Architecture

Long-audio pipeline

The repository keeps the long-audio path explicit:

1. LTXAudioDuration and LTXLongAudioSegmentInfo plan segments from the real source audio length.
2. LTXBuildChunkedStillVideo splits the song into chunk-sized windows.
3. Each chunk chooses one deterministic frame from the selected folder.
4. Internal dummy segment rendering produces a still-video slice for each chunk.
5. Image batches and audio segments are concatenated back into one previewable result.
6. LTXVideoCombine muxes the final MP4 preview payload.

Segment frame counts are quantized in blocks of 8 frames to stay compatible with LTX-style downstream assumptions.

Node families

Inputs and staging

• LTXLoadAudioUpload
• LTXLoadImageUpload
• LTXLoadImages
• LTXBatchUploadedFrames
• LTXRepeatImageBatch

Chunk planning and media assembly

• LTXAudioDuration
• LTXLongAudioSegmentInfo
• LTXRandomImageIndex
• LTXAudioSlice
• LTXBuildChunkedStillVideo
• LTXDummyRenderSegment
• LTXAppendImageBatch
• LTXAppendAudio
• LTXEnsureImageBatch
• LTXEnsureAudio
• LTXAudioConcatenate
• LTXVideoCombine

Loop control and workflow helpers

• LTXWhileLoopStart, LTXWhileLoopEnd
• LTXForLoopStart, LTXForLoopEnd
• LTXIfElse
• LTXCompare
• LTXSimpleMath
• LTXSimpleCalculator
• LTXIntConstant
• LTXIndexAnything
• LTXBatchAnything
• LTXSeedList
• LTXShowAnything

LTX utility replacements

• LTXVAELoader
• LTXImageResize
• LTXChunkFeedForward
• LTXSamplingPreviewOverride
• LTXNormalizedAttentionGuidance

Repository layout

docs/                         VitePress docs and shared SVG identity
samples/input/                lightweight sample frames and fallback audio
samples/workflows/            smoke workflow JSON
scripts/check_workflow_layout.py
scripts/run_comfyui_api_smoke.py
tests/                        import and workflow regression tests
nodes.py                      custom node implementations and mappings