/

/

TIDEBOUND (ANIMATION)

As the “Cloud Tide” threatens to engulf the only bridge to the sky for a week-long storm, a high school girl pushes her physical limits in a race against a countdown, determined to reach the boy she loves before the path vanishes into the white abyss.

Video Prompt List

SH010: On a bustling animated street, the girl on her bicycle stops and looking at the trem, watching intently as the vibrant yellow tram glides past her. Pedestrians move in the background. After the tram clears the frame, the camera slowly tilts up to reveal the detailed floating islands in the sky.

SH012: The girl on her bicycle pauses, her gaze fixed on the glowing digital sign displaying “CLOUD TIDE RISING” and a countdown. With a determined expression, she starts to pedal rapidly uphill. The camera smoothly tracks right, following her as she ascends the cobblestone street.

SH013: The girl on the bicycle pedals rapidly, her legs pumping with dynamic forward momentum. The bicycle wheels spin quickly as she moves along the elevated road into the drifting clouds. A low-angle cinematic tracking shot follows closely behind her.

SH015: A cinematic tracking shot follows behind the determined anime-style character on the bicycle, pedaling persistently along a high, curving bridge. The camera gradually moves closer and transitions to an over-the-shoulder perspective, revealing the thick, dark fog ahead that makes navigation difficult. Rain heavily splatters on the wet road, creating a fine mist as the bicycle wheels cut through puddles pushing deeper into the unknown.

SH014: A cinematic tracking shot, dolly right, following the determined character as they vigorously pedal their bicycle. A powerful wind visibly sweeps through their dark hair and yellow jacket. The scene begins engulfed in a thick, ethereal mist. As the character cycles forward, the dense fog slowly and gracefully dissipates, gradually revealing a breathtaking, clear sunset sky painted with warm orange, pink, and purple hues. The transition from mist to vibrant twilight is seamless and dreamlike.

SH017: The character on the bicycle pedals steadily forward on the illuminated bridge. The camera tracks closely behind the character, then smoothly transitions to an over-the-shoulder shot, culminating in a medium shot revealing the character and the majestic, glowing floating city in the distance. Wispy clouds and soft fog drift languidly around the bridge and beneath the floating island, catching the golden light of the vibrant sunset. Anime aesthetic, cinematic view.

SH018: A static wide shot. The character on the bicycle gasps and looks intently down at the immense drop, then begins to pedal the bicycle toward the right, disappearing behind the house. Clouds gently drift and swirl beneath the floating island.

SH020: A static shot. The girl blushes and looks up at the boy, then quickly fixes her messy hair and bows deeply in apology. The boy approaches her and gently raises his right hand toward her shoulder.

SH021: A static wide shot, dark blue twilight. Thick, ethereal clouds slowly and steadily rise from the deep chasm below the floating island, gracefully engulfing and obscuring the winding road that stretches into the distance. Simultaneously, the warm glow from the house windows and the solitary lamppost by the cliff edge softly flickers and wavers, casting dancing shadows. Anime aesthetic, cinematic mood.

This animated iteration of Tidebound serves as a direct evolution of my “Nano Banana Pro” experiments, pushing the boundaries of what these generative tools can achieve. For this piece, I leaned heavily into new Kling’s native audio capabilities (Kling 2.6), where the SFX, ambience, and foley were generated directly by the video model itself.

One of the most intriguing aspects of this workflow was observing the model’s own interpretation of my prompts. A perfect example of this is the dialogue at the end of the clip [SH020]; while I didn’t explicitly script those lines, the AI “hallucinated” or interpreted the scene’s context to include them—a surprising, yet fitting, creative choice! (Hahaha 😅).

For those curious about the visual foundations of this project, you can explore the full behind-the-scenes breakdown of the images used in this animation on this page.

Release Date

December 5, 2025

Runtime

00:01:08

Directed by

Taruma Sakti

Technique

txt2img, img2img, img2vid

AI Platform

OpenArt (Video), Higgsfield AI (Image), Google AI Studio (LLM)

Text/LLM Model

Gemini 2.5 Pro, Gemini 2.5 Flash

Image (with Editing) Model

Nano Banana Pro

Video Model

Kling 2.6

Other Tools

Shotbuddy v3 (AI Asset Manager), Typingmind (LLM Interface), Adobe Premiere Pro (Video Editor)

Music

Paper Crane by Egaoya feat. KOKORISA

posted by

Recent Works