HjMiXIllustrious

Hi everyone!

This is my first custom mix. It includes the well‑known WAI‑illustrious‑SDXL checkpoint, as well as the WAI‑ANIMA checkpoint and a bit of WAI‑SHUFFLE‑NOOB.

I’m sharing the result with you! The portrait generation turned out amazing — but that’s not all this mix has to offer.
The checkpoint already includes curated LoRAs optimized for facial detail, lighting, and overall aesthetic consistency. As the sample shows, it excels at ethereal, high‑end character art with rich volumetric lighting, intricate accessories, and smooth anatomy.

Feel free to experiment, share your results, and credit the source — I’d be happy to get any feedback!

Compatibility

• Fully works in ComfyUI

• Fully works in Automatic1111 / Forge / reForge

Recommended Settings

• Base Generation:

  • Sampler: euler_ancestral

  • Scheduler: beta

  • CFG: 5.5

  • Steps: 25

  • Size: 1024×1344

  • Post-process: Sharpen (radius 1, sigma 0.20, alpha 0.30)

• img2img Refinement:

  • Sampler: euler_ancestral

  • Scheduler: beta

  • CFG: 4.5

  • Steps: 25, start_at_step: 10

  • Post-process: Contrast Fix (blend_factor 0.35, soft_light) + Sharpen (radius 1, sigma 0.25, alpha 0.35)

• Upscale Stage:

  • Sampler: euler_ancestral

  • Scheduler: beta

  • CFG: 4.0

  • Steps: 25, start_at_step: 20

  • Post-process: Contrast Fix (blend_factor 0.30, soft_light) + Sharpen (radius 1, sigma 0.30, alpha 0.40)

Recommended Prompt:
Positive: masterpiece, best quality, very aesthetic, detailed eyes, detailed face, intricate details, beautiful lighting, volumetric lighting, sharp focus, highly detailed, amazing quality,

Negative: bad quality, worst quality, worst detail, sketch, censor, lowres, bad anatomy, bad hands, extra fingers, missing fingers, blurry, low quality, text, watermark, ugly, poorly drawn face, poorly drawn hands, logo, distorted anatomy,

Strengths

• Consistent, high‑quality anatomy and body structure in both portraits and full‑body shots

• Excellent lighting, depth, and volumetric effects

• Distinctive style

• Handles detailed accessories, hair, clothing, and complex scenes well

• Versatile across character art, fashion, and atmospheric compositions
  

Notes

• Best results at moderate CFG (3.5–6)

• Compatible with most SDXL LoRAs

• Optional FreeU_V2 or ADetailer can further improve details

• The included pipeline settings are tuned to maximize detail retention and style consistency