RedNode Rig In
The front door your external engine's render walks back in through
- image
- image
RedNode Rig In is the world's most honest node: an image goes in, the same image comes out, and its entire job is to be a labelled doorway. You use it because the pack's Rig Out already handed your prompt and sampler numbers to an outside engine - NovelAI, or anything else - and now that engine's picture needs to re-enter the graph somewhere that reads clearly. This is that somewhere.
Think of the pair as a bridge with a sign on each end. Rig Out is the sign that says "NAI starts here" - it hands the numbers out. Rig In is the sign on the far side: "NAI ends here." Without it you'd drag the external image straight into whatever node needs it, and a week later you'd be staring at a graph with a floating wire and no idea where that tensor came from. With it, the story is on the canvas: the external render enters the chain here, and everything downstream - the Studio Detailer, the i2i wired canvas, Review, Save - treats it like any other image.
What it actually does
Two inputs, one output. image is the external engine's rendered picture, and it's required - this node is a pass-through, so there's no meaning to it without one. rig is just a label, defaulting to (active rig), whose whole purpose is to appear in the console receipt: when the node runs it prints something like [RedNode Rig In] 'my-nai-rig' handed back 832 x 1216, so you get a timestamped breadcrumb of whose render closed which loop. That receipt is the feature. The image output goes straight onward, untouched.
Why you'd bother
If you're only ever running one engine inside ComfyUI, skip this node - it does nothing a wire wouldn't. It earns its place the moment you run a hybrid pipeline: local Krea 2 for the base frame, an external engine for a pass, then back into ComfyUI for the detailer and upscale. The labelled junction keeps that graph legible, which is the pack's whole design thesis - stages in subgraphs and groups, so the canvas stays small however much is running. This node is that thesis in miniature: a couple of sockets that make the flow readable at a glance.
Install
Same one-liner as the rest of the pack:
git clone https://github.com/RedNodeAI/ComfyUI-RedNodeStudio.git ComfyUI/custom_nodes/ComfyUI-RedNodeStudio
Restart ComfyUI, or search RedNode Studio in ComfyUI Manager. No pip dependencies, and nothing here needs Krea 2 - Rig In works in any model's graph.
The one thing to know
Because the node passes pixels through untouched, there's no failure mode of its own to troubleshoot - if the picture looks wrong downstream, the problem is upstream in whatever produced it, not here. The only practical gotcha is remembering the label is cosmetic. It's for you and for the console log, not for routing; the node doesn't check that the rig name matches anything, so it can't silently route your image to the wrong place. That's the whole point of keeping it a dumb, labelled junction instead of a clever one.
Inputs (2)
| Name | Type | Default | Description |
|---|---|---|---|
| image | IMAGE | The external engine's rendered picture. | |
| rigopt | STRING | (active rig) | A label saying whose render this closes; shown in the console receipt. |
Outputs (1)
| Name | Type | Description |
|---|---|---|
| image | IMAGE | — |