comfyui-myloadvideo

ComfyUI-MyLoadVideo

A custom node for ComfyUI that provides video loading and image batch splitting functionality.

Features

MyLoadVideo

Load video files and extract frames as IMAGE batch for ComfyUI workflows.

Parameters:

• video: Select video file from input directory
• skip_first_frames: Number of frames to skip at the beginning (default: 0)
• frame_load_cap: Maximum number of frames to load (default: 171)

Output:

• IMAGES: Image batch tensor
• frame_count: Number of frames loaded

Supported formats:

• MP4, AVI, MOV, MKV, GIF

MySplitImageBatch

Split or slice an IMAGE batch to extract specific frame ranges.

Parameters:

• images: Input IMAGE batch
• start_index: Starting frame index (0 = first frame)
• length: Number of frames to extract
• select_every_nth: Sample every Nth frame (optional, default: 1)

Output:

• IMAGES: Sliced image batch
• frame_count: Number of frames extracted

MyDropLastFrames

Drop the last N frames from an IMAGE batch.

Parameters:

• images: Input IMAGE batch
• drop_last_n: Number of frames to drop from the end (default: 5)

Output:

• IMAGES: Image batch with last N frames removed
• frame_count: Number of frames remaining

Installation

Method 1: Clone to custom_nodes (Recommended)

cd ComfyUI/custom_nodes
git clone https://github.com/1123156819/comfyui-myloadvideo.git

Note: The main file is already named __init__.py, so no renaming is needed!

Method 2: Download ZIP

1. Download from GitHub: Download ZIP
2. Extract to ComfyUI/custom_nodes/comfyui-myloadvideo
3. No renaming needed! The file is already __init__.py

Requirements

• ComfyUI
• OpenCV (opencv-python)
• NumPy
• PIL (Pillow)

These are typically pre-installed with ComfyUI.

Usage Examples

Example 1: Load Video and Extract First 100 Frames

Load Video → VAE Decode → MySplitImageBatch (start=0, length=100) → Video Combine

Example 2: Extract Middle Frames (50-150)

Load Video → VAE Decode → MySplitImageBatch (start=50, length=100) → Video Combine

Example 3: Drop Last 5 Frames

VAE Decode (200 frames) → MyDropLastFrames (drop_last_n=5) → Video Combine (195 frames)

Example 4: Frame Sampling (Every 2nd Frame)

Load Video → VAE Decode → MySplitImageBatch (start=0, length=50, select_every_nth=2) → Video Combine

Example 5: Combined Usage

VAE Decode (200 frames)
    ↓
MySplitImageBatch (start=10, length=150)
    ↓
MyDropLastFrames (drop_last_n=5)
    ↓
Video Combine (145 frames)

Parameter Details

MyLoadVideo

| Parameter | Type | Default | Description | |-----------|------|---------|-------------| | video | File | Required | Video file from input directory | | skip_first_frames | Int | 0 | Frames to skip at start | | frame_load_cap | Int | 171 | Maximum frames to load |

MySplitImageBatch

| Parameter | Type | Default | Description | |-----------|------|---------|-------------| | images | IMAGE | Required | Input image batch | | start_index | Int | 0 | Starting frame index (0-based) | | length | Int | 100 | Number of frames to extract | | select_every_nth | Int | 1 | Sample interval (1=all, 2=every 2nd, etc.) |

MyDropLastFrames

| Parameter | Type | Default | Description | |-----------|------|---------|-------------| | images | IMAGE | Required | Input image batch | | drop_last_n | Int | 5 | Number of frames to drop from end |

Use Cases

1. Video Trimming: Load a video and trim to specific frame range
2. Frame Sampling: Extract every Nth frame for analysis or processing
3. Batch Processing: Process only specific portions of video sequences
4. Memory Optimization: Load only needed frames to reduce memory usage
5. Remove Unwanted Frames: Drop last N frames to remove artifacts at video end

Troubleshooting

Video not showing in dropdown?

• Ensure video file is in ComfyUI's input directory
• Supported formats: MP4, AVI, MOV, MKV, GIF

Out of memory errors?

• Reduce frame_load_cap or length parameter
• Process video in chunks

Node not appearing?

• Restart ComfyUI after installation
• Check console for import errors

License

MIT License

Author

GitHub: @1123156819

Contributing

Contributions are welcome! Please feel free to submit issues or pull requests.