A ComfyUI custom node designed for LTX 2.3 MSR (Multiple-Subject-Reference) LoRA workflows.
Python
379
17 commits
updated Jul 14, 2026
Value not in list error for the frame_count property on newer ComfyUI frontends.frame_count dropdown values are now serialized using ComfyUI's supported string-based COMBO format and converted back to integers before processing.frame_count as an integer remain supported, while newer workflows that submit values such as "25" as strings now work correctly.A ComfyUI custom node for creating fixed-frame MP4 reference videos from multiple subject images and a background image, designed for LTX 2.3 MSR (Multiple-Subject-Reference) LoRA workflows.
1-4) are resized proportionally and always kept complete; unused canvas space is filled with white.49, 57, and 65 frame options.Licon MSRLicon/MSR1: optional image2: optional image3: optional image4: optional imagebackground: required imagewidth: output video widthheight: output video heightframe_count: 17, 25, 33, 41, 49, 57, or 65Images are processed in this fixed order:
1 -> 2 -> 3 -> 4 -> background
Disconnected 1 to 4 inputs are skipped. background is always required and always placed last.
Subject images are proportionally fitted inside the configured dimensions without cropping and centered on a white canvas. The background proportionally covers the full canvas and is center-cropped if its aspect ratio differs. Frames are assigned in LTX-aligned temporal groups, and the node outputs an IMAGE frame sequence for downstream video processing.
Clone this repository into the ComfyUI custom_nodes directory:
cd ComfyUI/custom_nodes
git clone https://github.com/liconstudio/ComfyUI-Licon-MSR
Install requirements if needed:
pip install -r ComfyUI-Licon-MSR/requirements.txt
Restart ComfyUI after installation.
17 commits
Python
100.0%
A ComfyUI custom node designed for LTX 2.3 MSR (Multiple-Subject-Reference) LoRA workflows.
Python
379
17 commits
updated Jul 14, 2026
Value not in list error for the frame_count property on newer ComfyUI frontends.frame_count dropdown values are now serialized using ComfyUI's supported string-based COMBO format and converted back to integers before processing.frame_count as an integer remain supported, while newer workflows that submit values such as "25" as strings now work correctly.A ComfyUI custom node for creating fixed-frame MP4 reference videos from multiple subject images and a background image, designed for LTX 2.3 MSR (Multiple-Subject-Reference) LoRA workflows.
1-4) are resized proportionally and always kept complete; unused canvas space is filled with white.49, 57, and 65 frame options.Licon MSRLicon/MSR1: optional image2: optional image3: optional image4: optional imagebackground: required imagewidth: output video widthheight: output video heightframe_count: 17, 25, 33, 41, 49, 57, or 65Images are processed in this fixed order:
1 -> 2 -> 3 -> 4 -> background
Disconnected 1 to 4 inputs are skipped. background is always required and always placed last.
Subject images are proportionally fitted inside the configured dimensions without cropping and centered on a white canvas. The background proportionally covers the full canvas and is center-cropped if its aspect ratio differs. Frames are assigned in LTX-aligned temporal groups, and the node outputs an IMAGE frame sequence for downstream video processing.
Clone this repository into the ComfyUI custom_nodes directory:
cd ComfyUI/custom_nodes
git clone https://github.com/liconstudio/ComfyUI-Licon-MSR
Install requirements if needed:
pip install -r ComfyUI-Licon-MSR/requirements.txt
Restart ComfyUI after installation.
17 commits
Python
100.0%