80s–90s Bombshell Style LoRA (IMG2IMG) – Vintage Cinema Glamour | Analog Film LookDescription: This SDXL LoRA is built from ~800 film stills capturing the look and presence of 1980s–early 1990s cinematic bombshells—an era defined by controlled glamour, confident screen presence, and rich analog lighting.Important context on the training: This dataset includes 200+ different actresses and was originally intended to be structured differently. As trained, it does not function as a consistent character LoRA—it blends too many identities into one. As a result, it will fail if used like a character trigger.However, that outcome produced something valuable: This LoRA works effectively as an IMG2IMG style model, translating images into a cohesive late-80s cinematic aesthetic.The visual language it captures draws from archetypes like the Bond girl, high-society antagonists, and leading women shaped by performances from Sharon Stone, Kim Basinger, Michelle Pfeiffer, Catherine Zeta-Jones, and Monica Bellucci.What it does well (intended use):<ul><li>Image-to-image style transfer into late 80s / early 90s film aesthetic</li><li>Adds analog film grain, soft diffusion, and warm highlight rolloff</li><li>Introduces cinematic lighting (lamps, window light, low-key interiors)</li><li>Enhances posture and presence toward confident, composed “leading lady” energy</li><li>Shifts wardrobe and mood toward upscale, dramatic, and filmic styling</li></ul>Core Style Traits:<ul><li>Soft, practical lighting (tungsten, lamp glow, window light)</li><li>Film grain and analog color response</li><li>Slightly warm tones with softened contrast</li><li>Glamorous but grounded realism (not exaggerated or anime)</li><li>Environments: luxury interiors, marble bathrooms, night scenes, poolside, glass-heavy architecture</li><li>Hair and makeup consistent with the era (volume, texture, bold refinement)</li></ul>Training Focus:<ul><li>Adult women with strong screen presence</li><li>Cinematic framing and composition</li><li>Late-night cable / B-movie visual tone</li><li>High-status, controlled, confident energy</li></ul>Suggested Workflow (Important):<ul><li>Use primarily with img2img, not txt2img</li><li>Start with a strong base image and let the LoRA restyle it</li><li>Keep denoise moderate to preserve structure while applying style</li><li>Pair with realistic or cinematic SDXL checkpoints</li></ul>Suggested Prompting: <code>80s film still, 90s glamour, cinematic lighting, analog film grain, elegant, confident, mature woman</code>Notes:<ul><li>Not a character LoRA—identity will drift and blend</li><li>Strength is in style unification across diverse inputs</li><li>Covers a wide aesthetic range across the late 80s into early 90s</li><li>Best results come from focusing on mood, lighting, and scene—not specific faces</li></ul>What came out of this is a broad, cohesive cinematic style translator—less about who the subject is, and more about making everything feel like it belongs to that era of film.

80s–90s Bombshell Style LoRA (IMG2IMG) – Vintage Cinema Glamour | Analog Film Look

SDXL_Elmore_epoch_9.safetensors

Sporty_Granny_epoch_6.safetensors

v2.0

Larry Elmore Style LoRA – Classic Fantasy Illustration | D&amp;D Legacy, Cinematic RealismDescription: This SDXL LoRA is a focused tribute to the visual language of Larry Elmore—one of the most influential fantasy illustrators of the modern era. His work helped define the look and tone of Western fantasy across tabletop gaming, novels, and visual media for decades.Larry Elmore rose to prominence through his iconic contributions to Dungeons &amp; Dragons during its formative years, creating imagery that shaped how entire generations تصور heroes, dragons, armor, and epic worlds. His paintings are known for their balance of realism and idealism—figures feel grounded and anatomical, yet elevated into myth.His influence extends far beyond tabletop:<ul><li>Early fantasy book covers and paperback art direction</li><li>The visual identity of RPGs and MMOs</li><li>Foundations for modern video game character design</li><li>Aesthetic DNA seen in fantasy films and cinematic concept art</li><li>The heroic proportion and staging now standard in illustration pipelines</li></ul>This LoRA aims to reproduce those core traits:<ul><li>Strong, heroic compositions with clear storytelling</li><li>Clean anatomy with idealized proportions</li><li>Painterly realism with controlled brushwork</li><li>Warm, natural lighting and outdoor environments</li><li>Classic armor, fabric, and material rendering</li><li>Expressive but grounded characters (not exaggerated anime forms)</li><li>Balanced color palettes with subtle saturation shifts</li></ul>Unlike heavily stylized or exaggerated fantasy models, this leans toward illustrative realism—where characters feel believable, but elevated into legend.Training Focus:<ul><li>Fantasy heroes and heroines</li><li>Armor, swords, cloaks, and traditional gear</li><li>Natural environments (mountains, forests, skies)</li><li>Painterly textures and classic illustration finish</li><li>Clean faces, readable emotion, composed posture</li></ul>Suggested Trigger / Prompting: <code>larry elmore style, fantasy illustration, heroic pose, detailed armor, natural lighting, painterly, realistic proportions</code>Recommended Use Cases:<ul><li>High fantasy character art</li><li>D&amp;D-style party compositions</li><li>Book cover illustrations</li><li>Cinematic concept art with classical grounding</li><li>Hero portraits with narrative weight</li></ul>Notes:<ul><li>Works best with SDXL checkpoints tuned for realism or painterly output</li><li>Avoid overly stylized/anime prompts if aiming for authentic results</li><li>Benefits from lighting descriptors (sunset, golden hour, diffused daylight)</li></ul>Official Site: <a target="_blank" rel="ugc" href="http://www.larryelmore.com/">http://www.larryelmore.com/</a>This model is built around discipline—composition, anatomy, and storytelling first. It’s less about exaggeration and more about presence, clarity, and legacy-quality fantasy imagery.

Files with SHA256: 701d59ba0246296fe42a645aa2786f2ae1f4eec5e1507c494beeff623d1a45df