DP Load Image Effects Small

The trimmed-down sibling of DP Load Image Effects - same idea (load an image, get it back in two differently-processed versions, plus whatever prompt metadata was embedded in the file), but two effect slots instead of four. If you only need "give me a before and an after" rather than four parallel variants, this is the lighter option and it keeps your graph tidier.

How it works

Pick image from ComfyUI's input folder, choose an effect for effect_A and a (possibly different) one for effect_B from the shared 24-option effect menu - grayscale, enhance, flips, rotations, posterize, sharpen, contrast, equalize, and a further set the node's brief doesn't spell out individually. effect_strength scales both. resize_image (on by default) forces both outputs to a common width/height.

Same as its bigger sibling, it also reads back any generation metadata baked into the source file - dp_prompt and dp_negative_or_other come out as strings if the image has that data embedded, which is handy when you're working from a previously-generated reference and want its original prompt without hunting through your history.

A common use for the two-slot pattern specifically: effect_A set to original and effect_B set to whatever transformation you're testing, so you get a clean before/after pair out of one node instead of duplicating a loader.

Inputs and outputs

Required:

• image - enum dropdown from ComfyUI's input folder.
• effect_strength - 0 to 1, default 1.
• effect_A / effect_B - each independently pick from 24 effects, default original.
• resize_image - boolean, default true.
• width / height - 64 to 2048, default 1024 each.

Outputs:

• image_A / image_B - the two processed variants.
• filename - source filename as a string.
• dp_prompt - extracted prompt, if present.
• dp_negative_or_other - extracted negative/other text, if present.

How to install it

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

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

Restart, find it under DP/image. No models or dependencies beyond a standard ComfyUI install.

Common issues & troubleshooting

dp_prompt/dp_negative_or_other are empty strings. Only populated when the source image has embedded generation metadata - a plain photo or a JPEG that stripped metadata during export will have nothing to extract. Not a bug, just nothing there to read.

One output slot looks like the unmodified source. Check it's not left on effect_A/effect_B's default, original - a genuine no-op, and easy to forget if you only set one of the two slots deliberately.

Resize distorts the image. resize_image defaults to on and does a hard resize to width×height, not an aspect-preserving fit - if your source's aspect ratio differs from the target, expect stretching. Turn it off to keep native dimensions.

Need more than two variants. Use the full DP Load Image Effects node instead - it's the same node with two extra effect slots (A through D rather than A/B).