What You Will Learn

• Install and configure Advanced Live Portrait and Video Helper nodes in ComfyUI
• Prepare and load optimal portrait images for animation
• Animate portraits using a driving video for realistic motion transfer
• Manually shape expressions with the Expression Editor and crop factor
• Sequence multiple expressions and use motion commands for timed animations
• Preview outputs, export frames to video, and troubleshoot common issues

Study Guide

Introduction: Bringing Still Portraits to Life with ComfyUI

Imagine grabbing a still photograph,a simple, neutral portrait,and making it blink, smile, tilt its head, or even mimic the subtle nuances of your own facial movements. That’s the magic behind ComfyUI’s Advanced Live Portrait workflow. This course is your deep-dive into the world of animating portrait images using ComfyUI, focusing not only on the technical process but also on the creative possibilities and real-world applications.
In this guide, you’ll learn how to install and set up custom nodes, transform static faces using both manual and video-driven methods, fine-tune facial expressions, and export your animated creations. Whether you’re aiming to enhance digital storytelling, create engaging social content, or experiment with AI-powered art, mastering these techniques will unlock a new level of creative control. We’ll walk through every detail,from the basics of node installation to the nuances of expression sequencing,ensuring you have the practical knowledge and insight to apply these skills effectively.

Understanding the Advanced Live Portrait Node: Core Functionality and Purpose

At the heart of this workflow lies the Advanced Live Portrait node,a custom ComfyUI node designed to animate still portraits and control facial expressions. Think of it as a tool that breathes life into a static image, either by transferring motion from a video or by tweaking facial features manually.
There are two main ways to use this node:
1. Use a driving video, transferring motion and expressions from the video onto the still portrait for dynamic, realistic animation. 2. Directly manipulate facial features and expressions using the Expression Editor, no video required.
The Advanced Live Portrait node only works with portraits of people. For the best results, use a sharp, well-lit photo of a face looking directly at the camera with a neutral expression. While it can work on cartoon-like characters, the output is consistently better with real faces.
Example 1: Animate a professional headshot to deliver a personalized video message, using a webcam video as the driving input.
Example 2: Transform a friend’s selfie to make them wink and smile by adjusting parameters in the Expression Editor, no video needed.

Setting Up: Installing Custom Nodes in ComfyUI

Before you can animate portraits, you need to install two essential custom nodes: "advanced live portrait" and "video helper." These are not included in the standard ComfyUI installation but are easily added via the manager.
- Open the ComfyUI manager interface. - Search for and install the "advanced live portrait" node. - Similarly, find and install the "video helper" node, which provides video-related utilities.
Once installed, restart ComfyUI. This ensures the nodes are loaded and ready for use. The first time you run the workflow, ComfyUI will automatically download any necessary models. This is a one-time process; subsequent runs will be much faster.
Tip: If you ever encounter an error or the nodes don’t appear, try restarting ComfyUI again. The "fix node" option (right-click on a node) can also help reset a node to its default state.
Example 1: After installing and restarting, drag the "advanced live portrait" node onto the canvas,it should now be available under custom nodes.
Example 2: Add the "video combine" node from "video helper" to assemble frames into a video later in your workflow.

Building Your First Workflow: Loading Images and Setting the Foundation

Every live portrait animation starts with a quality source image. Use the "load image" node to import your portrait. For optimal results:
- Choose a photo of a real person, face-on, well-lit, and with a neutral expression. - The image should be sharp, with clear facial features. Avoid images with strong shadows, extreme angles, or heavy makeup.
Example 1: Load a studio headshot where the subject looks directly at the camera, neutral expression, even lighting.
Example 2: Import a passport-style selfie taken in natural daylight.
Practical Tip: If you’re experimenting with cartoon-like images, understand that results may be mixed. The node is trained on real human faces, so photorealistic portraits work best.

Using the Expression Editor: Manual Control of Facial Features

The Expression Editor node is your control panel for animating facial features without a driving video. Add this node and connect it to your loaded image. You can now manipulate various aspects of the face:
- Blinking: Adjust "blink" to simulate natural or exaggerated eye closure. - Mouth Movement: Open or close the mouth for speaking or smiling effects. - Eyebrow Position: Move or arch the eyebrows to convey surprise, concern, or other emotions. - Head Rotation: Adjust pitch (up/down), yaw (left/right), and roll (tilt) for subtle or dramatic head movements. - Pupil Direction: Shift the gaze to make the subject look left, right, up, or down.
These adjustments are made by dragging sliders, clicking arrows, or typing values directly into the node’s interface.
Example 1: Set "blink" to -20 to create a blinking effect, making the portrait’s eyes close naturally for one frame.
Example 2: Rotate the head slightly with yaw and pitch, and raise the eyebrows for a curious expression.
Best Practice: Enable "Q on change" (found via the small arrow next to the Q button) to preview changes live as you adjust settings. This allows you to see the effect of each tweak instantly, streamlining experimentation.

