分辨率大师简化版 (Resolution Master Simplify)

Worth clearing up first, because the name will confuse you if you've used the real thing: Resolution Master is a genuinely popular custom node by Azornes (Comfyui-Resolution-Master) with an interactive canvas, aspect-ratio presets, and model-specific optimization for SDXL, Flux, and friends. This node, from a completely different pack, is a "simplified" homage to that idea - and it's simplified all the way down to just two integers. No canvas, no presets, no per-model logic. If you were hoping for the full thing, install Azornes's node instead; this one is for something much smaller.

What it actually is

Two required INT fields - width and height - each clamped 64–8192 in steps of 64, each defaulting to 1024. Two INT outputs, same names, passed straight through. That's the entire node. No math, no aspect-ratio math, no megapixel targeting.

Why you'd still reach for it

Honestly: mostly labeling and tidiness. A raw ComfyUI graph full of unlabeled Primitive nodes feeding an Empty Latent Image is hard to scan at a glance - which int is width, which is height, which one did you mean to change? This node gives you one clearly-named box for both, with sane bounds already baked in (the 64-step increment matches the alignment most diffusion models expect, so you can't accidentally set an odd resolution that breaks tiling). If your workflow just needs "a width and a height, please, in one place," this does that and nothing more.

The inputs and outputs that matter

• width (INT, 64–8192, step 64, default 1024)
• height (INT, 64–8192, step 64, default 1024)

Outputs mirror the inputs exactly - width and height - for wiring into your Empty Latent Image node (or anywhere else that wants a resolution pair).

How to install it

ComfyUI Manager: search "ComfyUI-Danbooru-Gallery" (this node ships as part of that pack, not standalone - its category label is the only thing separating it from the pack's other tools in the node menu). Manual:

cd ComfyUI/custom_nodes
git clone https://github.com/Aaalice233/ComfyUI-Danbooru-Gallery.git
cd ComfyUI-Danbooru-Gallery
pip install -r requirements.txt

Restart ComfyUI. No model downloads, no heavy dependencies for this one specifically - it's pure arithmetic-free UI.

Common issues & troubleshooting

You wanted aspect-ratio presets or megapixel targeting and this doesn't have them. That's not a bug, it's the whole point of the "Simplify" in the name - this is deliberately the bare-minimum version. For the real feature set (interactive canvas, SDXL/Flux-aware presets), you want Azornes's Comfyui-Resolution-Master pack, which is a separate, unrelated project.

Values snap to unexpected numbers. The step is 64 in both directions - if you type a value that isn't a multiple of 64, expect it to round. That's deliberate: most diffusion architectures want dimensions aligned to a power-of-two-friendly step, and this node won't let you set one that isn't.

Node doesn't appear in the menu. Standard pack-import issue, not specific to this node: confirm the repo is at ComfyUI/custom_nodes/ComfyUI-Danbooru-Gallery, confirm pip install -r requirements.txt ran clean, and check the ComfyUI console on startup for an import error.