ComfyUI 35 Grouped Nodes - Free Workflow - Stable Diffusion

Rudy's Hobby Channel
21 Apr 202406:50

TLDRThe video provides a tutorial on using the ComfyUI 35 Grouped Nodes feature in Stable Diffusion for a cleaner workflow. It demonstrates how to combine multiple nodes into a single grouped node, customize visibility of inputs and widgets, and rearrange them for a more organized interface. The presenter also shares their updated default workflow, explaining the step-by-step process from image generation to post-processing, and how to efficiently refine and upscale images using this feature.

Takeaways

  • 😀 ComfyUI has an updated function to group nodes, making the workflow more organized and user-friendly.
  • 🛠️ Users can now customize which inputs, outputs, and widgets appear within a grouped node, enhancing workflow efficiency.
  • 📋 The script demonstrates how to convert multiple nodes into a single 'grouped node' by right-clicking and selecting 'convert to group'.
  • 🔍 After grouping, users can manage the visibility of widgets within the group node through a management window.
  • 🧩 The video script showcases a neat and tidy workflow by rearranging and managing grouped nodes for a cleaner interface.
  • 🖼️ The default workflow is designed to work from left to right, making it easy to follow and less cluttered.
  • 🔄 The workflow includes steps like a loader group, sampler, latent upscaler, phase detailer, image upscaling, post-processing, and saving as JPEG.
  • ⏱️ The script mentions the speed of image generation with the turbo checkpoint, which is approximately six to seven seconds per image.
  • 🎨 The workflow allows for quick testing of new prompts and selection of preferred images for further processing.
  • 🔧 Users can enable or disable certain nodes like the latent upscaler, face detailer, and post-processing according to their needs.
  • 🔗 The video provides a download link for additional workflows and mentions the inclusion of an adapter or control net in the workflow setup.

Q & A

  • What is the main feature discussed in the script about ComfyUI 35?

    -The main feature discussed is the ability to group multiple nodes into a single grouped node in ComfyUI 35, which has been improved to allow users to manage which inputs, outputs, and widgets they want to see within the new node.

  • How does the new function of grouping nodes in ComfyUI 35 improve the workflow?

    -The new function allows users to select which widgets to display or hide within the grouped node, making the workflow more organized and less cluttered, thus enhancing the user experience.

  • What is the default workflow in ComfyUI 35?

    -The default workflow in ComfyUI 35 is designed to work from left to right, making it easy to follow and understand the process of image generation and processing.

  • How can users rearrange the nodes within a grouped node in ComfyUI 35?

    -Users can rearrange the nodes within a grouped node by right-clicking the node, selecting 'manage the group', and then using the six dots to drag and place the nodes in the desired order.

  • What is the purpose of the 'manage the group' option in ComfyUI 35?

    -The 'manage the group' option allows users to customize the visibility of inputs, widgets, and other elements within a grouped node, providing a more tailored and efficient workflow.

  • How does the script describe the process of converting nodes into a grouped node?

    -The process involves selecting multiple nodes using the control key and the left mouse button, right-clicking, and choosing 'convert to group'. After naming the group, the nodes are combined into a single grouped node.

  • What does the script suggest as the first step in the default workflow?

    -The first step in the default workflow is using a loader group that includes prompts, checkpoint, image size, and the styler.

  • What is the typical time it takes to generate an image using the default workflow in ComfyUI 35?

    -On the user's system with a turbo checkpoint, it takes close to six or seven seconds to generate one image.

  • How can users quickly try out multiple images using the default workflow?

    -Users can activate only the loader and the first sampler to quickly generate a set number of images, such as six, twelve, or twenty, and then select the ones they like for further processing.

  • What is the final step in the default workflow described in the script?

    -The final step in the default workflow is to save the image as a JPEG, which includes the post-processing steps like color adjustment, sharpening, vignette, and film grain.

Outlines

00:00

🛠️ Customizing Workflow with Group Nodes

The script discusses an enhancement to the 'COMI' software's built-in function for grouping nodes. It allows users to create a new node by combining several different ones, and has been updated to be more customizable. The user can now manage which inputs, outputs, and widgets are visible within the grouped node. The script provides a walkthrough of how to use this feature in the default COMI workflow, including rearranging nodes for a cleaner interface and managing group nodes to show or hide specific widgets as needed. The process of converting individual nodes into a group node called 'group loader' is demonstrated, and the benefits of this feature in streamlining the workflow are highlighted.

