ComfyUI-VideoHelperSuite

VHS_AudioToVHSAudio

Legacy method for external nodes that utilized VHS_AUDIO, VHS_AUDIO is deprecated as a format and should no longer be used

Video Helper Suite 🎥🅥🅗🅢/Audio · 1 input · 0 parameters · 1 output

VHS_BatchManager

Coordinates Video Helper Suite loaders and outputs in fixed-size meta-batches so very long frame sequences can be processed without materializing the whole sequence at once. It tracks open inputs, total frames, requeues, and the configured frames per batch.

Video Helper Suite 🎥🅥🅗🅢 · 0 inputs · 1 parameter · 1 output

VHS_DuplicateImages

Repeats the entire input image batch the requested number of times by concatenating copies along the batch dimension. It outputs the expanded image batch and its resulting image count.

VHS_DuplicateLatents

Repeats an entire latent batch the requested number of times, preserving every tensor field whose leading dimension matches the sample batch. It returns the expanded LATENT and its new count.

VHS_DuplicateMasks

Repeats the entire input mask batch the requested number of times by concatenating copies along the batch dimension. It outputs the expanded mask batch and its resulting mask count.

VHS_GetImageCount

Returns the number of images in an input image batch.

VHS_GetLatentCount

Returns the number of samples in an input latent batch.

VHS_GetMaskCount

Returns the number of masks in an input mask batch.

VHS_LoadAudio

Loads audio from a validated filesystem path or supported URL and returns ComfyUI AUDIO plus the loaded duration. Optional seek and duration controls select a segment of the source.

Video Helper Suite 🎥🅥🅗🅢/Audio · 0 inputs · 3 parameters · 2 outputs

VHS_LoadAudioUpload

Loads a selected audio file from ComfyUI’s input directory and returns the AUDIO plus its loaded duration. Start-time and duration controls can select a subsection.

Video Helper Suite 🎥🅥🅗🅢/Audio · 0 inputs · 3 parameters · 2 outputs

VHS_LoadImagePath

Loads an image from an arbitrary validated path through the pack’s FFmpeg image path and returns its RGB pixels plus an alpha-derived mask. Optional dimensions resize the image, and a supplied VAE can return latents instead of pixels.

VHS_LoadImages

Loads an ordered image sequence from a subdirectory of ComfyUI’s input folder and returns image, mask, and frame-count batches. It can cap, skip, or subsample files and participate in VHS meta-batching.

VHS_LoadImagesPath

Loads an ordered image sequence from an arbitrary validated directory and returns image, mask, and frame-count batches. It supports caps, initial skipping, every-nth selection, and VHS meta-batching.

VHS_LoadVideo

Loads an uploaded video from ComfyUI's input folder and exposes its frames, frame count, audio, and video information. It can resample, resize, cap, skip, or subsample frames and can emit latents when a VAE is supplied.

VHS_LoadVideoFFmpeg

Loads a video selected from ComfyUI’s input folder with FFmpeg and returns frames or latents, transparency as a mask, audio, and source/loaded video metadata. It supports frame-rate and size conversion, a frame cap, start time, formats, and meta-batching.

VHS_LoadVideoFFmpegPath

Loads a video from an arbitrary validated path or URL with FFmpeg and returns frames or latents, transparency as a mask, audio, and video metadata. It supports frame-rate and size conversion, a frame cap, start time, formats, and meta-batching.

VHS_LoadVideoPath

Loads a video from an arbitrary validated path or URL and returns frames or optional latents, frame count, audio, and source/loaded video information. It can resample frame rate, resize, cap, skip, subsample, and meta-batch the sequence.

VHS_MergeImages

Concatenates two image batches in A-then-B order and returns the merged batch and count. If dimensions differ, it scales one batch to match A, B, the smaller, or the larger resolution using the chosen scale and crop strategy.

VHS_MergeLatents

Concatenates two latent sample batches in A-then-B order and returns the merged LATENT and count. When latent dimensions differ, it rescales one side according to the selected template, scale method, and crop behavior.

VHS_MergeMasks

Concatenates two mask batches in A-then-B order and returns the merged masks and count. If their dimensions differ, it rescales one batch to the chosen A, B, smaller, or larger template.

