Load Video FFmpeg (Path) 🎥🅥🅗🅢(LoadVideoFFmpegPath)
The main function of this node is to parse video files into frame-by-frame images (or latent representations), while extracting audio tracks and detailed video information (such as frame rate, resolution, duration).
Load Video F Fmpeg Path
The Load Video FFmpeg (Path) 🎥🅥🅗🅢 node primarily functions to parse a video file into frame-by-frame images (or latent representations), while also extracting the audio track and video metadata (such as frame rate, resolution, duration).
The Load Video FFmpeg (Path) node is a core component of the Video Helper Suite toolkit, used to load video files from the local filesystem or network URLs. Built upon the powerful FFmpeg multimedia framework, it can decode and process various video formats, converting them into image sequences, audio, and metadata usable within ComfyUI workflows.
Node Functionality
The primary function of this node is to parse a video file into frame-by-frame images (or latent representations), while extracting the audio track and video details (like frame rate, resolution, duration). This allows users to seamlessly integrate video content into image generation, editing, or video processing pipelines—for example, applying AI image processing to each frame of a video or performing audio-visual separation analysis.
Node Parameter Description - Load Video FFmpeg (Path) 🎥🅥🅗🅢
Connection Inputs (Inputs)
| Parameter Name | Data Type | Required | Default Value | Range/Options | Description |
|---|---|---|---|---|---|
meta_batch | VHS_BATCHMANAGER | No | - | - | Connect a VHS_BatchManager to enable video batch processing or sequential loading functionality. |
vae | VAE | No | - | - | Connect a VAE model; the node will then output latent space representations of the video frames instead of raw pixel images. |
Control Parameters (Parameters)
| Parameter Name | Data Type | Required | Default Value | Range/Options | Description |
|---|---|---|---|---|---|
video | STRING | Yes | - | - | Specifies the full path or network URL of the video file to load. Supports common video formats like .mp4, .avi, .mov, etc. |
force_rate | FLOAT/INT (Combobox) | Yes | 0 | 0 – 60 (Step: 1) | Forces the frame rate of the output image sequence, range 0-60, step 1. Uses the video's original frame rate when set to 0. |
custom_width | INT | Yes | 0 | 0 – DIMMAX | Customizes the width of the output images, range 0-DIMMAX. Maintains the video's original width when set to 0. |
custom_height | INT | Yes | 0 | 0 – DIMMAX | Customizes the height of the output images, range 0-DIMMAX. Maintains the video's original height when set to 0. |
frame_load_cap | INT | Yes | 0 | 0 – BIGMAX (Step: 1) | Limits the maximum number of frames to load, range 0-BIGMAX, step 1. Loads all frames when set to 0. |
start_time | FLOAT | Yes | 0 | 0 – BIGMAX (Step: .001) | Sets the start time (in seconds) from which to begin loading the video, range 0-BIGMAX, step 0.001. |
Outputs
| Parameter Name | Data Type | Description |
|---|---|---|
image | IMAGE or LATENT | Outputs the loaded video frame sequence as an image tensor (if no VAE connected) or as a latent representation (if a VAE is connected). |
mask | MASK | Outputs the video's inherent transparency channel (Alpha channel) as a mask. |
audio | AUDIO | Outputs the audio track data extracted from the video. |
video_info | VHS_VIDEOINFO | Outputs video metadata such as frame rate, resolution, duration, and codec information. |
Usage Scenarios
The node supports fine-grained control over the loading process, such as specifying a start time, limiting the number of frames to load, forcing an output frame rate, or adjusting resolution. It is also compatible with batch managers, facilitating the handling of video sequences or clips. A typical use case is: a user loads a short video, adjusts its dimensions to match model requirements, and then feeds each frame into an image-to-image transformation node to achieve style transfer or content restoration.
Load Video FFmpeg (Path) 🎥🅥🅗🅢 Node Source Code Link
The Load Video FFmpeg (Path) 🎥🅥🅗🅢 node is from the ComfyUI-VideoHelperSuite node package.
Comments
Sign in with GitHub to join the discussion.