Live Preview and Output: Working with Preview and Save Nodes

The workflow can generate many frames or images, especially when animating expressions or sequencing multiple motions. To keep things efficient, use the "preview node" rather than saving every single frame immediately.
- The Preview node displays output images or frames within ComfyUI, allowing you to assess quality and select your favorites. - Right-click a frame in the preview to save it manually, reducing unnecessary file clutter. - The Save node can be added when you’re ready to export specific frames or batches to your computer.
Example 1: Animate a blink sequence and preview each frame, saving only the ones that look most natural.
Example 2: Adjust mouth movement frames for a speaking animation, using the preview to refine timing before saving.
Tip: When generating long expression sequences or video-driven animations, the preview node gives you quick feedback without overwhelming your storage with unused images.

Understanding and Using the Crop Factor

The "crop factor" in the Expression Editor is crucial for ensuring your animations look natural, especially when the head rotates or moves. The model works with a 512-pixel square around the face. The crop factor determines how much area around the face is included in the generation:
- Too small: Parts of the head, hair, or face may be cut off during movement. - Too large: The face may appear blurry or stretched, as the 512-pixel output must cover a bigger area.
Adjust the crop factor to balance detail and coverage. If you use multiple expression editors (for sequencing), adjust the crop factor for each to keep results consistent.
Example 1: Increase crop factor when the animation involves a big head turn, ensuring the ear and hair are not cut off.
Example 2: Use a smaller crop for subtle eyebrow movements, focusing tightly on the eyes and brow.
Advanced Tip: Right-click the crop factor setting and choose "convert crop to input." This lets you control crop size across several editors with a single primitive node,ideal for synchronizing crop values in complex workflows.

Animating with a Driving Video: Dynamic Motion Transfer

For realistic, dynamic animation, connect a driving video. This video provides the facial movements and expressions to be copied onto your portrait.
Steps:
- Use the "load video" node to import your driving video. - Connect both the source image and driving video to the "advanced live portrait" node. - The node analyzes the video, extracting facial motion and applying it to the still portrait.
Key Considerations:
- The driving video should feature a well-lit, clearly visible face, ideally facing the camera. - The expressions and movements in the video determine the animation on your portrait. - The source image should match the angle and lighting of the driving video as closely as possible for smoothest results.
Example 1: Record yourself smiling, blinking, and nodding, then use that video to animate a headshot of a colleague, making it appear as if they are performing those actions.
Example 2: Import a short video clip of an actor delivering a line, then transfer those expressions and mouth movements to a historical portrait for an educational project.
Tip: The first time you run this workflow, model files will be downloaded. This may take longer, but future runs will be much faster.

Outputting Animations: From Frames to Video

Once you have your sequence of generated frames,whether from manual expression editing or video-driven animation,decide how to export your results.
- Use the "video combine" node (from the video helper set) to assemble frames into a video file. - Specify parameters such as output format (e.g., h264 MP4), filename prefix, and frame rate for smooth playback.
Alternatively, save individual frames with the Save node for use in GIFs, slideshows, or further editing.
Example 1: Combine 60 frames of a talking portrait into a 2-second MP4, ready for social media sharing.
Example 2: Export single frames showing different expressions for use as avatars or profile pictures.

