Skip to content

DINKI_Image.md

Latest commit

 

History

History
235 lines (153 loc) · 11.2 KB

DINKI_Image.md

File metadata and controls

235 lines (153 loc) · 11.2 KB

Home

🖼️ DINKI Overlay

Preview

A powerful and versatile ComfyUI node designed to add watermarks, copyright text, subtitles, and logo overlays to your generated images with professional precision.

✨ Key Features

  • Dual Layering System: Add Text and Image overlays simultaneously or independently using simple toggle switches.
  • Advanced Text Styling:
    • Custom Fonts: Automatically detects .ttf and .otf files in the fonts folder for easy dropdown selection.
    • Stroke (Outline): Add colored outlines to your text for better visibility on complex backgrounds.
    • Drop Shadow: Create depth with adjustable shadow position (offset), blur (spread), and opacity.
    • Multiline Support: Perfect for subtitles or long copyright notices with automatic line spacing handling.
  • Precise Positioning: Choose from 7 preset positions (e.g., Top-Left, Bottom-Center, Center) and fine-tune with percentage-based margins.
  • Adaptive Sizing: Scale text and logos relative to the source image size (%) for consistent results across different resolutions (SDXL, Flux, etc.).
  • Transparency Control: Full support for Alpha/Masks (transparent PNGs) and adjustable opacity (0-100%) for both text and images.

📂 How to Add Custom Fonts

  1. Go to the node's directory: ~/ComfyUI/custom_nodes/ComfyUI-DINKIssTyle/fonts/
  2. Paste your .ttf or .otf font files into this folder.
  3. Restart ComfyUI. Your fonts will automatically appear in the font_name dropdown list.

💡 Usage Tip for Transparent PNGs (Logos)

To properly overlay a logo with a transparent background:

  1. Connect the IMAGE output of your Load Image node to overlay_image.
  2. Connect the MASK output to overlay_mask.
  3. (Optional) Use the overlay_opacity slider to blend the logo with the background.

🎛️ Input Parameters

Parameter Description
font_name Select a font from the fonts folder.
text_content Enter your text here. Supports multiple lines (enter key).
text_opacity Adjust text transparency (0-100).
enable_stroke Toggle text outline. Set color and width.
enable_shadow Toggle drop shadow. Adjust offset (X/Y), spread (blur), and opacity.
overlay_mask (Optional) Connect a mask here to support transparent PNG logos.

📸 DINKI Photo Specifications

Preview

A smart utility node designed to calculate the optimal resolution for AI generation by selecting target megapixels and real-world standard aspect ratios.

Eliminate the guesswork of manual pixel entry. This node ensures your images are generated at the perfect size for models like SDXL, Flux, and Z-Image Turbo.

✨ Key Features

  • Real-World Standards: Supports a wide range of formats, from standard Photography ratios (3:4, 4:6) to professional Cinema/Film specifications (Academy, IMAX, Super 35).
  • AI Optimization: Automatically adjusts Width and Height values to multiples of 8, preventing encoding errors and ensuring compatibility with latent diffusion models.
  • Megapixel Targeting: Select from 1MP to 4MP based on your model's capacity (Base: 1MP = 1024x1024). It maintains consistent quality by preserving the total pixel area across different aspect ratios.
  • Instant Orientation: Easily toggle between Portrait and Landscape modes without recalculating.

💡 Workflow Tip

I found this node to work especially well with Z-Image Turbo workflows, ensuring fast generation at the most efficient resolutions.

🎛️ Supported Formats

Category Aspect Ratios
Photo 3:4, 3.5:5, 4:6, 5:7, 6:8, 8:10, 10:13, 10:15, 11:14
Cinema 35mm Academy (1.37:1), 35mm Flat (1.85:1), 35mm Scope (2.39:1)
Premium 70mm Todd-AO (2.20:1), IMAX 70mm (1.43:1)
Super Super 35 (1.85:1 / 2.39:1), Super 16 (1.66:1 / 1.78:1)

📤 Outputs

  • width (INT): Calculated width (multiple of 8).
  • height (INT): Calculated height (multiple of 8).
  • info_string (STRING): Summary of current settings (e.g., 896x1152 (Photo 3.5:5, 1MP)).

📚 DINKI Batch Images

A smart utility node designed to combine multiple individual images into a single image batch.

Unlike standard batch nodes that error out when image dimensions differ, this node automatically resizes all incoming images to match the resolution of the first image, ensuring a seamless batching process.

✨ Key Features

  • Mass Input: Connect up to 10 different images at once.
  • Auto-Resizing: Automatically scales all images to match the dimensions (Width/Height) of the first input image. No more "Shape Mismatch" errors!
  • Mode Switching: Easily toggle between creating a batch or just passing through the first image for testing.

