No sound, animation, or art should hide the task, cover text, or confuse the next move.
Art, voice, music,
and motion that teach.
Dungeon Academy uses media like a classroom instrument: AI art sets the world, voices guide the lesson, music shapes the room, and SFX make proof feel real.
Future AI upgrades should keep the same asset filenames where possible so the app code does not churn.
Motion should stay optional or gentle for kids who need calmer screens.
Media previews should not require child data or public child profiles.
Six layers that make the school feel alive
Hero, dragon, subject wings, Lyra, mastery moments, and lesson scenes.
Sets the emotional world before a child reads the first prompt.
Short subject loops for math, science, physics, engineering, and AI.
Makes abstract ideas feel visible and memorable.
Professor Lyra, narrator, dragon, and familiar voice packs.
Turns instruction, encouragement, mistakes, and mastery into a cast.
Domain-aware loops for lesson, evidence, challenge, reward, and mastery states.
Gives each room a mood without crowding the task.
Proof filed, rune tap, artifact found, dragon gate, reward, and record save.
Makes learning actions feel satisfying and legible.
Arena lighting, stage rails, character presence, and cinematic transitions.
Keeps the world alive while the lesson remains organized.
Stable slots for better art later
xAI/Grok Imagine source-to-video loop for the front desk hero, with audio stripped for autoplay.
Clean source frame used as the reduced-motion and loading fallback for the dragon/apprentice hero.
Used by public pages, intro surfaces, and the first-room school story.
Shows the five STEM domains and public orientation world.
Stable left-stage character slot for future higher-end replacements.
Stable dragon opponent slot used by the arena and cast-door imagery.
Mentor art for lesson guidance, practice coaching, and classroom presence.
Used for parent reports, handbook, and capstone-style school framing.
The next upgrades to make it cinematic
Regenerate Lyra portrait, speaking poses, and lesson-stage variations with a stable character reference.
Create short image-to-video clips for wake, challenge, shield crack, miss, mastery, and next-room beats.
Upgrade each STEM wing with higher-end camera motion, particles, lab effects, and kid-readable concept framing.
Replace robotic fallbacks with warmer character voices and shorter, more expressive educational clips.
Add satisfying but gentle sounds for practice, evidence, proof, mistake recovery, record filing, rewards, and course completion.
Layer quieter study loops with clear transitions for Teach, Try, Lab, Test, Record, and Mastery.
The media layer should make concepts stick.
The next production pass should replace good placeholders with character-consistent art, warmer voices, richer SFX, and short lesson clips while keeping the learning path calm.
