ComfyUI-LTXLongAudio

Native LTX workflow surface

Bundle loop control, chunk planning, media assembly, and preview output without relying on extra node packs.

App mode-ready smoke path

Ship one focused workflow with four exposed controls, sample folder defaults, and a deterministic frame-per-chunk preview path.

Publish with QA

Validate layout contracts, App mode metadata, and prompt API execution with repository-native scripts before you share a graph.

Workflow Preview

Repository Header

ComfyUI-LTXLongAudio

The repository header uses this workflow overview screenshot so the bundled long-audio graph is visible at a glance.

App Screen

The App mode surface stays intentionally small: choose a frame folder, upload source audio, set chunk length, and keep deterministic seed control.

What ships

Native input nodes for audio upload, image upload, batched frames, and folder selection.

Long-audio helpers for segment timing, deterministic frame selection, sliced audio, and final audio concatenation.

Native loop-control nodes and utility replacements used by the bundled LTX workflow.

A smoke workflow under samples/workflows/ plus lightweight sample image folders in samples/input/.

What to expect from the sample

The bundled workflow keeps the UI surface intentionally small:

App mode input	Purpose
`Frames Folder`	Picks the folder that feeds per-chunk still frames.
`Source Audio Upload`	Lets you upload the long-form song directly in ComfyUI.
`Segment Seconds`	Controls chunk length, defaulting to 20 seconds.
`Random Seed`	Keeps frame selection deterministic across reruns.

App mode input

Purpose

Frames Folder

Picks the folder that feeds per-chunk still frames.

Source Audio Upload

Lets you upload the long-form song directly in ComfyUI.

Segment Seconds

Controls chunk length, defaulting to 20 seconds.

Random Seed

Keeps frame selection deterministic across reruns.

The workflow currently analyzes to 3 groups and 9 nodes, with one final output node selected for App mode preview.

Verification entry points

uv run pytest

uv run python scripts/check_workflow_layout.py samples/workflows/LTXLongAudio_CustomNodes_SmokeTest.json --require-all-nodes-in-groups --require-app-mode

uv run python scripts/run_comfyui_api_smoke.py --workflow samples/workflows/LTXLongAudio_CustomNodes_SmokeTest.json --comfy-root /path/to/ComfyUI

TIP

run_comfyui_api_smoke.py keeps ltx-demo-tone.wav as the tracked default. If you also keep a longer local HOWL AT THE HAIRPIN2.wav, the script prefers it and stages it under the tracked widget filename.

ComfyUI-LTXLongAudioLong-audio chunks, native loops, and still-video previews

Native LTX workflow surface

App mode-ready smoke path

Publish with QA

Workflow Preview ​

Repository Header ​

App Screen ​

What ships ​