💡 Workflow Tip

This node works perfectly with DINKI LM Studio Assistant. Use it to batch multiple reference images together and send them to a Vision LLM for bulk analysis or captioning in a single pass.

🎛️ Parameters

Parameter Description
batch_image True (multiple): Resizes and merges all connected images into one batch.
False (single): Ignores the rest and outputs only the first image found (Pass-through mode).
image1 ~ 10 Connect your images here. Inputs can be left empty; the node automatically detects active connections.

▦ DINKI Grid

Preview

An essential ComfyUI node for compiling up to 10 images into a customizable grid layout. Perfect for creating comparison sheets, storyboards, or organized image galleries.

✨ Key Features

  • Flexible Matrix Layout: Define your own grid structure by setting Columns and Rows (e.g., 2x3, 4x4). Images fill the grid from Left-to-Right, Top-to-Bottom.
  • Smart Resolution Handling:
    • Base Resolution: The grid cell size is automatically determined by the resolution of Image 1.
    • Adaptive Resizing: Subsequent images are automatically resized to fit the cell using methods like Fit, Crop, or Stretch.
  • Upscale Comparison Mode:
    • No Resize (Top-Left): A specialized mode where images are placed at their original scale without resizing. Ideal for comparing Upcaled vs. Original images side-by-side to visualize detail enhancement.
  • Custom Styling:
    • Frames: Add spacing between images with adjustable Frame Thickness (supports 0 for seamless grids).
    • Background: Customize the background color (Hex code) for frames and empty cells.
  • Output Optimization:
    • Size Limiter: Toggle limit_output to prevent generating massive files. Automatically downscales the final grid to fit within max_width / max_height while maintaining aspect ratio.

💡 Layout Logic Example

If you set the grid to 2 Columns × 3 Rows (Total 6 cells) but connect only 5 images:

  1. Images 1-2 fill Row 1.
  2. Images 3-4 fill Row 2.
  3. Image 5 fills the first slot of Row 3.
  4. The last empty slot will be filled with your specified Background Color.

🎛️ Input Parameters

Parameter Description
image_1 ~ 10 Connect up to 10 images. Unconnected slots are ignored.
cols / rows Set the number of columns and rows for the grid.
frame_thickness Width of the border around each image (in pixels). Set to 0 for no gap.
bg_color_hex Hex color code for the background/frame (e.g., #000000, #FFFFFF).
resize_method Choose how images fit the cell: Fit, Crop, Stretch, or No Resize.
limit_output Enable to restrict the maximum pixel dimensions of the final image.
max_output_w/h The maximum allowed width/height if limit is enabled.

👁️ DINKI Image Preview

A robust preview node that handles empty signals gracefully. If no image is provided (e.g., a skipped step due to a switch), it automatically generates a custom placeholder image containing text instead of crashing or showing an error.

Preview

🎛️ Parameters Guide

Parameter Description
images (Optional) Connect your image here. If disconnected/None, the placeholder is shown.
placeholder_text Text to display on the placeholder (e.g., "Bypassed").
width / height Dimensions of the placeholder image.
bg_gray / fg_gray Background and Text brightness (0-255 grayscale).
font_path Path to a custom .ttf file. If empty, it attempts to find a system font.

📦 DINKI Base64 Image Embedding Suite

Preview
Download DINKI_Base64_to_Image.json

A set of nodes designed to make your ComfyUI workflows fully self-contained and portable. By converting images into Base64 strings, you can embed essential reference images, masks, or logos directly inside the workflow .json file.

You can embed explanatory images or sample results directly within the Workflow.

🖼️ DINKI Image To Base64

Prepares your image for embedding by converting it into a text-based format. Use this to generate the data needed for the Base64 String Input node.

🎛️ Parameters Guide

Parameter Description
image The source image you want to embed (e.g., a specific ControlNet reference or style image).

Workflow Tip: Connect an image, run the queue, and copy the resulting string. You can then paste it into the DINKI Base64 String Input node to permanently store it in your workflow.

💾 DINKI Base64 String Input

The core storage node. It allows you to paste the Base64 code, effectively saving the image data inside the node itself. When you save and share your workflow .json, the image travels with it.

🎛️ Parameters Guide

Parameter Description
base64_string Paste your Base64 code here. This text field acts as the permanent container for your embedded image.

Key Benefit: Eliminates external file dependencies. Users downloading your workflow will have the correct image loaded instantly, without needing to download separate assets.

👁️ DINKI Base64 Image Viewer

Unpacks and restores the embedded image data for use in generation. It visualizes the stored Base64 string and converts it back into a standard IMAGE format.

🎛️ Parameters Guide

Parameter Description
base64_string Connects to the DINKI Base64 String Input node to retrieve the stored image data.

Smart Decoding: Automatically handles standard Base64 headers.

Output: Returns a standard IMAGE tensor, allowing the embedded image to be used immediately in KSampler, ControlNet, or Image-to-Image processes.