ComfyUI XY Video Plot Suite

A suite of ComfyUI nodes for generating parameter sweep video matrices with labeled grids. Create comprehensive visual comparisons of how different parameter combinations affect your video generation results.

Example Output

To load the example workflow: Download and drag examples/wan_xy_plot.json into ComfyUI

▶️ View Full Video Output

A suite of ComfyUI nodes for generating parameter sweep video matrices with labeled grids. Create comprehensive visual comparisons of how different parameter combinations affect your video generation results.

Features

• Parameter Sweep Automation: Test multiple parameter combinations automatically
• Labeled Grid Output: Clear X/Y axis labels showing which parameters generated each video
• For Loop Integration: Works seamlessly with comfyui-easy-use For Loop nodes
• Customizable Layout: Adjust cell sizes, fonts, and colors
• Batch Organization: Optional batch IDs for organized file management
• Auto Cleanup: Optionally delete source videos after grid assembly

Installation

1. Install Dependencies

cd ComfyUI/custom_nodes/comfyui-xyvideoplot2
pip install -r requirements.txt

Or manually:

pip install moviepy

2. Clone Repository

cd ComfyUI/custom_nodes
git clone https://github.com/yourusername/comfyui-xyvideoplot2

3. Restart ComfyUI

Quick Start

Basic Workflow Structure

XY Plot Setup
    ↓ initial_collection_data → For Loop Start (initial_value1)
    ↓ total_iterations → For Loop Start (iterations)
    ↓ xy_config (pass through)
    ↓ batch_id (optional - use in filename_prefix)

For Loop Start
    ↓ loop_index → XY Plot Get Values
    ↓ value1 (collection_data) → XY Plot Collect Video

XY Plot Get Values
    ↓ x_value → KSampler (or other parameter)
    ↓ y_value → KSampler (or other parameter)
    ↓ x_label, y_label → XY Plot Collect Video

[Your Generation Workflow]
    KSampler → Images → VHS Video Combine
    ↓ video_filepath → XY Plot Collect Video

XY Plot Collect Video
    ↓ collection_data → For Loop End (value1)

For Loop End
    ↓ final_value1 → XY Plot Video Grid

XY Plot Video Grid
    ↓ video_grid (IMAGE tensor) → Save Image/Video

Example Setup

XY Plot Setup:

• X-axis: seed (type: INT, values: 1, 3, 5)
• Y-axis: cfg (type: FLOAT, values: 2.0, 7.5)

This creates a 3×2 grid testing all 6 combinations.

Node Reference

1. XY Plot Setup

Initialize parameter sweep configuration.

Inputs:

• x_axis_name (STRING): Display name for X-axis (e.g., "seed")
• x_axis_type (ENUM): Data type - INT, FLOAT, or STRING
• x_values (STRING): Comma-separated values (e.g., "1, 2, 3")
• y_axis_name (STRING): Display name for Y-axis (e.g., "cfg")
• y_axis_type (ENUM): Data type - INT, FLOAT, or STRING
• y_values (STRING): Comma-separated values (e.g., "7.0, 8.0, 9.0")

Outputs:

• initial_collection_data (STRING): Empty array "[]" for For Loop Start
• total_iterations (INT): Number of iterations (len(x) × len(y))
• xy_config (STRING): JSON configuration (pass through workflow)
• batch_id (STRING): Unique timestamp ID (YYYYMMDD_HHMMSS) for organizing outputs

Note: Connect batch_id to your video generation node's filename_prefix for organized file naming.

2. XY Plot Get Values

Extract current X/Y parameter values for the loop iteration.

Inputs:

• xy_config (STRING): From XY Plot Setup
• loop_index (INT): From For Loop Start

Outputs:

• x_value (INT/FLOAT/STRING): Current X value (dynamically typed)
• y_value (INT/FLOAT/STRING): Current Y value (dynamically typed)
• x_label (STRING): Display label (e.g., "seed: 1")
• y_label (STRING): Display label (e.g., "cfg: 7.5")
• xy_config (STRING): Pass-through

Important: Connect x_value/y_value directly to your parameter inputs (e.g., KSampler seed, cfg). ComfyUI handles type conversion automatically.

3. XY Plot Collect Video

Accumulate video paths and labels during loop execution.

