DP Image Effect Processor

Normally, running four different post-processing effects on four different images means four separate effect nodes cluttering your graph. This one collapses that into a single node with four independent slots - plug in up to four images, pick an effect per slot, get four processed images back out. It's less "one clever effect" and more "workflow tidiness": a batch finishing pass that doesn't eat half your canvas in wires.

How it works

Each of the four slots (01 through 04) is its own mini pipeline: an optional image input, an effect choice from a shared 24-option list, and a shared effect_strength that applies across all four. The effect list covers the usual post-processing toolkit - grayscale, enhance, flips (flip_h/flip_v), 90/180/270-degree rotations, posterize, sharpen, contrast, equalize, and a further dozen the schema summarizes as "+12 more" (the full menu isn't enumerated in the node's brief, so treat that list as illustrative rather than exhaustive - check the dropdown in ComfyUI for the complete set). The default for every slot is original, which is a deliberate no-op passthrough - nothing changes until you pick something else.

There's an optional resize pass too: flip resize_image on and set width/height and all four outputs get resized to match, which is handy if your four source images came in at different resolutions and you want them consistent before saving or batching further downstream.

Inputs and outputs

Required:

• effect_strength - 0 to 1, default 1, applies globally across whichever slots have an effect selected.
• image_01_effect through image_04_effect - each an independent pick from the 24-option effect list, default original.
• resize_image - boolean, default off.
• width / height - 64 to 2048, default 1024 each, used only if resize_image is on.

Optional:

• image_input_01 through image_input_04 - the four images, each independent. You don't have to fill all four; leave a slot's image input disconnected and that output just won't have anything meaningful to process.

Outputs: image_output_01 through image_output_04, one per slot, each carrying whatever effect you picked for it (or the untouched original if you left it at original).

How to install it

Search ComfyUI-Desert-Pixel-Nodes in ComfyUI Manager and install, or:

cd ComfyUI/custom_nodes
git clone https://github.com/DesertPixelAi/ComfyUI-Desert-Pixel-Nodes

Restart. It lives under the DP category in the node browser with the rest of the pack. No models, no extra dependencies beyond a standard ComfyUI/PyTorch setup per the README.

Common issues & troubleshooting

A slot "isn't doing anything." Check that its effect dropdown isn't still on original - that's the default for all four slots, and it's a genuine passthrough, not a bug. Easy to forget you left one slot untouched when you're focused on the other three.

One output is blank or errors. If you didn't connect anything to that slot's image_input, there's nothing to process - the four slots are independent, so leaving image_input_03 disconnected while using 01, 02, and 04 is fine, but don't expect image_output_03 to produce anything from thin air.

Resize looks like it's stretching instead of fitting. resize_image forces every connected slot to the same width/height, which is a hard resize, not a crop-to-fit or letterbox - if your four inputs have different aspect ratios, they'll each get squashed or stretched to match the target dimensions rather than padded. If aspect ratio matters, resize your sources to a consistent ratio before this node rather than relying on it to do aspect-aware fitting.

Effect not showing up as expected. With 24 choices and only 12 named in the node's summary, some effects will be unfamiliar by name alone - if a pick looks like a no-op, it may just be a subtle effect (equalize and enhance in particular tend to be gentler than posterize or sharpen); push effect_strength up to 1 if you dialed it down and want to confirm the effect is actually wired correctly before troubleshooting further.