Overview

This LoRA aims to replicate the style of Magic the Gathering legend Magali Villeneuve. Subjectively, I find that it replicates the rendition of light on skin and on clothing faithfully i.e. it is rendered with thick brushstrokes with a lot of texture. Eyes and faces are rendered in a more realistic manner, i.e. closer to human proportions as opposed to big anime eyes, but I wouldn't say that the LoRa completely captures the way Villeneuve does faces. It's still a versatile LoRA for those looking to generate illustrations in the style commonly seen in MtG card art.

Comparison

Usage, Settings and Workflow

All images in the image carousel have ComfyUI workflows attached.
TLDR:

• Sampler: res_multistep

• Scheduler: Karras

• Lora Strength: 0.9-1.0

• CFG: 6

• Steps: 20

Note: A Hi-res fix or a face detailer workflow significantly improves the quality of hands and faces. All the samples in the image carousel were hi-res fixed 1.5x from1000x1600 to 1504x2400 at a denoise of 0.60.