Sequencing Expressions: Working with Multiple Expression Editors

For advanced animations, chain multiple Expression Editor nodes together. This lets you define a series of expressions or motions to be played in sequence,without relying on a driving video.
- Each Expression Editor defines a different facial pose or movement. - The "motion link" connects editors, passing the output from one to the next.
To control the order and duration of these expressions, use motion commands in the Advanced Live Portrait node. The format:
motion_number = changing_frame_length : length_of_frames_waiting_for_the_next_motion
- motion_number: Identifies which expression to use (0 is the original image, 1 is the first editor, etc.). - changing_frame_length: Number of frames spent transitioning into this expression. - length_of_frames_waiting_for_the_next_motion: Number of frames held before moving to the next.
Example 1: Animate a sequence where the subject blinks (editor 1), then smiles (editor 2), then returns to neutral (editor 0):
1=10:5; 2=15:10; 0=5:0
This means: transition into blink over 10 frames, hold 5 frames; transition to smile over 15 frames, hold 10 frames; transition back to neutral over 5 frames.
Example 2: Create a subtle nod, then a surprised eyebrow raise, then a glance to the left, each defined in separate editors and sequenced with tailored frame lengths and wait times.
Best Practice: The "changing frame length" (the first number after the equal sign) must not be zero,otherwise, the workflow throws a division by zero error. Always specify at least one frame for transitions.

Animating Without a Driving Video: The “Animate Without Video” Setting

You don’t always need a driving video to create animation. The "animate without video" setting in the Advanced Live Portrait node allows you to rely solely on the defined sequence of expressions (via editor nodes).
- Set "animate without video" to true. - The node will ignore any driving video input and use your motion commands and editors instead.
Example 1: Design a looping sequence where the portrait winks, smiles, and looks around,all controlled by editor nodes and played in order, no video required.
Example 2: Create a step-by-step animation for an educational explainer, with each step defined by a different expression editor.
Tip: This approach gives you full creative control over timing and expression, perfect for stylized or non-naturalistic animations.

Advanced Workflow Techniques: Synchronizing Crop Factors and Expressions

Complex workflows often involve multiple Expression Editors, each with its own settings. Consistency is key:
- Synchronize crop factors across editors by using the "convert crop to input" option and connecting a shared primitive node. - This ensures all motion steps use the same framing, preventing sudden jumps or mismatches in head size/position.
Example 1: Use a primitive node set to "1.2" as input for the crop factor of five different editors, guaranteeing smooth transitions.
Example 2: Adjust the primitive node value to fine-tune the crop area for the entire sequence, instantly updating all editors.

Limitations and Troubleshooting: Common Issues and Solutions

No workflow is perfect. Here are issues you might face and how to address them:
1. Blurry Faces: The model outputs at 512x512 pixels. If you use a larger area (via crop factor), the face can appear blurry when stretched.
Solution: Keep the crop factor as small as possible while still including all relevant features. Avoid upscaling the output excessively.
2. Visible Crop Lines: Sometimes a line or cut appears around the crop area, especially at the neck.
Solution: Use a darker or less detailed background to make the transition less noticeable. Adjust lighting or background color in the original image to help blend the edges.
3. Model Download/First Run: The first workflow run is slow as it downloads required models.
Solution: Be patient. Subsequent runs are much faster.
4. Division by Zero Error: If a motion command has a changing frame length of zero, it will cause an error.
Solution: Always specify a positive, nonzero number for changing frame length in motion commands.
5. Output Management: Generating hundreds of frames can clutter your workspace.
Solution: Use the Preview node to select and save only the best results.
Example 1: If you notice the animation skips or jumps, check if all crop factors match and motion commands are correct.
Example 2: If artifacts appear around the neck, add a dark vignette to the background in your source image.

Creative Applications and Extensions: Beyond Basic Portraits

The Live Portrait workflow is versatile, but works best with real human faces. While you can experiment with cartoon or stylized characters, expect less realistic results. The model may not fully capture exaggerated or non-human features.
- For best results, stick to photos of real people, straight-on, neutral, and well-lit. - Use cartoon or stylized faces for fun experiments, but review outputs carefully.
Example 1: Animate your own selfie for a personalized video greeting.
Example 2: Try animating a digital painting or a character from a game, observing how the workflow interprets non-photorealistic features.
Tip: Record your own facial expressions as driving videos for maximum creative control.

Comparing Methods: Expression Editor vs. Driving Video

Each animation method has unique strengths:
Expression Editor: - Full manual control over every facial parameter. - Perfect for stylized, step-by-step, or non-naturalistic animation. - No video input required, so you can script precise sequences.
Example: Animating a portrait to teach sign language facial cues, with each expression clearly defined.
Driving Video: - Captures subtle, natural movements from real video. - Fast way to transfer speech, emotion, and gestures to a still image. - Limited to the expressions and angles present in the video.
Example: Making a historical figure "speak" a modern message by mimicking your own facial movements.
Choose the method that fits your creative intention. Use both in combination for even greater flexibility.

