Load Video (Path) 🎥🅥🅗🅢 (LoadVideoPath) - ComfyUI-VideoHelperSuite

The Load Video (Path) 🎥🅥🅗🅢 node offers extensive video processing options, allowing users to finely control the loading process.

The Load Video (Path) node is a core component of the Video Helper Suite toolkit, used to load video files from the local filesystem or network URLs. Its primary function is to decode video files into a series of consecutive image frames (IMAGE) for subsequent image processing, analysis, or generation tasks within ComfyUI workflows.

Node Functionality

This node provides extensive video processing options, allowing users to finely control the loading process. For example, users can limit the number of frames loaded, skip a specified number of initial frames, sample frames at intervals, and even force the output frame dimensions and frame rate. This makes it highly suitable for handling large video files or scenarios requiring specific preprocessing steps.

Node Parameter Description - Load Video (Path) 🎥🅥🅗🅢

Connection Inputs (Inputs)

Parameter Name	Data Type	Required	Default Value	Range/Options	Description
`meta_batch`	VHS_BatchManager	No	-	-	: Connects to a batch manager for coordinating the loading of multiple videos in batch mode.
`vae`	VAE	No	-	-	: Connects a VAE model. If provided, the node directly outputs the latent space representation (LATENT) of the video frames; otherwise, it outputs images (IMAGE).

Control Parameters (Parameters)

Parameter Name	Data Type	Required	Default Value	Range/Options	Description
`video`	STRING	Yes	-	-	: Specifies the path to the video file or network URL to load. Supports common video formats (e.g., .mp4, .mov, .avi).
`force_rate`	COMBO	Yes	0	0 - 60 (Step: 1)	: Forces the frame rate (FPS) of the output frame sequence. Range is 0 to 60, step is 1. Setting it to 0 uses the video's original frame rate.
`custom_width`	INT	Yes	0	0 - DIMMAX	: Forces the width of the output images. Range is 0 to DIMMAX. Setting it to 0 preserves the video's original width.
`custom_height`	INT	Yes	0	0 - DIMMAX	: Forces the height of the output images. Range is 0 to DIMMAX. Setting it to 0 preserves the video's original height.
`frame_load_cap`	INT	Yes	0	0 - BIGMAX (Step: 1)	: Limits the maximum number of frames to load. Range is 0 to BIGMAX, step is 1. Setting it to 0 loads all frames.
`skip_first_frames`	INT	Yes	0	0 - BIGMAX (Step: 1)	: Skips a specified number of frames from the beginning of the video before starting to load. Range is 0 to BIGMAX, step is 1.
`select_every_nth`	INT	Yes	1	1 - BIGMAX (Step: 1)	: Loads every Nth frame for downsampling. Range is 1 to BIGMAX, step is 1.

Output

Parameter Name	Data Type	Description
IMAGE	UNKNOWN	: The output video frame sequence (image or latent space representation).
frame_count	INT	: The total number of video frames successfully loaded.
audio	AUDIO	: The audio track data extracted from the video.
VHS_VIDEOINFO	VHS_VIDEOINFO	: Video metadata including original dimensions and frame rate.

Usage Scenarios

A key feature of this node is its support for connecting to a VAE (Variational Autoencoder). When a VAE model is provided, the node directly encodes video frames into a latent space representation (LATENT) for output, bypassing the explicit image decoding step. This can significantly improve processing efficiency, especially in workflows that require feeding video into latent space-based generative models (such as Stable Diffusion).

Notes

Additionally, the node outputs video metadata (such as original dimensions and frame rate), the extracted audio track, and the total number of frames actually loaded. These outputs provide necessary contextual information for subsequent nodes, facilitating the construction of complex video processing pipelines.

Load Video (Path) 🎥🅥🅗🅢 Node Source Code Link

The Load Video (Path) 🎥🅥🅗🅢 node is from the ComfyUI-VideoHelperSuite node package.

Load Video (Path) 🎥🅥🅗🅢(LoadVideoPath)

Load Video Path