Part 1. Introducing Vidu Q1: Giving Form to the "Memory" of Sight and Sound
The first time I used Vidu Q1, what stunned me most was the dramatic improvement in the atmosphere and depth of the visuals. While maintaining the high consistency of illustrations that I'd come to expect, the added layers of depth and lighting nuances gave the entire piece a cinematic quality.
That said, in some scenes, the 3D-like texture felt slightly overemphasized, softening the original hand-drawn warmth and subtlety. Still, these were minor adjustments that could be refined in post-production. Overall, the work transcended the typical "AI-generated" feel, achieving something uniquely expressive.
Another standout was the evolution of AI Sound Effects. With Vidu AI video generator, you can layer multiple sound effects and precisely time their triggers within a 0-10 second range, allowing for meticulous synchronization with the emotional flow and rhythm of the scene.
During the beta phase, prompt effectiveness—whether in Japanese or English—was inconsistent. Crafting sound effects for this video involved heavy trial and error. Some AI-generated sounds were used as-is, but most required editing—splicing, pitch-shifting, or speed adjustments—to fit seamlessly. The official release, however, showed marked improvement, especially with English prompts. What once took painstaking effort (like generating realistic footsteps) now works smoothly with simple instructions.
Part 2. The Creative Process: When a Story "Emerges" on Its Own
This project, The Nameless Sound, began as an experiment to create an opening sequence for a fictional anime. My initial vision leaned toward cyberpunk fantasy, but as I worked, the theme of memory pulled me in deeper.
I hadn't planned a full narrative upfront. Instead, the story unfolded organically—a single prompt evoking a character's emotion, a flicker of light suggesting "this must be someone's memory." It felt less like deliberate construction and more like discovering a story that was already there.
The protagonist, Lyra, became a girl who "senses others' memories through touch"—yet no traces of her own past exist in the city. When this concept crystallized, the core idea took shape: to create something that lingers in memory, not just in records.
Collaborating with ChatGPT was like co-writing. It helped refine the setting, deepen character psychology, and even craft the title, dialogue, poetic phrasing, and social media captions. The dialogue shaped the project's identity piece by piece.
For the visual composition and editing, I trusted my instincts. Though I had storyboards and timing sheets, I treated them as guidelines rather than rigid rules, adjusting based on the emotional resonance of each scene. Editing became part of the creative act—a way to uncover the story as I went.
