>>> LEDGER — 01_WORKS
## MooMooIL
- Base family: Illustrious
- Verified showcase settings: 1024×1024 / Euler a / 28 steps / CFG 5
- Observed visual identity:
- mature adult character proportions
- cinematic directional lighting
- fashion/editorial illustration influence
- strong foreground / midground / background separation
- expressive full-body staging
- capable multi-character ensemble composition
- stylistic continuity between portrait and ensemble scenes
- Current strongest differentiator:
scene construction and character staging rather than isolated character rendering
- Current weakness:
facial-beauty convergence, especially among female characters; background-face fidelity decreases in dense ensembles
- Public claim ceiling:
"multi-character staging / ensemble scenes"
Do not yet claim persistent individual identity across complex crowds.
- Recommended showcase order:
ensemble scene → dynamic full-body scene → mature portrait → intimate portrait → additional ensemble evidence
Description
<iframe width="560" height="315" src="https://www.youtube.com/embed/LmnbWZLbZlI?si=3Xq280TqCLaUBkca" title="YouTube video player" frameborder="0" allow="accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture; web-share" referrerpolicy="strict-origin-when-cross-origin" allowfullscreen></iframe>MooMooIL v2.0 is a major update focused on broader visual storytelling, stronger character diversity, and more cinematic composition.
This version significantly expands beyond conventional portrait-oriented illustration, with improved performance across narrative scenes, environmental compositions, character interactions, and cinematic framing.
Key improvements in v2.0:
Improved cinematic and narrative composition
Stronger male and female character rendering with more distinct facial structures and body proportions
Better multi-character interaction and physical posing
Improved emotional expressions and facial acting
Better handling of close-ups, medium shots, full-body shots, and wide establishing shots
Stronger environmental and background generation
Improved integration between characters, props, lighting, and scenery
More natural semi-realistic 2.5D rendering
Refined skin, hair, fabric, and material rendering
Improved dramatic lighting, rim light, backlight, and atmospheric depth
Better support for everyday scenes, romance, fantasy, surreal imagery, creatures, animals, vehicles, and cinematic environments
Reduced dependence on conventional “AI beauty portrait” aesthetics
The goal of v2.0 is not simply to generate attractive characters, but to create illustrations that feel like frames taken from a larger story.
The model continues to be developed primarily from my own curated and hand-created illustration dataset, with additional refinement focused on visual consistency, composition, and storytelling capability.
FAQ
Comments (1)
A very interesting model! I tried both versions of your model, and v2 is a clear improvement. But unfortunately, my face still looks a bit blurry, especially the eyes. If you could work a little more to eliminate artifacts on the face, it would be a really cool model👍