05:00

🖼️ Streamlining Image Generation with Group Nodes

This paragraph continues the discussion on using group nodes to streamline the image generation process in COMI. It details the workflow from the initial image generation using a loader and sampler to the selection of preferred images for further processing. The script explains how to activate the latent upscaler and face detailer, if necessary, and proceed with image upscaling. It also mentions the option to include or exclude post-processing steps such as color adjustment, sharpening, and adding film grain. The user's typical workflow is described, starting with a basic setup and expanding it based on the needs of the image being processed. The script concludes with a mention of optional components like adapters or control nets and provides a download link for additional resources.

Mindmap

Keywords

💡ComfyUI

ComfyUI refers to a user interface design that is easy and pleasant to use, often characterized by a clean layout and intuitive controls. In the context of the video, ComfyUI is a feature within the Stable Diffusion workflow that allows users to create a more organized and user-friendly environment by grouping nodes together.

💡Grouped Nodes

Grouped Nodes is a functionality that enables the combination of multiple nodes into a single, cohesive unit. This concept is central to the video, as it demonstrates how users can simplify their workflow by grouping related nodes, making the interface neater and the process more efficient.

💡Workflow

A Workflow in this context is a sequence of steps or processes involved in completing a task or project within a software application. The video discusses the default workflow in ComfyUI and how it can be enhanced by using Grouped Nodes to streamline the process of image generation.

💡Convert to Group

Convert to Group is an action within the ComfyUI that allows users to turn selected nodes into a single grouped node. The script mentions this action as a way to create a more organized workspace by consolidating multiple nodes into one, which simplifies the user's interaction with the software.

💡Widgets

Widgets in the video script refer to the interactive elements within the ComfyUI, such as sliders, buttons, and text fields, that users can manipulate to control various parameters of the image generation process. The video explains how to manage the visibility of these widgets within a grouped node.

💡Checkpoint

A Checkpoint in the context of the video is a saved state of the image generation process that can be loaded and used as a starting point for further modifications. The script describes how the checkpoint is part of the grouped nodes and is essential for continuing the workflow from a specific point.

💡Sampler

A Sampler in the video is a node responsible for the initial generation of images based on the input parameters. It is mentioned as a critical component in the workflow, especially in the first step where multiple images are created for selection and further processing.

💡Latent Upscale

Latent Upscale is a process mentioned in the script where an existing image is enhanced or scaled up in quality. This term is used to describe a step in the workflow where an initially generated image is refined to improve its resolution or detail.

💡Post Processing

Post Processing refers to the final adjustments made to an image after the main generation process. In the video, it includes steps like color correction, sharpening, and adding effects such as vignette and film grain. This is an important part of the workflow to achieve the desired final look of the image.

💡JPEG

JPEG is a common image file format used for storing and sharing images, known for its balance between quality and file size. The script mentions saving the final output as a JPEG, which is a standard practice for sharing images online or in print.

💡G-Drive

G-Drive in the script likely refers to a storage device or cloud storage service where additional resources or files related to the workflow are stored. It is mentioned as a place where users can find additional grouped nodes or other workflow components.

Highlights

ComfyUI 35 introduces a new function to combine different nodes into a grouped node.

The function has been updated to allow more customization of inputs, outputs, and widgets.

Users can now select which widgets to make visible within the grouped node.

A demonstration of converting multiple nodes into a single grouped node named 'Group Loader'.

Managing the grouped node to customize the visibility of widgets.

The importance of the checkpoint name widget being visible in the workflow.

Making certain widgets invisible to simplify the workflow interface.

Rearranging the order of widgets within the grouped node for better workflow organization.

The updated workflow makes it easier to focus on essential parameters and steps.

Workflow starts with a loader group including prompts, checkpoint, image size, and styler.

The pipe sampler is essential for the initial creation and selection of images.

Further processing can involve a latent upscale and a phase detailer if needed.

Image upscaling and post-processing steps are part of the workflow.

Workflow concludes with saving the image as a JPEG.

The default workflow is designed to be easy to follow from left to right.

Workflows can be customized to include or exclude certain widgets for specific needs.

An example of how to work with a new prompt, starting with the loader and the first sampler.

Using the latent upscaler and face detailer based on the needs of the selected image.

The process of enabling the JPEG output and post-processing for image upscaling.

The final image produced using the updated default workflow.