Glossary of Key Terms

- Advanced Live Portrait: Custom node for animating still portraits. - Expression Editor: Node for manual adjustment of facial features. - Load Image: Node for uploading static images. - Load Video: Node for video input, used as a driving video. - Preview Node: Displays output for immediate review. - Save Node: Saves generated images or frames. - Video Combine: Combines frames into a video file. - Q on change: Enables live preview on parameter change. - Crop Factor: Sets the area around the face included in generation. - Driving Video: Provides motion and expression data to animate the portrait. - Animate without video: Enables animation sequences without a driving video. - Motion Link: Connects multiple Expression Editors in sequence. - Motion Number: Identifies a specific expression for sequencing. - Frame Length: Number of frames for each motion step. - Wait Time: Pause before the next expression. - Fix node: Resets a node to default settings. - Convert crop to input: Makes crop factor adjustable across multiple nodes. - Primitive node: Sets and outputs a specific value (e.g., crop factor). - Yaw, Pitch, Roll: Directions of head movement (left/right, up/down, tilt).

Conclusion: Mastering Live Portrait Animation in ComfyUI

You now have a comprehensive roadmap for animating still portraits with ComfyUI’s Advanced Live Portrait workflow. From installing custom nodes and selecting the right images, to mastering manual and video-driven animation, synchronizing expression sequences, and troubleshooting common pitfalls, you’re ready to bring your creative visions to life.
Remember: - Quality source images and videos yield the best results. - Experiment with both manual (Expression Editor) and video-driven methods to find your ideal workflow. - Use preview nodes for efficient review and output management. - Pay attention to crop factors and motion commands for smooth, professional-looking animations. - Don’t hesitate to try new combinations,record your own expressions, sequence multiple motions, or test out cartoon portraits for fun.
The possibilities are limited only by your imagination and your willingness to experiment. Take these skills, apply them to your projects, and watch your static images transform into engaging, dynamic works of art.

Frequently Asked Questions

This FAQ section is a practical guide for anyone working with ComfyUI’s Live Portrait & Face Expressions workflow. It answers the most common questions,both technical and strategic,about animating still portraits, using facial expressions, and getting the best results for business or creative projects. Whether you’re just starting or looking to refine your process, you’ll find clear, actionable answers below.

What is Live Portrait in ComfyUI?

Live Portrait is a tool available as a custom node in ComfyUI that allows you to add animation and facial expressions to still portrait images.
It can bring a static photo to life by either applying preset facial expression changes or by transferring facial movements from a driving video. This is particularly useful for enhancing presentations, marketing materials, or interactive content where dynamic visuals make a stronger impact.

How do I install the necessary nodes for Live Portrait in ComfyUI?

To use Live Portrait, you need to install the 'advanced live portrait' node through the ComfyUI Manager.
If you plan to use video input or output, you'll also need the 'video helper' nodes. After installation, restart ComfyUI for the nodes to become available. This ensures all dependencies are loaded and the workflow operates smoothly.

What kind of image works best as a source for Live Portrait?

For optimal results, the source image should be a sharp, well-lit portrait of a person's face looking straight at the camera.
A neutral expression on the subject often provides the smoothest base for animation. The tool is primarily designed for real people, though it can sometimes work with cartoon-like characters. Using high-quality images reduces issues like blurriness or unnatural movements.

How can I animate a still portrait using only settings in Live Portrait?

After loading your portrait and adding an 'expression editor' node, you can directly manipulate various settings like 'blink', 'open mouth', 'move eyebrows', 'rotate pitch', 'yaw', and 'roll', or 'change pupil direction'.
By adjusting these values and using 'Q on change' for live previews, you can create custom facial animations without a video input. You can chain multiple expression editors to build sequences of different movements and expressions tailored to your scenario.

How can I animate a still portrait using a video as a driver?

To use a video for animation, you'll need the 'advanced live portrait' node and a 'load video' node connected to the 'drive images' input.
The facial movements and expressions from the video will be transferred to your still portrait. For best results, the driving video should have a clear, well-lit face with visible movements. This approach is powerful for creating realistic talking avatars or mimicking specific expressions.

What is the purpose of the 'crop factor' setting in Live Portrait?

The 'crop factor' setting ensures the area you want animated,like hair or head parts,is captured during the process.
Since the model is trained on 512-pixel images, increasing the crop factor can prevent important features from being cut off during movements. However, setting it too high may include unnecessary space and reduce the clarity of the face. For workflows with multiple editors, consider converting crop to an input for easy, synchronized adjustments.

