Pixie Queen from Magic the Gathering (Legends)
What would happen if I used a single low-resolution, not very detailed card illustration to train a LoRA?
It’s not perfect — I think I can do better, even with just one image. It’s definitely experimental, but I’m happy with it and wanted to share!
Usage Instructions
Use with the following settings:
Basic trigger words:
pqmtg, green hair, fairy wings, green lipsOptional for eyes:
no pupils, no scleraClothing trigger words:
choker, errings, jewelry, staffStrength: 0.8–0.9
Add skin color (pale skin for exemple) cause the Lora want her to have green skin.
Add Solo to the prompt
I hope you enjoy using this LoRA! Feel free to rate it, leave comments, and share your creations. Your feedback is always appreciated — thank you !
Description
How to generate correct image with Arthemy Comics Illustrious
Recommended settings for text-to-image:
Sampler: Euler A (Karras)
Sampling steps: 30
CFG scale: 8
Then, send your result to image-to-image with:
Sampling steps: 40
CFG scale: 8
Denoising strength: up to 0.75 (lower values for a closer match to the original image)
Extras: ADetailer (if you want her to have the original eyes, dont use Adetailer for eyes)