Save 4 hours! We auto-setup your workflow! Free!

Drop your workflow.json — we handle every dependency, custom node, and model. Just open the link and run.

Auto-Setup Workflow Json (Free) Now!
ComfyUI > Nodes > toobusy > toobusy LTX2.3 Empty AV Latents

ComfyUI Node: toobusy LTX2.3 Empty AV Latents

Class Name

LTX23EmptyAVLatents

Category
toobusy/Make
Author
nicekriss (Account age: 1318days)
Extension
toobusy
Latest Updated
2026-10-25
Github Stars
0.02K

How to Install toobusy

Install this extension via the ComfyUI Manager by searching for toobusy
  • 1. Click the Manager button in the main menu
  • 2. Select Custom Nodes Manager button
  • 3. Enter toobusy in the search bar
After installation, click the Restart button to restart ComfyUI. Then, manually refresh your browser to clear the cache and access the updated list of nodes.

Visit ComfyUI Online for ready-to-use ComfyUI environment

  • Free trial available
  • 16GB VRAM to 80GB VRAM GPU machines
  • 400+ preloaded models/nodes
  • Freedom to upload custom models/nodes
  • 200+ ready-to-run workflows
  • 100% private workspace with up to 200GB storage
  • Dedicated Support

Run ComfyUI Online

toobusy LTX2.3 Empty AV Latents Description

Creates aligned LTX 2.3 video and audio latents with configurable timing, size, and optional custom audio.

toobusy LTX2.3 Empty AV Latents:

LTX23EmptyAVLatents prepares the starting latent data for an LTX 2.3 audio-and-video workflow. A latent is an intermediate representation used by generation nodes rather than a finished video or playable sound. The node combines the setup normally handled by separate LTXV video and audio nodes, keeping the two parts aligned to the same video duration and batch size. It can also prepare supplied audio as the audio latent instead of starting with an empty one. This makes it useful when you want a single place to establish the basic shape and timing of an audio-video generation.

toobusy LTX2.3 Empty AV Latents Input Parameters:

audio_vae

A required VAE connection used to create the empty audio latent or encode supplied audio. Connect the audio VAE appropriate for your LTXV workflow. This is a connection, not a numeric setting, so it has no minimum, maximum, or default value.

ratio_preset

A required aspect-ratio selection that determines the intended relationship between video width and height. The node uses it with megapixels to calculate a resolution, then rounds the dimensions according to divisible_by; the final ratio can therefore differ slightly from the preset. The default is 16:9. The complete list of available options comes from RATIO_PRESETS and is not shown in the provided source, so select from the options displayed in your node.

megapixels

Sets the approximate video image area in millions of pixels. Together with ratio_preset, it determines the calculated width and height before rounding to the requested divisibility. A higher value generally produces larger dimensions, while a lower value produces smaller ones; the final pixel count may differ because of rounding. The default is 1.0, the minimum is 0.01, the maximum is 16.0, and the adjustment step is 0.01.

divisible_by

Sets the multiple used when rounding the calculated video width and height. This lets you request dimensions aligned to a particular size requirement, although rounding can change the exact resolution implied by megapixels. The default is 32, the minimum is 1, and the maximum is 256.

length

Sets the number of frames requested for the empty video latent. The node also passes this value to the empty audio latent setup as its frame count, keeping that setup tied to the video duration. The default is 97, the minimum is 1, and the maximum is 16384.

frame_rate

Sets the frames-per-second value passed to the empty audio latent setup and returned for use elsewhere in your workflow. In combination with length, it describes the intended video timing: for example, 96 frames at 24 frames per second represents about four seconds. The default is 24, the minimum is 1, and the maximum is 120.

batch_size

Sets how many samples the node requests from each empty-latent setup node. Increase it when your workflow needs multiple items in a batch, while keeping in mind that larger batches can require more resources. The default is 1, the minimum is 1, and the maximum is 4096. In custom-audio mode, this value is not passed to the audio encoder, so do not assume it changes the batch size of encoded audio.

use_custom_audio

Chooses how the audio latent is prepared. When False, the node uses the empty audio latent it created. When True, it replaces that latent with one encoded from the optional audio connection and applies a noise mask. The default is False; the options are False and True. If you choose True, an audio connection is required.

