Nuke Constant

Sometimes you need a rectangle of pure color, and you need it now, with exactly the right RGBA values. NukeConstant is the ComfyUI version of Nuke's Constant node: a generator with no image input that outputs a solid-color image at whatever size and color you ask for. It's the most boring node in the pack, and that's its superpower - a boring, predictable color reference is exactly what a lot of pipelines need.

You'll find it useful more often than you'd think. Need a background to composite a foreground over without loading an image file? This. Need to test whether your save pipeline preserves a specific color value exactly? This. Need a neutral mid-gray frame to check a grade? This. It's also the building block for mockups - put a logo render over a Constant background instead of hunting for a placeholder image.

How it works

It builds a tensor of the given width and height filled with the red, green, blue, and alpha values (each 0 to 1). Nothing more. The one decision that matters: output_alpha. With it off (default), you get a plain RGB fill. Turn it on and the output carries an RGBA channel with the alpha you set - which is what you want if this color is going to be composited with the alpha doing real work, like making a translucent color overlay.

The optional batch_size (1 to 64) generates multiple identical frames, handy when you're feeding a video workflow that expects a batch.

The inputs that matter

• width / height - output size, up to 8192.
• red / green / blue - the fill color, 0 to 1 each. For a mid-gray test card: 0.5, 0.5, 0.5.
• alpha - the alpha value, 0 to 1 (default 1).
• output_alpha - whether to include an RGBA channel or just RGB.
• batch_size - number of identical frames.

Output is a single IMAGE.

Installing it

It's part of the Nuke Nodes pack:

cd ComfyUI/custom_nodes
git clone https://github.com/sumitchatterjee13/nuke-nodes-comfyui.git nuke-nodes
cd nuke-nodes
pip install -r requirements.txt

Or search "Nuke Nodes" in ComfyUI Manager and restart. It's under the Nuke category, pure PyTorch tensor generation - no dependencies beyond the pack itself.

Common issues

• The color looks different on screen than the values I set - your display is doing gamma/color management. This node outputs exactly the numeric values you set; how they look depends on your monitor. That's a feature: it's a reference, not a rendering.
• It's transparent when I expected it to be opaque - check output_alpha and alpha. If output_alpha is on and alpha is low, you've built a translucent fill on purpose.
• "Why do I need a solid color node?" - you'll know the first time a workflow needs a background and you don't want to save and load an image file. Also: gray test cards are the honest way to verify a grade.

It won't impress anyone, but when a workflow needs a color, this is the cleanest way to hand it one.