Cuphead Aesthetic (Illustrious)

I am admittedly kind of a noob at this, but if you enjoy my humble work, any donations are greatly appreciated: https://ko-fi.com/insult_to_ninjary

Your generosity helps me stay motivated to keep improving and to keep trying new things!

Trigger Words:

Main trigger: cuph34d

Helper prompts: parody, style parody, retro artstyle

The LoRA can potentially work with no triggers, but it may not be strong/consistent.

For a stronger effect, use the main trigger: cuph34d. You can also weight it numerically or with parentheses - I usually do (((cuph34d))).

If the (weighted) main trigger is still not enough by itself, next try the helper prompts, individually or in combination, along with the main trigger. They can be weighted as well - I usually do them all together, like (((cuph34d, parody, style parody, retro artstyle))).

The "retro artstyle" prompt is sometimes very helpful, but sometimes it creates too much of a retro anime effect instead of the intended Western cartoon style effect. If you're getting the retro anime effect, don't use this prompt (or maybe you can counteract it with negative prompting).

At a Glance:

Base model: Illustrated style (I use WAI-illustrious-SDXL)

LoRA weight: 1 (but slowly go lower or higher if necessary)

Suggested parameters: Euler a, Automatic, 30 steps, 832x1216/1216x832, CFG 5 (but feel free to experiment!)

Hires. fix: 1.5x, 30 steps, use a quality upscaler (like 4x-UltraSharpV2), denoising strength 0.5 - 0.3

Extra: use ADetailer to fix faces and/or hands if necessary

Description:

This LoRA produces art in a Cuphead-esque Western retro cartoon art style. The LoRA doesn't mimic the exact art style too closely in my opinion, but rather it kind of extrapolates the game's general aesthetic onto whatever you prompt for. I think the effect is pretty cool, but it might not be "on-the-nose" enough for you if you're looking for true mimicry. You may come across some other issues as well.

First, being a retro style, it's kind of blurry/fuzzy by nature, so while it does usually nail that intended effect well, this can also decrease the fidelity of fine details, especially in faces, hands, and small background elements. You may want to use ADetailer for faces and/or hands if you find them not defined enough.

In addition, the training dataset had a lot of screenshots, so the model has a tendency to want to zoom out, like a platformer level would look like. The helper prompts also seem to intensify this. To counteract it (if you want to), you can prompt for the specific composition you want (portrait, upper body, cowboy shot, etc.), and this usually works well enough.

Finally, you'll occasionally be visited by deformed little Cupheads (or other game characters). Just inpaint them out or re-gen (or lean into it if you're brave!). You might also sometimes get HUD elements, usually in the lower-left corner. Again, you can inpaint them out or just re-gen if you don't like that.

For more potentially relevant information, you can check out the original SD1.5 description. Below, there's a troubleshooting section in case you're having any problems.

Troubleshooting:

If you're having persistent problems with anatomy, fine detail, and/or other jank, just try your best with all the usual fixes - different prompting (including negative prompt), try a different base model, use a different VAE, try different clip skip values, adjust the LoRA weight, use adetailer, inpaint, etc.

Outro:

If you make anything awesome, please consider posting them to the gallery! That goes for all of my models! I really like seeing what kind of stuff you all are able to make!

If you have any questions, feel free to ask! I will try my best to help! Also, if you have any ideas for future models, please don't hesitate to share them with me!