TrimVideoLatent Node

Trim video frames in latent space

The TrimVideoLatent node is used to trim video frames in latent space (LATENT). It is commonly used when processing video latent sequences to remove unwanted frames from the beginning, achieving “forward trimming” of the video.

Basic usage: Input the video latent data to be trimmed into samples, and set trim_amount to the number of frames to trim. The node will trim the specified number of frames from the beginning of the video and output the remaining latent sequence. Typical scenarios: Used in video generation, video editing and other scenarios to remove unwanted leading frames, or to work with other nodes to achieve video segment splicing and processing.

Parameters

Input Parameters

Parameter	Type	Required	Default	Description
samples	LATENT	Yes	None	Input latent video data
trim_amount	INT	Yes	0	Number of frames to trim (from start)

Output Parameters

Parameter	Type	Description
samples	LATENT	Trimmed video latent data

Usage Example

Wan2.1 VACE Video Generation Workflow Example

Wan2.1 VACE Video Generation Workflow Example

Source Code

class TrimVideoLatent:
    @classmethod
    def INPUT_TYPES(s):
        return {"required": { "samples": ("LATENT",),
                              "trim_amount": ("INT", {"default": 0, "min": 0, "max": 99999}),
                             }}

    RETURN_TYPES = ("LATENT",)
    FUNCTION = "op"

    CATEGORY = "latent/video"

    EXPERIMENTAL = True

    def op(self, samples, trim_amount):
        samples_out = samples.copy()

        s1 = samples["samples"]
        samples_out["samples"] = s1[:, :, trim_amount:]
        return (samples_out,)