Inputs:

• video_filepath (STRING): Path from video generation node (e.g., VHS Video Combine)
• x_label (STRING): From XY Plot Get Values
• y_label (STRING): From XY Plot Get Values
• xy_config (STRING): Pass-through
• collection_data (STRING, optional): From For Loop Start value1

Outputs:

• collection_data (STRING): Updated JSON array → For Loop End value1
• xy_config (STRING): Pass-through

4. XY Plot Video Grid

Assemble all collected videos into a labeled grid.

Inputs:

• collection_data (STRING): From For Loop End final_value1
• xy_config (STRING): Pass-through from workflow
• max_cell_width (INT, default: 512): Maximum width per cell (maintains aspect ratio)
• banner_font_size (INT, default: 20): Title text size
• label_font_size (INT, default: 14): Axis label text size
• font_color (STRING, default: "white"): Text color
• bg_color (STRING, default: "black"): Background color
• cleanup_files (BOOLEAN, default: True): Delete source videos after assembly

Outputs:

• video_grid (IMAGE): Assembled grid as tensor (connect to Save Image/Video)
• metadata (STRING): JSON with grid info (shape, labels, fps, duration)

Connection Guide

Critical Connections

1. For Loop Start Configuration:

   • initial_value1 ← XY Plot Setup: initial_collection_data
   • iterations ← XY Plot Setup: total_iterations

2. Collection Data Flow (through loop):

   • For Loop Start: value1 → XY Plot Collect Video: collection_data
   • XY Plot Collect Video: collection_data → For Loop End: value1
   • For Loop End: final_value1 → XY Plot Video Grid: collection_data

3. Config Pass-Through (direct connections):

   • XY Plot Setup: xy_config → XY Plot Get Values: xy_config
   • XY Plot Get Values: xy_config → XY Plot Collect Video: xy_config
   • XY Plot Collect Video: xy_config → XY Plot Video Grid: xy_config

Common Mistakes

❌ DON'T manually type "[]" into For Loop Start - use initial_collection_data output
❌ DON'T connect collection_data directly between nodes - it must flow through For Loop ports
❌ DON'T forget to connect both x_label and y_label to XY Plot Collect Video

Troubleshooting

Videos not in correct grid positions

• Check that your Y/X value ordering matches your expectations
• Grid is row-major: iterates Y values (outer) then X values (inner)

Type mismatch errors

• Verify x_axis_type/y_axis_type match your parameter requirements
• Most ComfyUI nodes auto-convert STRING to INT/FLOAT when possible

Missing videos in grid (black cells)

• Check video generation is completing successfully
• Verify video file paths are valid
• Check console for "Video file not found" warnings

Text clipping or alignment issues

• Increase max_cell_width for larger grid
• Adjust label_font_size if text is too large/small
• Ensure all source videos have the same aspect ratio

For Loop not completing

• Verify total_iterations connects to For Loop Start iterations
• Check loop isn't set to "Pause" mode in For Loop settings

Advanced Usage

Using Batch ID for Organization

Connect batch_id output from XY Plot Setup to your video node's filename_prefix:

XY Plot Setup: batch_id → [VHS Video Combine] filename_prefix

Result: Videos named like 20251003_143052_00001.mp4, 20251003_143052_00002.mp4, etc.

Custom Video Dimensions

The grid automatically detects aspect ratio from the first video. For best results:

• Keep all videos at the same resolution
• Adjust max_cell_width based on your source video size (e.g., 512 for 512×512, 1024 for 1024×1024)

Preserving Source Videos

Set cleanup_files to False in XY Plot Video Grid if you want to keep individual videos after grid assembly.

Example Workflow

See examples/wan_xy_plot.json for a complete working example. Load it in ComfyUI to see the full node setup.

View examples/wan_vace_xy_plot.mp4 to see the output grid format.

Requirements

• ComfyUI (latest version recommended)
• comfyui-easy-use custom nodes (for For Loop)
• Python packages:
  • moviepy>=1.0.3
  • torch (included with ComfyUI)
  • numpy (included with ComfyUI)
  • pillow (included with ComfyUI)

Example Workflow

See example_workflow.json for a complete working example.

License

[Your License Here - e.g., MIT]

Credits

Created for the ComfyUI community. Built with:

• ComfyUI
• comfyui-easy-use
• MoviePy

Support

• Report issues: GitHub Issues
• Discussions: GitHub Discussions