Glossary of Terms

This glossary provides definitions for common terms and concepts you'll encounter while using Fater and reading this documentation.


A

  • Activity Type (Editor Mode): Refers to the different specialized workspaces within the Fater Editor, such as the Image Editor, Asset Editor, or Video Editor. Each mode tailors the UI and tools for specific tasks. (See Activity Types (Editor Modes))

  • Agent: Fater's integrated AI assistant, accessible via the Settings Tab. Agents can help with prompts, explain features, and are configurable with different system prompts, LLM models, and temperatures. (See The Agent & How to Use It)

  • AI Model Directory: A reference section in this documentation listing available AI models, their user-facing names, categories, typical uses, and key parameters. (Link to AI Model Directory)

  • Aspect Ratio: The proportional relationship between an image's width and its height (e.g., 16:9, 1:1 for square).

  • Asset (3D): A three-dimensional model or object, which can be generated using AI in Fater's Asset Editor.

  • Auto-Save: Fater's system for automatically saving your project progress in the background after short periods of inactivity. (See Understanding Auto-Save)

B

  • Batch (Generation Batch): A set of one or more outputs (images, assets, videos) generated from a single click of the "Generate" button with a specific set of parameters and prompt.

  • Bounding Box (Generation Area): The dashed rectangular outline on the Inpaint Canvas that defines the active region for AI processing and masking. (See The Generation Area (Bounding Box))

C

  • Canvas (Inpaint Canvas): The main interactive workspace in the Image Editor where layers are displayed, masks are drawn, and direct edits are made.

  • Canvas (Asset Canvas / Video Canvas): The main viewing area in the Asset Editor (for 3D assets) or Video Editor (for video playback).

  • Checkpoint (Model Checkpoint): A specific version or trained state of an AI model. Some Fater models (like certain upscalers) allow you to choose a checkpoint, which can influence the style or quality of the output.

  • Control Layer (Edges/Depth): An image layer (either an Edges map or a Depth map) designated to provide structural or spatial guidance to "Controlled" AI models. (See Control Layers (Edges/Depth))

  • Credits: The virtual currency used to perform AI generation tasks in Fater. Acquired through plans. (See Billing & Subscription)

  • Cutter (Edge Drawing Tool): A direct layer editing tool in the Image Editor used for drawing straight white lines, often for creating Edges maps.

D

  • Dashboard: The central hub you see after logging in, used for managing projects, organization settings, and accessing your account.

  • Depth Map: A grayscale image where pixel intensity represents distance, used as a Control Layer to guide AI with 3D spatial information.

  • Denoising Strength: A concept in some inpainting/image-to-image models that controls how much the AI alters existing pixels under a mask versus generating entirely new content. A strength of 1.0 usually means full replacement. (See "Destroy Fill Area Behind Mask" parameter).

  • Destroy Fill Area Behind Mask: A parameter for some "Edit" category AI models. If "Yes," the AI completely ignores original pixels under a mask during inpainting. If "No," original pixels can influence the result.

E

  • Edges Map: A black and white image where white lines represent outlines and contours, used as a Control Layer to guide AI with structural information.

  • Editor Modes: See Activity Type.

F

  • Floating Prompt Area: The persistent UI element at the bottom-center of the Editor, containing the text prompt input, Generate button, and Generation Task List.

G

  • Generation Area: See Bounding Box.

  • Generation Task List: The row of thumbnails in the Floating Prompt Area displaying recent AI generation results. (See Working with the Generation Task List)

  • Guidance Scale (CFG Scale): A common AI parameter that controls how strictly the generated output adheres to the text prompt.

I

  • Image Editor: The Fater Editor Mode for 2D image creation, AI generation, and layer-based editing.

  • Inpainting: An AI technique where a masked area of an image is filled in or modified based on a prompt and the surrounding image context.

  • Inpaint Toolbar: The floating toolbar in the Image Editor providing tools for masking, direct layer editing, and canvas actions.

L

  • Layer: An individual image or element in the Image Editor that can be manipulated independently. Layers are stacked to form a composite image. (See The Layer Stack)

  • Layers Tab: The tab in the Right Sidebar of the Editor used for managing the layer stack, layer properties, and Layer Workspaces.

  • Layer Workspace: A separate, independent layer stack within the Layers Tab, used for organizing complex projects or variations.

  • LLM (Large Language Model): The type of AI model that powers Fater's Agent.

M

  • Mask: A selection that defines a specific area of an image for AI operations or targeted edits. Usually visualized as a semi-transparent overlay. (See Masking for AI)

  • Mask Alpha Radius: A parameter for some AI models (inpainting, upscaling) that controls the transparency feathering at the edge of the generated/affected area for smoother blending.

  • Mask Extra Radius: A parameter for some inpainting models where the AI generates content slightly beyond the defined mask to improve blending.

  • Model Selector: The UI component in the Left Sidebar of the Editor used to browse and select AI models based on categories and types.

N

  • Negative Prompt: A text description telling the AI what elements, styles, or qualities to avoid generating.

O

  • Opacity: The transparency level of a layer. 100% opacity is fully opaque; lower values are semi-transparent.

  • Outpainting: An AI technique, often related to inpainting, used to extend the canvas and generate new content beyond the original image boundaries, guided by a prompt and existing image context.

P

  • Parameter: A configurable setting for an AI model that influences its behavior and output (e.g., Steps, Guidance, Seed).

  • Pencil Tool: A direct layer editing tool in the Image Editor for freehand drawing on a selected layer.

  • Prompt (Text Prompt): The primary textual description given to an AI model to guide its generation.

R

  • Reference Image (External): A separate image uploaded by the user to provide visual guidance (style, content) to certain AI models, in addition to a text prompt.

  • Restore Layer Config: A feature on AI-generated layers that reapplies the model and parameters originally used to create that layer.

S

  • Seed: A number that initializes the random process in AI generation. Using the same seed (with other inputs constant) produces the same result. (See Seeds)

  • Segment Mask (Prompt to Mask): An AI-powered feature where the user provides a text prompt, and the AI generates a new layer representing a mask for the described object/region.

  • Steps (Diffusion Steps): A common AI parameter controlling the number of refinement iterations. More steps often mean more detail but longer generation time.

U

  • Upscaling: The process of increasing an image's resolution and enhancing detail using AI.

Last updated