layers section of the Layers Tab.

  2. Select an AI Model:

    • Go to the Model Selector in the Left Sidebar.

    • Choose the Category (Generate, Edit, Upscale), SubCategory (Basic, Controlled, Special), and Control Type (Contour/Edges, Depth) if applicable.

    • Select the specific user-facing model name you want to use.

  3. Configure Model-Specific Parameters:

    • Once a model is selected, its parameters will appear in the Left Sidebar below the Model Selector.

    • Adjust settings like Prompt, Negative Prompt, Steps, Guidance, Seed, and any other controls specific to that model (e.g., Style Type, Mask Alpha Radius, Upscale Scale).

    • (Refer to the AI Models for details on each model's parameters).

  4. Write/Refine Your Text Prompt:

    • In the Floating Prompt Area at the bottom of the editor, enter a clear and descriptive text prompt.

    • If you're editing, describe the changes you want within the masked area. If generating, describe the entire scene or subject.

    • You can also use the Voice Input button here (On Chrome).

  5. Upload Reference Images (if applicable):

    • If the selected model supports External Reference Images, upload them using the designated input area in the Left Sidebar.

  6. Initiate Generation:

    • Click the Generate button (✨ Sparkles icon) in the Floating Prompt Area.

    • The button will indicate "Generating..." while the request is being preparred.

    • If the task was launched correctly, it will appears in the task list as loading, you can continue working and launch different generations during the task processing time.

  7. Review and Use Results:

    • The first generated image is automatically added as a new layer at the top of your layer stack, typically positioned correctly (e.g., within the masked area for inpainting).

    • Thumbnails for all generated outputs appear in the Generation Task List above the prompt input.

    • Click a different thumbnail to swap the current result layer with the new selection.

    • You can then further edit this new layer, adjust its opacity, visibility, or merge it down.


Iterative Process

AI image generation is often an iterative process:

  • Analyze: Look at the generated result. Does it match your vision?

  • Refine:

    • Adjust your Prompt or Negative Prompt.

    • Tweak Model Parameters (e.g., Guidance, Steps, Seed).

    • Modify your Mask or Control Layer.

    • Change the Generation Area Bounding Box.

  • Regenerate: Click "Generate" again.

Using a locked Seed can be very helpful for making small adjustments to prompts or parameters while keeping the overall composition similar. Restoring a layer's configuration using the "Restore Layer Config" feature can also help you revisit and iterate on previous settings.

By following these steps and iterating, you can achieve a wide range of creative outcomes with Fater's AI image generation tools.

Last updated