AnimateDiff Legacy Animation v5.0 [ComfyUI]

Jerry Davos AI
15 May 202406:00

TLDRIn this tutorial, the creator guides viewers through the process of crafting an animation using Comfy UI and Anime Dall-E workflows. The video begins with setting up the first workflow, which includes inputs, animation, properties, and control settings. The creator shares the output folder path for rendering frames, selects the model 'Concept Pyromancer Laura' for a fire effect, and adjusts the weight to 0.5. The workflow continues with the control net, open pose reference images, and exporting settings. The creator then moves on to upscaling the video, using a specific model and adjusting the target resolution and FPS. Finally, the video is enhanced with a face fixer workflow to improve facial details. The tutorial concludes with a note on the creator's Patreon, where more in-depth tutorials and resources are available for free to support the community in learning and improving their AI-generated art.


  • 🎨 Use Comfy UI and Anime to create animations with a specific workflow.
  • 📂 Drag and drop the first workflow to start, which includes inputs, animation, properties, and controls.
  • 🔗 Include a link in the description to the workflow for easy access.
  • 📊 Select the output folder path where frames will be rendered and choose the output dimension.
  • 🔥 Choose a model like 'Mune Anime' and 'Concept Pyromancer' for cool fire effects.
  • 📝 Use a prompt for the Anime Diff model and control net settings.
  • 🖼️ Unmute the Directory Group to use open pose reference images from previous renders.
  • 📹 Adjust the FPS for the exporting video to control the speed of the animation.
  • 🔍 Render the queue and wait for the animation to finish.
  • 📈 For upscaling, use a video upscale workflow with specific model settings and an upscale value.
  • 📼 Use the video2video face fixer workflow for enhancing facial details and smoothness.
  • 💖 Support from Patreons helps keep tutorials free and accessible for everyone.

Q & A

  • What is the main topic of the video?

    -The video is about creating an animation using Comfy UI and Anime, including various workflows such as animation, settings, and video export.

  • What are the key components in the animation process described in the video?

    -The key components include inputs, animate D, prps, control net, batch or single op option, case sampler, settings, and video export.

  • What is the purpose of the 'control net' in the animation workflow?

    -The control net is used to manage the open pose reference images, which are essential for the animation process.

  • How does one add fire effects to the animation?

    -The video suggests using a concept pyromancer, Laura model and adjusting its weight to around 0.5 to add cool fire effects to the animation.

  • What is the recommended batch size for the output in the tutorial?

    -The recommended batch size for the output in the tutorial is 72.

  • How can one extract open pose images for the animation?

    -One can extract open pose images using the CN passes extractor workflow.

  • What is the frame rate (FPS) set for exporting the video in the tutorial?

    -The frame rate (FPS) set for exporting the video in the tutorial is 12.

  • What is the purpose of the 'upscaling' workflow in the animation process?

    -The upscaling workflow is used to enhance the resolution of the video, making it more detailed and visually appealing.

  • What is the target resolution set for the upscaled video?

    -The target resolution set for the upscaled video is 1200.

  • How does one ensure the video's speed matches the desired pace?

    -One can adjust the FPS (frames per second) according to the video's requirements for faster or slower speed.

  • What additional step is taken after rendering to improve the video's quality?

    -After rendering, the video undergoes face fixing using the video2video face fixer workflow to enhance the details of the faces in the animation.

  • Where can viewers find more tutorials and support the creator?

    -Viewers can find more tutorials and support the creator on the creator's Patreon page.



🎨 'Animating with Comfy UI and Anime'

This paragraph outlines the process of creating an animation using Comfy UI and the Anime software. It begins with the setup of the first workflow, which includes dragging and dropping the initial workflow and setting up various components such as inputs, animation, properties, and control. The tutorial provides specific instructions on how to use the 'net' feature with batch or single operation options, how to configure case sampler settings, and how to export the video. It also covers the selection of the anime model and the addition of effects like fire, adjusting the weight of these effects, and using the correct prompts for the anime diffusion model. The paragraph concludes with the rendering of the animation and moving on to the upscaling workflow.


📚 'Upscaling and Face Fixing in Video Workflow'

The second paragraph details the steps for upscaling the video and fixing faces in the animation. It starts with dragging and dropping the video into the upscaling workflow and setting the output path, model, and other settings. The paragraph explains how to copy the video path and adjust settings such as the load cap and target resolution. It also touches on the use of an IP adapter and the rendering process. Finally, the paragraph describes the use of the video2video face fixer workflow, which includes similar settings and the addition of prompts for more detailed faces. It concludes with a note on the importance of adjusting the FPS according to the video's requirements and the final rendering process.