How can I combine different facial expressions in a sequence?

Create complex sequences by chaining multiple 'expression editor' nodes using 'motion link'.
In the 'advanced live portrait' node, specify which motion to display, for how many frames, and the transition length using commands like [motion number]=[changing frame length]:[waiting for next motion length]. Set 'animate without video' to true if you don't have a driving video. This structure gives you precise control over the animation flow.

What are some potential limitations or issues when using Live Portrait?

Limitations include blurriness due to 512-pixel face generation and visible lines where the automatic crop is applied.
Scaling up outputs can reduce sharpness. Uniform or light backgrounds may reveal crop lines. Use darker backgrounds to minimize this or adjust the crop factor for better alignment. When chaining motions, the changing frame length must be greater than one, and you must include a second length after the colon to avoid errors.

What is the primary function of the Advanced Live Portrait node?

The Advanced Live Portrait node animates still portraits by adding facial movements and expressions.
It can work with manual expression settings or by transferring motion from a driving video. This node is the core of the workflow, handling the transformation from static to animated visuals.

Why should I use the Preview node instead of the Save node when working with Live Portrait?

The Preview node lets you review generated results and save only the best frames by right-clicking.
Since many images can be generated in a single session, this saves time and storage space. You can quickly experiment and iterate without cluttering your files with unwanted outputs.

Why does the system download models the first time a workflow is run?

Necessary models are downloaded on first use to provide the data required for the nodes to function.
This is an automatic process that ensures you have the correct files to generate animations. Once downloaded, they are stored locally for future sessions, so you won’t experience repeated delays.

How can I see the effect of changing Expression Editor settings quickly?

Enable the 'Q on change' option from the small arrow menu next to the Q button.
This triggers a workflow run automatically whenever you adjust a parameter, giving you immediate feedback in the Preview node. It’s an efficient way to fine-tune expressions without manually cuing each update.

Which node is used to add a video input to the Advanced Live Portrait workflow?

The Load Video node is used to bring a video file into the workflow.
Connect its output to the 'drive images' input of the Advanced Live Portrait node. This setup allows the system to analyze movements in the video and apply them to your static portrait.

How do I specify the order and duration of expressions when combining multiple Expression Editor nodes?

Add commands in the Advanced Live Portrait node using the format: motion_number=frame_length:wait_time.
This command sequence tells the system which expression to display, for how long, and how long to pause before changing. For example, "1=30:10" means show motion 1 for 30 frames, then wait 10 frames before transitioning.

What setting allows animation without a driving video input?

The 'animate without video' setting in the Advanced Live Portrait node must be set to true.
This enables animation based solely on your defined sequence of expressions and transitions, removing the need for a driving video. It’s especially useful for custom presentations or simple facial animations.

What is the purpose of the Expression Editor node?

The Expression Editor node lets you manually adjust facial features and parameters.
You can set values for blinking, mouth movement, eyebrow position, or head rotation. This hands-on control is ideal for fine-tuning expressions to match specific business needs, such as making a portrait smile or look curious in a pitch video.

Can I use Live Portrait on cartoon or non-photorealistic images?

Live Portrait is primarily designed for real photos but can sometimes work with cartoon-like characters.
Success depends on how closely the cartoon matches a real human face in structure and features. Results might vary,some cartoon faces animate well, while others may show artifacts or unnatural movements. Testing with different images is the best way to evaluate fit.

How do I export my animation as a video?

Use the Video Combine node (from video helper nodes) to join a sequence of frames into a video file.
After generating your frames, connect them to the Video Combine node, specify output settings, and save the final video. This makes it easy to share animated portraits in presentations, emails, or social media.

What should I consider when selecting a driving video?

The driving video should feature a clearly visible, well-lit face with meaningful expressions or movements.
Avoid videos with extreme angles, rapid lighting changes, or heavy occlusion (like hands covering the face). The closer the driving face matches the source portrait in orientation and style, the more natural the animation will look.

What are Motion Link and Motion Number in the workflow?

Motion Link connects multiple Expression Editor nodes to form a sequence, while Motion Number identifies each motion in commands.
For example, motion 0 is usually the original image, while motion 1, 2, etc., refer to specific edited expressions. This system streamlines complex animation sequences.

What is the role of Primitive nodes in this workflow?

