Wiki
Edit history

<div style="color:#e0e0e0; background-color:#0e0e0e; font-family:Segoe UI, sans-serif; padding:1em; line-height:1.7; font-size:17px;">

  <p><span style="font-size:22px; color:#ff4081;"><strong>Animating FusionFall: Building a Workflow for Consistent Fan Media</strong></span></p>

  <p>FusionFall has a very specific and recognizable visual style. Its characters use an inverse hull toon shader, giving them a distinctive silhouette defined by bold toon outlines. Preserving this look is essential when creating AI-generated videos or reinterpretations based on FusionFall assets.</p>

  <p>This page documents the process I’ve been developing to generate consistent animation and fan media from in-game screenshots, without drifting away from the original art direction.</p>

  <p>For a recent project, I created animations of Lance and Ilana from <em>Sym-Bionic Titan</em>. Below, I’ll walk through the steps we used, the tools involved, and how we adapted prompts to better match our visual goals.</p>

  <p>Our team currently uses <a href="https://runwayml.com" style="color:#88c0ff;">Runway ML’s Unlimited Plan</a>, but for smaller or amateur productions, <a href="https://leonardo.ai" style="color:#88c0ff;">Leonardo AI</a> offers a cheaper and often more effective solution.</p>

  <h3 style="color:#ff4081;"><strong>Prompting Lance: Starting with Animation</strong></h3>

  <p>I first asked ChatGPT:</p>
  <blockquote style="border-left: 3px solid #ff4081; padding-left: 1em; font-style:italic; color:#ccc;">
    “Describe how you would animate Lance from Sym-Bionic Titan and then write a prompt for that.”
  </blockquote>

  <p>Here’s what it generated:</p>

  <div style="background-color:#1a1a1a; padding:1em; border-radius:6px; border-left:4px solid #ff4081;">
    <p><strong>Prompt:</strong> Animate a character inspired by Lance from <em>Sym-Bionic Titan</em> in the visual style of Genndy Tartakovsky. Use sharp, clean linework with flat cel shading and bold silhouettes. The character should appear stoic and disciplined, with reserved facial expressions, mostly neutral or slightly stern. Animate subtle motions: slight brow movement, tight mouth, controlled posture. When in action, depict powerful, efficient body language with punchy, angular movement arcs. Emphasize hold frames on key impacts and avoid excessive smear frames. Background should be minimal or abstract, focusing attention on the figure’s silhouette and precise animation timing.
    </p>
  </div>

  <p style="margin-top:1em;">This prompt produced solid stylistic results, but the animation had too much movement for a character like Ilana, who needed more subtlety. So we changed our approach.</p>

  <h3 style="color:#ff4081;"><strong>Prompting Ilana: Focused Performance</strong></h3>

  <p>I started with a similar prompt for Ilana:</p>
  <blockquote style="border-left: 3px solid #ff4081; padding-left: 1em; font-style:italic; color:#ccc;">
    “How would you do Ilana? Give me a prompt for her.”
  </blockquote>

  <p>But again, it added too much motion. So I refined the request:</p>
  <blockquote style="border-left: 3px solid #ff4081; padding-left: 1em; font-style:italic; color:#ccc;">
    “Just make her talk. Maybe the camera zooms in.”
  </blockquote>

  <p>And this was the improved result:</p>

  <div style="background-color:#1a1a1a; padding:1em; border-radius:6px; border-left:4px solid #ff4081;">
    <p><strong>Prompt:</strong> Animate a close-up talking scene of a stylized princess character inspired by Ilana from <em>Sym-Bionic Titan</em>, using Genndy Tartakovsky’s visual style. Begin with a medium shot and slowly zoom in toward her face as she speaks. She has an elegant, poised demeanor with soft, expressive facial acting: subtle brow movement, gentle blinking, and clearly articulated lip sync. Her eyes occasionally glance to the side or shift slightly as she thinks. Her posture is upright and noble. Her gestures are delicate, such as one hand near her chest as she emphasizes a point. Use clean linework, flat cel shading, and a minimal background to keep focus on her graceful expression and composed delivery.
    </p>
  </div>

  <h3 style="color:#ff4081; margin-top:1.5em;"><strong>Example Output: Lance Animation</strong></h3>

  <p>Below are the resulting animations created using the Lance prompt above:</p>

  <div style="text-align:center; margin:1.5em 0;">
    <iframe width="560" height="315" src="https://youtube.com/embed/tc0jcUFrD80" title="Lance Animation Preview" frameborder="0" allowfullscreen style="border:2px solid #ff4081; border-radius:8px;"></iframe>
  </div>
  <div style="text-align:center; margin:1.5em 0;">
    <iframe width="560" height="315" src="https://youtube.com/embed/_QMx6UX3R5s" title="Lance Animation Preview" frameborder="0" allowfullscreen style="border:2px solid #ff4081; border-radius:8px;"></iframe>
    <p style="color:#aaa; font-size:14px;">↑ This was generated from the first Lance animation prompt.</p>
  </div>

  <h3 style="color:#ff4081;"><strong>Bonus: Flux Reinterpretation of Lance</strong></h3>

  <p>Before starting animation, I accidentally ran the Lance video prompt through Leonardo using the Flux model by Kontext. Instead of creating a video, it generated a single still image. The result was unexpected but surprisingly effective. It reimagined Lance with a completely different visual tone that still felt adjacent to FusionFall. We might use this approach more going forward as a stylized reinterpretation technique.</p>

  <div style="text-align:center; margin:1.5em 0;">
    <img src="https://fusionfall.org/wiki/lance.jpg" alt="Flux reinterpretation of Lance" style="max-width:100%; border:2px solid #ff4081; border-radius:8px;" />
    <p style="color:#aaa; font-size:14px;">↑ This image was generated in Leonardo using the Flux model. It wasn’t planned, but it gave us a great visual direction to explore.</p>
  </div>

</div>
Wiki