AnimateDiff is a term that refers to a software or tool used in the video for creating animations. It is a key component in the workflow described in the video, enabling the user to generate animated content. In the script, AnimateDiff is used in conjunction with ComfyUI to demonstrate the process of animation creation, highlighting its importance in the tutorial's main theme of animating with specific software tools.


ComfyUI appears to be the user interface of the software or platform being used to create the animation. It is mentioned alongside AnimateDiff, suggesting that it provides a comfortable and user-friendly environment for designing animations. The script emphasizes the ease of use of ComfyUI in the animation process, which is central to the tutorial's goal of teaching viewers how to make animations efficiently.


In the context of the video, a workflow refers to a sequence of steps or procedures followed to complete a task or project. The script outlines a specific workflow for animation creation, starting with inputs and moving through various stages like 'animate', 'prps', 'control', and others, which are crucial for understanding the process of making an animation as demonstrated in the tutorial.


The term 'input' in the script refers to the starting point or the initial data fed into the animation software. It is the first step in the workflow and is essential for setting up the parameters for the animation. The script mentions 'inputs' as the initial stage where users begin their animation project, indicating its foundational role in the animation process.

💡batch size

Batch size in the script pertains to the number of operations or tasks processed at one time in a workflow. The video specifies keeping the batch size at 72, which likely refers to the number of frames or operations handled simultaneously during the animation rendering process. This concept is important for understanding how the software manages rendering tasks efficiently.

💡anime model

An 'anime model' mentioned in the script refers to a specific style or template used in the animation software to create characters or elements that resemble Japanese animation (anime). The script mentions 'mune anime model' and 'concept pyromancer, Laura', which are examples of anime models used to add specific effects and styles to the animation, showcasing the customization options available in the software.

💡open pose

Open pose is a term used in the script to refer to a method or tool for extracting pose data from images. It is mentioned in the context of using 'open pose reference images', which implies that it is a technique for capturing and applying character poses in the animation. The script suggests using 'CN passes extractor' to obtain these images, indicating its role in the animation's pose and motion design.


FPS stands for 'frames per second', a measure used in the script to indicate the speed at which the animation plays. The video mentions changing the FPS to 12 when exporting the video, which affects the perceived speed of the animation's movement. This term is crucial for understanding how the pacing and fluidity of the animation are controlled.


Upscaling in the context of the video refers to the process of increasing the resolution or quality of a video. The script discusses an 'upscaling workflow' where the video is processed to enhance its resolution, using specific settings and models. This term is important for understanding how the final output of the animation is refined for better visual quality.

💡video2video face fixer

The term 'video2video face fixer' in the script seems to refer to a specific tool or process used to refine or correct the facial features in a video. It is part of the final stages of the workflow, where the video is further improved by enhancing the details of the faces. The script mentions adding prompts and upscaling to achieve better facial details, indicating its role in the final touch-ups of the animation.

💡frame interpolation

Frame interpolation is a technique mentioned in the script used to create smoother transitions between frames in a video. It is used to enhance the smoothness of the animation by adding intermediate frames. The script refers to adding frame interpolation for smoothness, which is an important aspect of the post-production process to achieve a polished animation result.


Learn to create animations using ComfyUI and AnimeD workflows.

Link to the tutorial in the description below.

Start by dragging and dropping the first workflow.

Set up inputs, animateD, prps, control, and case sampler settings.

Choose between batch or single op options in the control net.

Copy and paste the output folder path for rendered frames.

Select the dimension and batch size for the output.

Use the Mune anime model and choose the concept pyromancer, Laura.

Add cool fire effects with a weight of around 0.5.

Select the anime diff model for prompts.

Control net is turned off by default.

Use the directory for open pose reference images.

Unmute the Directory Group and enable open pose.

Change the FPS of the exporting video to 12.

Render the queue and wait for the results.

Proceed to the upscaling workflow.

Input the video and set the output path and settings.

Select the model, settings, prompts, and upscale value.

Use the video2video face fixer workflow for improved details.

Set the load cap, video settings, and model as used before.

Add prompts for more detailed faces and upscale for better quality.

Render the video with adjusted FPS for desired speed.

Use frame interpolation for smoothness with Flow Frames.

Find more workflow tutorials and resources on Patreon.

Support from Patreon helps keep tutorials free for everyone.