audio

An optional AUDIO connection containing the sound you want to encode when use_custom_audio is True. It is not used when the switch is False. There is no numeric range or default value; leave it unconnected for empty-audio setup, but connect it before enabling custom audio.

toobusy LTX2.3 Empty AV Latents Output Parameters:

video_latent

A LATENT containing the empty video starting point created at the calculated width and height, requested frame count, and batch size. Connect it to a compatible downstream LTXV video workflow; it is not a finished video.

audio_latent

A LATENT containing either the empty audio starting point or, when custom audio is enabled, the encoded supplied audio with a noise mask applied. Its meaning depends on the selected audio mode, and it is not directly playable sound.

length

An INT repeating the requested frame count. Use it to pass the same duration setting to later nodes without entering it again.

frame_rate_int

An INT version of the requested frame rate. This is the value the node passes to the empty audio latent setup and can be connected to downstream settings that expect an integer frame rate.

frame_rate_float

A FLOAT version of the requested frame rate. It carries the same frames-per-second setting as frame_rate_int in a type suitable for downstream inputs that expect a floating-point number.

width

An INT giving the calculated video width in pixels after rounding according to divisible_by. Use this actual value, rather than an estimate from the megapixel setting, when matching dimensions in later nodes.

height

An INT giving the calculated video height in pixels after rounding according to divisible_by. Along with width, it tells you the dimensions used to create the video latent.

toobusy LTX2.3 Empty AV Latents Usage Tips:

  • Start with the default 16:9 ratio, 1.0 megapixels, and 32 divisibility if you do not yet have a specific resolution requirement. Check the width and height outputs to see the dimensions actually chosen.
  • Set length and frame_rate together to plan timing. For example, 96 frames at 24 frames per second describes about four seconds; pass the frame-rate output to later nodes that need the same timing.
  • Keep use_custom_audio set to False for an empty audio starting point. To use a sound clip, connect it to audio first, then enable use_custom_audio.
  • Begin with batch_size set to 1 when checking a workflow. Increase it only when you need multiple samples and your available resources can support them.
  • Choose divisible_by to meet the dimension requirements of your downstream workflow. Because dimensions are rounded, use the node’s width and height outputs when configuring other size-dependent nodes.

toobusy LTX2.3 Empty AV Latents Common Errors and Solutions:

Connect an AUDIO input or turn off use_custom_audio.

  • Explanation: Custom-audio mode is enabled, but nothing is connected to the audio input. The node stops rather than creating an encoded audio latent without a source.
  • Solution: Connect an AUDIO output to audio, or turn use_custom_audio off to use the empty audio latent.

Required audio_vae input is missing

  • Explanation: The node requires an audio VAE connection for both empty-audio setup and custom-audio encoding. Your workflow may fail input validation if it is not connected.
  • Solution: Connect a compatible audio VAE before running the workflow.

LTXV node is unavailable

  • Explanation: This node calls EmptyLTXVLatentVideo and LTXVEmptyLatentAudio internally. Custom-audio mode also calls LTXVAudioVAEEncode, SolidMask, and SetLatentNoiseMask. If a required node is unavailable, the workflow cannot complete; the exact error text depends on the installed environment.
  • Solution: Install or enable the LTX 2.3 (LTXV*) node set required by this workflow, then restart or reload your node environment.

Input value is outside its allowed range

  • Explanation: Your workflow may reject a value outside the node’s declared limits, such as length below 1 or megapixels above 16.0. The validation message depends on your environment.
  • Solution: Reset the affected control to its default or enter a value within the minimum and maximum listed above.

A downstream node rejects the dimensions

  • Explanation: The generated width and height are rounded according to divisible_by and may not match dimensions entered separately elsewhere in your workflow. The exact downstream error depends on that node.
  • Solution: Connect this node’s width and height outputs where possible, and check that your chosen divisible_by meets the downstream node’s requirements.

toobusy LTX2.3 Empty AV Latents Related Nodes

Go back to the extension to check out more related nodes.
toobusy
RunComfy
Copyright 2025 RunComfy. All Rights Reserved.

RunComfy is the premier ComfyUI platform, offering ComfyUI online environment and services, along with ComfyUI workflows featuring stunning visuals. RunComfy also provides AI Models, enabling artists to harness the latest AI tools to create incredible art.

toobusy LTX2.3 Empty AV Latents