VELVET VICE — AUTONOMOUS VISION
LTX 2.3 Video Workflow
AUTONOMOUS VISION is a modular ComfyUI workflow for LTX 2.3 video generation.
The integrated Vision Director analyzes the source image and creates an optimized video prompt based on a short user instruction. Manual prompting is also fully supported.
Main Features
Autonomous image analysis and prompt generation
Manual and assisted generation modes
Structured three-pass rendering
Power LoRA AV system
Independent Quick LoRA Search
Adaptive video preview
Integrated watermark controls
Clean Velvet Vice interface
Multilingual installation guides
The workflow is designed to simplify advanced LTX video generation while still giving experienced users direct control over prompts, LoRAs and render settings.
Models and required third-party custom nodes are not included. Please follow the included installation guide and model checklist.
v1.0 Player Hotfix
Fixed audio continuing in the background after pausing the adaptive video preview
Fixed duplicated audio when muting or unmuting the player
Prevented the legacy preview player from running alongside the Velvet Vice adaptive player
Updated the included
ComfyUI-Velvet-Vice-LTXcustom-node frontend
No new dependencies or additional custom nodes are required.
Important: Please replace your existing ComfyUI-Velvet-Vice-LTX folder with the updated folder included in this release. Restart ComfyUI and refresh the browser with Ctrl + F5.
Creator Name Update
My previous workflows were published under the name ILLUMINATE AI.
VELVET VICE is the same creator under a new name and visual identity. This is a rebranding, not a change of ownership or developer.
This is the first official release under the new VELVET VICE identity.
Feedback and detailed bug reports are welcome.
Created by VELVET VICE
Previously known as ILLUMINATE AI
Description
Version 1.0
First official release under the new VELVET VICE creator identity.
Previously published as ILLUMINATE AI — same creator, new name and visual identity.
Included in this release
Autonomous Vision Director
Manual and assisted prompt modes
Three-pass LTX 2.3 video generation
Power LoRA AV system
Independent Quick LoRA Search
Adaptive video preview
Integrated watermark controls
Fully English workflow interface
Renamed Velvet Vice custom-node package
Updated installation guides in English, German and Russian
Important
Remove older installations of:
ComfyUI-ILLUMINATE-AI-LTX
and
ComfyUI-Velvet-Vice-LTX
before installing this release. Do not merge the folders.
FAQ
Comments (4)
3
Errors detectedResolve before running the workflow
17 nodes — 3 errors
3Missing Node Packs
Install missing packs to use this workflow.
To install missing nodes, first run pip install -U --pre comfyui-manager in your Python environment to install Node Manager, then restart ComfyUI with the --enable-manager flag.
comfyui-dasiwa-nodes9
DaSiWa_NodeStatusSwitch
DaSiWa_NodeStatusSwitch
DaSiWa_NodeStatusSwitch
DaSiWa_NodeStatusSwitch
DaSiWa_NodeStatusSwitch
DaSiWa_NodeStatusSwitch
DaSiWa_NodeStatusSwitch
DaSiWa_NodeStatusSwitch
DaSiWa_ResolutionScaleCalculator
ComfyUI-Velvet-Vice-LTX
Unknown pack7
DaSiWa_NodeStatusSwitch
DaSiWa_NodeStatusSwitch
DaSiWa_NodeStatusSwitch
DaSiWa_NodeStatusSwitch
DaSiWa_NodeStatusSwitch
DaSiWa_NodeStatusSwitch
DaSiWa_NodeStatusSwitch
The workflow itself is not broken. ComfyUI is reporting missing required custom nodes.
Please install:
comfyui-dasiwa-nodes
This provides DaSiWa_NodeStatusSwitch and DaSiWa_ResolutionScaleCalculator.
ComfyUI-Velvet-Vice-LTX
This custom-node folder is included in the downloaded workflow bundle and must be copied into ComfyUI/custom_nodes/.
After installing both, completely restart ComfyUI and refresh the browser with Ctrl + F5.
Please do not only load the JSON file—the included custom nodes and installation instructions are required for the workflow to work.
Looks cool, I think you guys put some serious work into it. Seems to have a lot of other issues with Comfy Versions and the ltx director node versions. Who knows, maybe ill get it running today or in a few weeks...
@CeruleanDiffuser Thanks! Yeah, a lot of work went into this one. 😄
The workflow itself is running stable on the setup it was built and tested with, but ComfyUI and especially some LTX Director/custom node versions can cause compatibility issues if the installed versions differ.
Also, the latest ComfyUI update currently seems to be causing some additional compatibility problems with certain custom nodes, so that may be part of what you're running into as well.
If you want, send me the exact missing nodes or the error message you're getting. I can take a look and tell you which dependency or version is causing it. Hopefully we can turn those "few weeks" into a few minutes. 😄