From One Image to an AI Short Film in 48 Hours: The Making of 'In the Quiet Between'
- ARTHART STUDIO
- Apr 29, 2025
- 3 min read
We recently participated in the Runway Gen:48 Film Festival — a unique filmmaking challenge that pushes creators to the edge of their imagination, and timeline. In this 48-hour sprint, participants were tasked with creating a short film based on a surprise theme released by Runway. From concept to final cut, every frame had to be imagined, generated, and delivered within just two days.
For us, this wasn’t just about fast storytelling. It was about discovering how far AI tools could stretch the limits of independent filmmaking. The result: In the Quiet Between — a surreal, philosophical short crafted entirely with the help of generative AI.
The 48-Hour Sprint: Where Creativity Meets Constraint
The clock started ticking the moment the theme dropped. With only 48 hours to conceptualize, visualize, animate, and finalize a complete short film, there was no room for error or inefficiency. We needed a production pipeline that was agile yet visually cohesive.
Enter Runway’s References, a brand-new feature launched exclusively for Gen:48 participants.
This tool became the cornerstone of our workflow. It allowed us to take a single reference image — in our case, generated via Midjourney — and build an entire cinematic universe around it.
The Challenge of Visual Consistency in AI
One of the biggest hurdles in AI-driven content creation has been visual consistency. While tools like Midjourney or Flux can create breathtaking one-off images, generating an entire sequence that visually "sticks together" has been notoriously difficult.
Characters might look slightly different between frames. Background elements shift unpredictably. Style and atmosphere often lack continuity.
Runway’s References changes that. By feeding the model a single image as a visual anchor, we were able to extend scenes, generate alternate angles, and maintain a consistent character and mood throughout the film. It allowed us to direct with intention, not compromise.
The Making of In the Quiet Between
Our story centered around silence, reflection, and transformation — themes best expressed through strong, symbolic visuals. We started with a single Midjourney frame that felt emotionally resonant: a lone rabbit-headed figure in a barren desert surrounded by floating mirror frames.
From that one image, we:
Generated consistent environmental scenes
Created multiple angles of the same character
Produced transition frames to support storytelling flow
Using Runway References, we were able to ensure the protagonist looked and felt the same across every shot. The lighting, color tone, even small visual motifs stayed intact — something that would have been near impossible without this technology.
Redefining the Filmmaker's Toolkit
This project showed us what’s possible when generative AI is integrated seamlessly into the filmmaking pipeline. What once required a team of concept artists, animators, and weeks of iteration, now becomes achievable in a single weekend by a compact creative team.
It isn’t about replacing human creativity. It’s about amplifying it.
We still wrote the narrative, crafted the emotional arc, and directed the visual tone. But AI tools like Runway handled the heavy lifting — fast, consistently, and beautifully.
What This Means for the Future of Cinema
Generative AI is poised to democratize filmmaking in ways we’re only beginning to understand. With tools like Runway References, small studios can compete with big production houses. Ideas that once felt impossible due to visual complexity are now within reach.
In the Quiet Between is our small proof-of-concept: a complete, emotionally-driven film built on a single image. It’s not just a film. It’s a sign of what’s to come.
Watch the Film
Join the Conversation
We believe this is just the beginning. If you’re experimenting with AI tools in your creative practice or are curious about the future of AI filmmaking, we’d love to hear from you.
Feel free to share your thoughts, questions, or collaborations with us. Let’s shape the future of cinema — one frame at a time.
Comments