VHS_PruneOutputs

Deletes selected intermediate and utility files from a VHS_FILENAMES output while retaining the intended final output. It refuses to remove files outside ComfyUI’s output or temporary directories.

Video Helper Suite 🎥🅥🅗🅢 · 1 input · 1 parameter · 0 outputs

VHS_SelectEveryNthImage

Skips an optional number of leading images, keeps every nth remaining image, and returns the reduced batch plus its count.

VHS_SelectEveryNthLatent

Skips an optional number of leading latent samples, keeps every nth sample across matching latent tensor fields, and returns the reduced LATENT plus its count.

VHS_SelectEveryNthMask

Skips an optional number of leading masks, keeps every nth remaining mask, and returns the reduced batch plus its count.

VHS_SelectFilename

Extracts one path from a VHS_FILENAMES output by integer index, with negative indexing available for selecting from the end.

Video Helper Suite 🎥🅥🅗🅢 · 1 input · 1 parameter · 1 output

VHS_SelectImages

Selects and reorders images using comma-separated indices and Python-style ranges, including negative indices and range steps. Optional checks control errors for missing indices or an empty selection.

VHS_SelectLatents

Selects and reorders latent samples using comma-separated indices and Python-style ranges, applying the same selection to matching tensor fields. Optional checks control errors for missing indices or an empty result.

VHS_SelectLatest

Experimental virtual node that resolves the most recently modified file matching a folder/prefix and postfix, then supplies that path to connected widgets. Its browser extension performs the substitution; the server method is intentionally unreachable.

Video Helper Suite 🎥🅥🅗🅢 · 0 inputs · 2 parameters · 1 output

VHS_SelectMasks

Selects and reorders masks using comma-separated indices and Python-style ranges, including negative indices and range steps. Optional checks control errors for missing indices or an empty selection.

VHS_SplitImages

Splits an image batch at the requested index and returns the leading and trailing batches together with each count.

VHS_SplitLatents

Splits a latent batch at the requested index, dividing every matching tensor field consistently, and returns both LATENT groups with their counts.

VHS_SplitMasks

Splits a mask batch at the requested index and returns the leading and trailing batches together with each count.

VHS_Unbatch

Combines a list of batched values into one output: tensors are concatenated along the batch dimension, latent or audio dictionaries have their sample or waveform tensors concatenated, and other values are added together. This is separate from VHS meta-batching.

Video Helper Suite 🎥🅥🅗🅢 · 1 input · 0 parameters

VHS_VAEDecodeBatched

Decodes a latent sequence with a VAE in configurable sub-batches and concatenates the decoded images, reducing peak memory for long sequences.

VHS_VAEEncodeBatched

Crops and encodes an image sequence with a VAE in configurable sub-batches, then concatenates the encoded samples into one LATENT output.

VHS_VHSAudioToAudio

Legacy method for external nodes that utilized VHS_AUDIO, VHS_AUDIO is deprecated as a format and should no longer be used

Video Helper Suite 🎥🅥🅗🅢/Audio · 1 input · 0 parameters · 1 output

VHS_VideoCombine

Combines an image sequence, or a latent sequence when a VAE is supplied, into a selected animation or video output. It controls frame rate, looping, ping-pong playback and saving, can add optional audio, and returns the produced filenames.

Video Helper Suite 🎥🅥🅗🅢 · 1 output

VHS_VideoInfo

Unpacks a VHS video-information value into both source and loaded FPS, frame count, duration, width, and height outputs.

Video Helper Suite 🎥🅥🅗🅢 · 1 input · 0 parameters · 10 outputs

VHS_VideoInfoLoaded

Unpacks the post-processing or loaded FPS, frame count, duration, width, and height from a VHS video-information value.

Video Helper Suite 🎥🅥🅗🅢 · 1 input · 0 parameters · 5 outputs

VHS_VideoInfoSource

Unpacks the original source FPS, frame count, duration, width, and height from a VHS video-information value.

Video Helper Suite 🎥🅥🅗🅢 · 1 input · 0 parameters · 5 outputs