Inspiration
This project began first and foremost with my work as an author. I’ve been building original fantasy sagas — published on Amazon and beyond — where lore, characters, and worldbuilding always come first. Writing has been my foundation: from The Last Igigi to The Eternal Lighthouse, each story explores human emotion through myth, memory, and the supernatural.
Fairy Dust emerged as part of that lineage — a written epic reimagined into a visual medium. What inspired me was the challenge of taking the same care and craft I put into my novels and applying it to animation, music, and narration. Rather than being a content creator who dabbles in story, I wanted to prove that an author’s vision can expand across formats without losing depth.
The twin queens — Aurea of Light and Velmourn of Storm — come directly from that storytelling drive: characters conceived on the page, then brought to life through cinematic AI. And at the center, the First Tree, Thalyndra, stands as the kind of mythic symbol I’ve always written into my books: a monument to memory, destiny, and magic.
What it does
How we built it
Story & Worldbuilding: Written as part of the Jai & Zeek Universe, the lore was developed across multiple chapters, weaving characters, dragons, and councils into a cohesive kingdom history.
Visuals: Veo AI for cinematic animations and sequences. Visionstory AI for AI Talking photo DALL·E & ImageGen for ultra-realistic character portraits and fantasy landscapes. Tiktok for Close Captions
Audio: ElevenLabs for narration (Deyrian, the Master Dustweaver’s voice). ElevenLabs for background music
Editing & Assembly: Cut and synced in DaVinci Resolve, combining narration, visuals, and music into one cohesive cinematic trailer.
Challenges we ran into
Consistency Across Mediums: Different AI models interpret prompts in different ways. Achieving visual continuity (armor styles, wings, dragons) across Veo, DALL·E, and Sora required constant refinement.
Technical Limits: Some tools imposed strict aspect ratios (often defaulting to 1:1), making true 16:9 cinematic frames difficult to enforce. This meant re-rendering scenes multiple times until they matched.
Narration & Timing: Deyrian’s Loom narration often ran longer than expected, so additional scenes had to be crafted (e.g., Kerrin’s scouting sequence) to balance voice and visuals.
AI Restrictions: Tools occasionally flagged original AI renders as “prominent persons,” preventing uploads — forcing creative workarounds by relying solely on descriptive continuity instead of reference images.
Accomplishments that we're proud of
What we learned
What's next for Deyrian Narrates
continue posting on all prominent media platforms creations to enhance my skills
Built With
- alotofpatience
- chatgpt
- davinciresolve
- elevenlabs
- tiktok
- veo
- visionstory
Log in or sign up for Devpost to join the conversation.