🎬 CineWatch puts Overwatch heroes into a cinematic CGI look: film-style lighting, real depth of field, moody shadows and detailed backgrounds, like frames from a cinematic trailer.📚 An Anima checkpoint trained on 432 curated cinematic renders of 25 Overwatch heroes. Every image is captioned with booru tags plus a full sentence, so it understands both prompt styles.<hr /><h3 id="what-it-does-well">✨ What it does well</h3><ul><li>🦸 Heroes on-model by name, no trigger word: <code>tracer \(overwatch\), overwatch</code> gives Tracer with her outfit and gear.</li><li>💡 Lighting that follows your prompt: backlight and silhouettes, neon, firelight, storms, low-key dark scenes.</li><li>🎯 Keeps Anima's prompt following: colours, left/right placement, counts and short text still work.</li><li>🌐 Non-Overwatch prompts stay neutral. Other characters come out in the same 3D CGI look, and adding <code>anime screencap, anime coloring, flat color, cel shading, 2d</code> gives clean 2D anime 🎨</li></ul><h3 id="heroes-in-the-dataset">🦸‍♀️ Heroes in the dataset</h3>Ana, Anran, Ashe, Brigitte, D.Mon, <a target="_blank" rel="ugc" href="http://D.Va">D.Va</a> (and MEKA), Domina, Freja, Illari, Juno, Junker Queen, Kiriko, Mei, Mercy, Moira, Pharah, Shion, Sierra, Sojourn, Sombra, Symmetra, Tracer, Vendetta, Widowmaker, Zarya.The dataset is mostly women; male heroes come from Anima's own knowledge.<hr /><h3 id="how-to-use-(comfyui)">⚙️ How to use (ComfyUI)</h3><ul><li>🧠 Load Diffusion Model: <code>CineWatch_AnimaV1.safetensors</code> (diffusion model only, bf16)</li><li>📝 Text encoder: <code>qwen_3_06b_base.safetensors</code> (Load CLIP, type stable_diffusion)</li><li>🖼️ VAE: <code>qwen_image_vae.safetensors</code></li><li>Same text encoder and VAE as the Anima base model.</li><li>🎛️ ModelSamplingAuraFlow shift 3 · 30 steps · CFG 4 · Euler · Simple</li><li>📐 Around 1.3 MP: the gallery uses 864x1536 and 1536x864</li></ul><h3 id="prompt-format">✍️ Prompt format</h3><code>masterpiece, best quality, score_7, safe, 1girl, solo, &lt;hero&gt; \(overwatch\), overwatch, &lt;tags&gt;. &lt;One or two sentences describing the moment, the camera angle and the light.&gt;</code>💬 Escape the brackets in hero tags in ComfyUI: <code>kiriko \(overwatch\)</code>.🎥 Tip: naming the camera angle and the light source in the sentence ("worm's-eye view", "lit only by the forge", "backlit by floodlights") makes the biggest difference.<h3 id="negative-prompt">🚫 Negative prompt</h3><code>worst quality, low quality, score_1, score_2, score_3, blurry, jpeg artifacts, sepia, bad anatomy, bad hands, extra fingers, missing fingers, extra digit, fewer digits, extra limbs, deformed, watermark, text, signature, artist name, logo</code><hr /><h3 id="known-limits">⚠️ Known limits</h3><ul><li>"cowboy shot" often still gives a full body.</li><li>Lettering on clothes (jerseys, logos) comes out garbled.</li><li>Domina tends to get sneakers instead of her silver boots.</li><li>Two-person action (who holds the ball, who blocks) can take a few seeds.</li></ul>📸 Every gallery image has its full prompt and settings attached.

Files with SHA256: 1e52dc9a141f5b23412e6eb2165aaf1e52c66bab95aa2aef09dca909bd0d1c93