<h1 id="seph's-illustrious-cinematic-anime">Seph's Illustrious - Cinematic Anime</h1>Preview images are "full final", only prompts (no LoRA's) -&gt; ADetailer &amp; Upscaled! If you feel that there are too few preview images, enable rating R, as I usualy aim for that.Aims for:<ul><li>dark with vibrant colors (goth, neo-/noir)</li><li>generaly artistic images</li><li>2.5d characters (realistic, but still anime)</li></ul><h3 id="settings:">Settings:</h3>For my full workflow (with Forge and DynamicPrompts), please see: <a target="_blank" rel="ugc" href="https://civarchive.com/articles/14013">https://civarchive.com/articles/14013</a><ul><li>Sampler: Euler &amp; Simple</li><li>CFG: 5-8</li><li>Steps: 12-14 as draft/preview, 24, 32, 42 for finals, 64 for super favorites</li><li>Aspect Ratio: should work with most, preview images are: 9:16 , 21:9 / render at: 720x1280, 1536x640, or 720x1536 (10:16 -&gt; cellphone "portrait" wallpaper)</li></ul><h3 id="prompt:">Prompt:</h3><ul><li>+ high_quality, highres, beautiful, detailed, masterpiece, best_quality, absurdres, 8k, HDR , very aesthetic , intricate_details , detailed_hands, elegant_fingers</li><li>- poorly_detailed, jpeg_artifacts, worst_quality, bad_quality</li></ul><h3 id="happy-imagination!">Happy imagination!</h3>Also, I really would like to see what you guys created! :)<h1 id="history-recommendation:">History / Recommendation:</h1>All statements are in regards to 21:9. Meaning either 1536x640 or 1720x720 "draft" renders (14 steps)!v1:<ul><li>Initial release</li></ul>v2-experimental:<ul><li>Raw quality droped - was workaround-able with img2img upscale + adetailer</li><li>Tried for slighly more realism, worked</li><li>Tried to preserve anime char, failed</li><li>Tried to improve "adult appearance" (on certain characters), failed</li><li>Failed to preserve the "intense black/colors"</li><li></li></ul>v3-VAE:<ul><li>Raw quality improved (back as v1 or even better)</li><li>Preserved "intense black/colors", yes</li><li>Age up characters, success (for most ; others are at least better than they were in v2)</li><li>Prompt adherence better than v2</li><li>Recommended IF: you prefer high quality characters with decent but obviously AI landscape/backgrounds</li></ul>v4-VAE-fp16<ul><li>Better "dark/black"</li><li>with few exceptions, most character appear more adult (at least when prompted for)</li><li>better backgrounds</li><li>either good "prompt adherence" or "good misunderstanding"</li><li><s>Info</s><s>: stay at CFG 6 or lower, seems by increasing contrast an "overcooking" has happend, allthough effects are rare.. about 1-2% at CFG:6-8.. HOWEVER, for local gen, this chance can be totaly worth it as it also produces incredible images.</s> --&gt;EDIT: This was caused because i prompted for things, that are no longer required (as much).</li><li>Recommended IF: you prefer characters with high quality landscape/backgrounds, while "full body" is not as good, portrait/macro shots are awesome</li></ul>v4.1-VAE-NOLORA-fp16<ul><li>same as v4</li><li>no included loras</li><li>thus no unwanted effects</li><li>less "glowing artefacts" (hands, breasts, etc)</li></ul>

SephIL-CA (Cinematic Anime)

Basicly: same 4.0 but without integrated LoRa'sThis helps to improve "base" quality, and reduces certain unwanted "glow artefacts" (boobs, hands, etc)

Files with SHA256: 88ebc79d9d0f56f9e1b78614e49fd90c8a52207b2d1c20ccf06a3ebbb5be186d