Primitive nodes are used to define and output specific values, like numerical inputs for the crop factor.
By converting the crop factor to an input on multiple Expression Editor nodes, you can link them to a single primitive node for consistent adjustments across the workflow. This saves time and maintains uniformity.

How can I fix a node to its default settings?

Right-click a node and select the Fix node option.
This recreates the node with its original settings, which can be useful if you've made multiple changes and want to start over without deleting and re-adding the node.

What do Yaw, Pitch, and Roll mean in the Expression Editor?

Yaw refers to left/right head rotation, Pitch is up/down, and Roll is tilting side to side.
Adjusting these lets you simulate head movements for more dynamic animations. For example, a slight yaw can make a portrait appear to look at someone off-camera, adding engagement in a business context.

How can I make sure the face does not get cut off during animation?

Increase the crop factor to include more area around the face.
If you notice parts like hair or the chin being clipped during rotations, bump the crop factor up slightly. Test with different values to find the sweet spot between coverage and image clarity.

How do I make my animation look more natural?

Use subtle, smooth adjustments in the Expression Editor and select driving videos with realistic, moderate facial movements.
Overly exaggerated expressions or rapid transitions may look artificial. For business presentations or marketing, aim for expressions that feel genuine and relatable.

How can I troubleshoot blurry outputs?

Blurriness often results from scaling up 512-pixel outputs or using low-quality source images.
Try starting with higher-quality input images, avoid excessive scaling, and adjust the crop factor to keep the face area as focused as possible. For marketing or client-facing content, consider post-processing tools to enhance sharpness if needed.

What does Q on change do and why is it helpful?

'Q on change' automatically cues a workflow run whenever you tweak a parameter in a connected node.
This gives instant feedback in the Preview node, making it easier to fine-tune animations without repetitive manual steps. It's a productivity booster, especially when iterating on facial expressions.

Can I use this workflow for business or corporate presentations?

Absolutely. Live Portrait animations can add engagement to slides, demos, or explainer videos.
Business professionals use animated avatars to communicate messages, demo products, or personalize communications. Ensure you have the right to use the source images, especially for client work or public presentations.

What are best practices for saving final outputs?

Use the Preview node to review and select your best frames, then save them to your chosen directory.
When exporting videos, double-check resolution and format settings for compatibility with your intended platform (e.g., PowerPoint, social media, websites).

What strategies can help reduce visible crop lines in final animations?

Use darker backgrounds in source images and adjust the crop factor carefully.
Test with a variety of settings and consider subtle background gradients or overlays to blend any noticeable lines. This ensures a cleaner, more professional result.

How do I handle errors related to motion sequencing?

Ensure your sequence commands have changing frame lengths greater than one and include a wait time after the colon.
For example, '1=10:0' is valid, but '1=1' or missing the colon may trigger errors. Double-check your command syntax if you run into issues.

Are there any privacy or ethical considerations when animating portraits?

Always have permission to use and animate someone's image, especially in professional or public contexts.
For marketing, training, or HR content, ensure your use of avatars aligns with company policies and respects personal privacy. Avoid misrepresenting real people with manipulated expressions.

Can I automate or batch process animations with ComfyUI?

ComfyUI workflows can be designed for batch processing by chaining nodes and using parameterized inputs.
For repetitive tasks, consider creating templates where only the input image or driving video changes. This is efficient for generating multiple animated avatars for a team or client list.

How does Live Portrait differ from other animation tools?

Live Portrait in ComfyUI offers node-based, customizable control over facial animations.
Unlike some automatic animation apps, it allows for granular adjustments and integration with broader image or video workflows. This flexibility appeals to those who want both creative control and technical depth.

What should I do if my output video is choppy or has missing frames?

Check your frame lengths and transition settings in the Advanced Live Portrait node.
If the animation is too abrupt, increase frame counts or add longer wait times between expressions. Also, ensure your system resources are sufficient for processing larger batches of images smoothly.

How can I share my animated portraits with others?

Export animations as GIFs or video files and share via email, cloud storage, or social media.
For business settings, embed videos in presentations or upload them to internal platforms for easy access by your team or clients.

Where can I find more examples or community support for ComfyUI Live Portrait?

Community forums, Discord servers, and online groups focused on ComfyUI are valuable resources.
You’ll find workflow templates, troubleshooting tips, and inspiration from other users. Sharing your own results often leads to helpful feedback and ideas for improvement.