1. Story Written Across Time
A cinematic AI journey
“Story Written Across Time” is a romantic journey across historical epochs and the infinite expanse of the cosmos. It’s a visual poem about love that transcends time — from the thrones of kings to the temples of gods, from the sands of empires to the vastness of galaxies. This is more than just a video. It’s a cinematic story brought to life through AI, art, and editing.
Concept and Script
I started with a simple question: “What if this love had lived a thousand lives?”
From there, I built scenes spanning from the Medieval Age to Ancient Greece, the Roman Empire, Ancient Egypt, and a futuristic world among the stars.
Each era tells a chapter of this eternal love, using color and atmosphere to evoke distinct moods.
From there, I built scenes spanning from the Medieval Age to Ancient Greece, the Roman Empire, Ancient Egypt, and a futuristic world among the stars.
Each era tells a chapter of this eternal love, using color and atmosphere to evoke distinct moods.
Reference Gathering and Moodboards
I gathered visual references for each era:
Medieval – grand halls, royal thrones
Ancient Greece – temples, marble statues
Roman Empire – colosseums, imperial architecture
Ancient Egypt – pyramids, hieroglyphs, pharaoh’s symbols
Future – cosmic pathways, stardust portals
AI Image Generation
Using MidJourney and Dzine.ai, i generated the key visuals, testing different angles and compositions to find the most cinematic and emotional shots. Here you can see the options that did not make it into the final edit:
Editing and Color Grading
The scenes were seamlessly stitched together with transitions evoking portals through time and space. Colors were carefully balanced to enhance the atmosphere of each era.
Sound Design and Music
Narration by ElevenLabs
Music and sound effects from Envato Elements for deeper immersion
Key Tools
MidJourney / Dzine.ai – image generation, face swap
Kling, Sora, Runway – video generation
Adobe After Effects / Capcut – editing, visual transitions and captions
ElevenLabs – narrator’s voice
Envato Elements – music and sound design
I created a looping video that starts and ends in the same place — a testament to the eternal nature of love. Each era is a self-contained story, yet all are connected by a single theme: “No matter where we are, we are always together.”
Final Render
This project is about love that defies time and space — something you can feel in every shot, something deeply personal yet universally relatable.
2. Kyiv, 482 CE
Cozy Nostalgia
A Cinematic Journey into the Heart of Ancient Kyiv
Welcome to Kyiv, 482 CE, a cozy and atmospheric video project that reimagines the dawn of one of Europe’s oldest cities through a warm, cinematic lens. Inspired by the legendary tale of Kyi, Shchek, Khoryv, and Lybid, this project transports viewers to the early medieval world of Kyiv, capturing the daily life, serene landscapes, and timeless rituals of a Slavic settlement on the banks of the Dnipro.
Project Concept
My goal was to create a sense of intimate nostalgia, blending historical authenticity with cozy and cinematic visuals. From sun-drenched wooden interiors and riverside morning mist to bustling markets and quiet evenings by the fire, every frame is meticulously crafted to evoke the gentle rhythms of life in early Kyiv.
3. Kyivan Rus Through the Ages
A Hyper-Realistic Journey
This project is a series of 8 hyper-realistic visual narratives exploring pivotal moments in the early history of Kyiv. Each scene transports the viewer to a different chapter of the region’s formative centuries, from the arrival of the Varangians to the baptism of Prince Volodymyr in 988 AD. My goal was to create immersive, historically inspired visuals that blend cinematic lighting, rich details, and authentic cultural context.
This series is a tribute to the mythical and historical roots of Kyiv, weaving together folklore, archaeology, and artistic vision. It’s a glimpse into a world where history and legend meet.
Watch the video on YouTube:
LINK
LINK
4. Baptism of Prince Volodymyr
Visual Storytelling through AI
This cinematic video recreates the key moments of the Baptism of Kyivan Rus by Prince Volodymyr the Great in 988 AD. Using AI-generated hyper-realistic imagery, we explore the significant historical episodes that shaped Ukrainian spirituality and culture.
Each frame captures the powerful atmosphere of the time, blending detailed medieval costumes, lifelike human expressions, and carefully crafted lighting. From the quiet reverence of a baptismal ceremony to the vibrant pageantry of medieval Kyiv, the video immerses the viewer in a journey through faith, politics, and cultural transformation.
5. Ivan Mazepa: A Visual Chronicle of a Hetman’s Life
Visual Storytelling through AI
This cinematic video project immerses viewers in the life of Ivan Mazepa, one of the most significant figures in Ukrainian history. From his youth at the royal court of the Polish-Lithuanian Commonwealth to his final days in exile, every scene has been carefully crafted with MidJourney’s photorealistic rendering.
We follow Mazepa through his political rise and fall: as a young courtier in Poland, his return to Ukraine, and his service as a general officer of the Zaporizhian Army. Scenes highlight his coronation as Hetman of Ukraine in 1687, the baroque churches he helped build, and his tense negotiations with Tsar Peter I and King Charles XII of Sweden.
Key moments include the tragic burning of Baturyn, the chaos of the Battle of Poltava, and Mazepa’s dramatic escape and final years in exile in Bender, Moldova. Beyond military and political episodes, the video pays tribute to Mazepa’s role as a patron of Ukrainian culture, featuring him in artistic and spiritual settings that shaped the cultural heritage of Ukraine.
Each frame is an atmospheric, photorealistic window into the 17th and early 18th centuries: richly detailed costumes, weathered battlegrounds, baroque architecture, and the faces of warriors and artists. The project concludes with a modern monument to Mazepa in Poltava, linking his historical impact to today’s cultural memory.
During the creation of these projects, I mastered various techniques:
- For image generation, it's essential to use multiple tools instead of relying solely on Midjourney. Midjourney still struggles when there are many people in an image—their faces and hands can become distorted. Also, photorealism heavily depends on individual profile settings and the reference images you provide. To create the best prompts, I used a ready-made tool in ChatGPT "ロMidjourneyロ -- MJ Prompt Generator".
- For transitions, you need to create 2 or 3 videos. For example, I generated an image of a cosmic star tunnel, then in Kling, I set it as the End frame, while for the Start frame, I used a simple green screen (a solid color image). For the second video, I simply used the cosmic tunnel as the Start frame, and the prompt should ensure the camera movement continues smoothly from the previous video (using automated prompt tools like ChatGPT's "Kling 1.6 image to video prompt helper" can save time).
For the third video, I generated it starting from the last frame of the second video to seamlessly continue the motion and fly back out onto a clean green screen. After generating the videos, I used After Effects to remove the green color at the start and end of each clip with simple keying, creating a seamless transition effect. Alternatively, you could get by with just two videos: the first one as described above, and the second generated in reverse (Start frame – the tunnel, End frame – the green screen). Just make sure to refine the prompt so that the camera doesn’t stop in the middle but keeps moving smoothly—this might take more generations than you initially expect.
- Nowadays, there are plenty of tools available for creating consistent characters, even multiple people in one shot. My approach was:
Generating people in Midjourney with realistic body proportions
Using the Face Swap tool in Dzine.ai
Cleaning up and editing in Photoshop
- Sometimes the generated video isn't long enough. In those cases, I used the Premiere Pro Beta's generative tool, which can extend a video by 1–2 seconds at the start or end. It's a very useful feature.
-
Here you can see the generations that didn't make it into the final render. Some of them have very unrealistic camera movement, or strange artifacts. Also, you can see that the faces of the dancing couple change to completely different people.
Thanks for watching my project!
Let me know your thoughts in the comments.
AI videogenerative aiAI animationmotion designNeural NetworksCreative Aicinematic AIvideo experimentAi Artvisual research