ADC Awards
ADC Annual Awards is the oldest continuously running industry award show in the world, with an incredible legacy of over 100 years. These awards celebrate the very best in advertising, digital media, graphic and publication design, packaging and product design, motion, experiential and spatial design, photography, illustration and fashion design – all with a focus on artistry and craftsmanship.
Category
Film / Moving Image / Fully AI-Generated Film / Short
URL
Click to launch
Annual ID
ADC105_ART002G
Railbound is a poetic short film built from the work of Mike Brodie — the prolific American photographer who, in his early twenties, train-hopped across the country documenting a subculture of young drifters moving by freight rail. His images captured a raw, intimate portrait of youth on the edge: freedom, restlessness, and the families we find along the way.
The film blurs the line between documentary and fiction, imagining the emotional stories inside those photographs — stories of movement, belonging, and letting go. The project explores whether generative video can extend photographic truth without rewriting it. Rather than fabricate new worlds, we treated AI as a temporal tool: a way to reveal motion already latent inside stillness.
Created during the inaugural Google Labs Flow Sessions residency, Railbound was developed over six weeks as part of a cohort exploring the future language of generative cinema. Working directly with Brodie — who opened both his published and unreleased archive — we built a hybrid workflow combining generative motion (Google Flow / Veo 3), AI Studio refinements, manual VFX correction, and traditional film disciplines.
Each frame began as a real photograph. Over 180 shots were iterated repeatedly to preserve lighting, grain, and expression. Imperfections were protected. Grain was reintroduced. Movement was restrained. If animation called attention to itself, it was rebuilt.
The narrative — a father and daughter in conversation — connects images taken years apart into a single emotional memory.
Railbound proposes a new cinematic grammar: photography as duration, archive as performance, and AI not as spectacle, but as preservation.
The film blurs the line between documentary and fiction, imagining the emotional stories inside those photographs — stories of movement, belonging, and letting go. The project explores whether generative video can extend photographic truth without rewriting it. Rather than fabricate new worlds, we treated AI as a temporal tool: a way to reveal motion already latent inside stillness.
Created during the inaugural Google Labs Flow Sessions residency, Railbound was developed over six weeks as part of a cohort exploring the future language of generative cinema. Working directly with Brodie — who opened both his published and unreleased archive — we built a hybrid workflow combining generative motion (Google Flow / Veo 3), AI Studio refinements, manual VFX correction, and traditional film disciplines.
Each frame began as a real photograph. Over 180 shots were iterated repeatedly to preserve lighting, grain, and expression. Imperfections were protected. Grain was reintroduced. Movement was restrained. If animation called attention to itself, it was rebuilt.
The narrative — a father and daughter in conversation — connects images taken years apart into a single emotional memory.
Railbound proposes a new cinematic grammar: photography as duration, archive as performance, and AI not as spectacle, but as preservation.
2026 Awards
Total Points: 45
Gold Cube
Credits
Animator
Kris Cave
Editor
Jake Bann
Producer
Alex Naghavi
Jonathan Perry
Sound Designer
Jeffrey Reed
Actor
Jeffrey Reed
Violet Geissler
Casting
Jake Lyon
Colorist
Dave Dixon
Original Photography
Mike Brodie
Original Score
Dan Michaelson
Production Assistant
Kemetrea Spearman
Sound mixer
Matt Wymer
Title Design & Graphics
Alex Naghavi
VFX Digital Artists
Ali Khan
M. Shahrukh
Norman Davis
Visual Director & AI Artist
Alex Naghavi
Written & Directed By
Jonathan Perry
Related Awards
