Case Study · Digital Pipeline · Warner Bros. · 2003–2005
No Roadmap.
No Precedent.
The World's First.
In 2003, there was no established process for shooting a stop-motion feature film digitally. The Corpse Bride pipeline was built from first principles, under extreme production pressure, while the cameras were already rolling. It became a highly referenced case study and the beginnings of a foundation the entire stop-motion industry was eventually built on.
First
Stop-motion feature shot entirely digitally
160K+
Frames processed through the colour pipeline
2003
Year the digital stop-motion pipeline was first built
FCP + DSLR
First feature to use both Final Cut Pro and digital SLR cameras
The Context
Tim Burton's Corpse Bride was the first stop-motion animated feature film shot entirely digitally — moving from the physical film cameras (Mitchell cameras) that had defined the medium since its inception to a fully digital, "born digital" foundation. There was no roadmap for what this meant technically, no established colour pipeline, and no precedent for delivering a digitally originated stop-motion feature to a studio for distribution.
I was one of the first hires on the production and the last person to leave — closing the doors and handling the complete handover to Warner Bros. archives when the film was finished.
Image — The Production
The first stop-motion feature shot entirely digitally. Corpse Bride, Warner Bros., 2005. ©Warner Bros.
01
Building the Colour Pipeline
The colour pipeline involved building a system that could process 160,000+ individual frames through a consistent colour grading workflow. The production shot on Canon EOS-1D Mark II digital SLR cameras — a still photography camera, not a motion picture camera, chosen six weeks before principal photography began when the decision was made to go digital. The raw files coming off the camera's CMOS chip were decoded using dcraw, an open-source RAW image decoder written by Dave Coffin, and processed by custom FilmLight software into Cineon images that emulated the colour space and gamma of Kodak 5248 negative — the film stock Pete Kozachik had originally selected when the project was still going to be shot on celluloid.
The solution was to use dcraw and 3D colour lookups to emulate Kodak 5248 film stock at scale. This allowed the digital frames to retain the visual language of the medium — the warmth, the grain character, the organic quality that had always defined stop-motion's look — while gaining all the practical advantages of digital origination: instant review, non-destructive adjustments, and the ability to course-correct within the same day rather than waiting for film dailies.
The on-set review system used Vinton Capture — software originally written for Will Vinton Studios specifically for stop-motion animators. The editorial database was built in FileMaker Pro, with a Python utility written to convert Final Cut Pro's XML output into a flattened reel, validating each shot for naming and length against the known shot list. Every shot was a folder of images. Every clip was treated as a reel. This was the first stop-motion feature edited in Apple Final Cut Pro and the first to choose digital cameras over film specifically on the basis of image quality.
02
Proving the Case — To Everyone
The Warner Bros. relationship was one of two cases that needed to be made simultaneously. The studio needed confidence that digital origination wouldn't lose the "soul" of the hand-crafted puppets. The camera crew needed something different — and harder to give.
The early digital tests returned images that were darker than the crew expected. Camera teams who had spent careers lighting film sets were lighting the digital sets the same way they always had — and getting results that looked underexposed by any conventional measure. Raising this directly was technically accurate but practically counterproductive. Telling an experienced camera professional that their lighting was wrong — that there was something underexposed about the result — triggered an entirely reasonable defensive response: there is nothing wrong with my lighting. I have been doing this for years. The problem is your system.
They weren't wrong about their lighting. They were right about their lighting — for film. The digital sensor read light differently. The same setup that produced a beautifully balanced film exposure could return a dark digital image, because the relationship between light levels and recorded data was fundamentally different. The solution was to flood the sets with more light, capture a flat image with as much data as possible, and then shape the look in the digital processing stage — using the colour pipeline to achieve what the camera operator's craft would previously have done on set.
That solution worked technically. It didn't work humanly. A perfectly flat, technically correct digital capture removed the camera team's creative contribution from the equation. Lighting a set became a purely scientific exercise, and that drained something essential from the production. Film-making is an art form at its core, and the camera team's craft — their ability to create atmosphere, tension, mood through light — was not incidental to the work. It was part of it.
The answer was to learn their process properly — to understand the terminology, the workflow, the decisions a camera team makes when lighting for film — and then build digital equivalents that felt familiar rather than alien. Tests were developed that matched their existing workflow as closely as possible, creating a bridge between what they knew and what the digital pipeline needed. A combination of lighting adjustments, careful work on aperture and exposure settings, and technical refinement of the RAW file processing eventually found the balance. The crew retained their creative input. The pipeline got what it needed. Most of the production went smoothly.
Not every shot was perfect. There is one technically underexposed shot in the finished film — Pastor Gaswell holding a candle — that never quite resolved the way it should have. Pete Kozachik, the colourists, and the team at Moving Picture Company found a way to make it work within the film. It is still in the movie. Some crew members chose not to continue with a digital production and moved on to other projects. Others joined specifically because they were excited by what was being built. That exchange — of people who wanted the old way for people who wanted to discover the new one — was part of how the production found its footing.
03
The Lasting Impact
When the production wrapped, we hadn’t just finished a movie; we had drafted the blueprint for an entire industry. The workflow established on Corpse Bride—RAW decoding, digital color emulation, and the FileMaker Pro editorial database—transferred directly to LAIKA for Coraline.
Beyond the US, this architecture became a highly referenced case study for the global stop-motion community. During the development of Wes Anderson's Fantastic Mr. Fox and The Pirates! when it started production at Aardman, I visited their teams to consult on their transition to a digital feature pipeline. By sharing the lessons learned on Corpse Bride and the systems we were evolving at LAIKA, we helped ensure their production could build upon a proven foundation rather than reinventing the wheel.
This cross-studio collaboration helped solidify a foundational digital capture language that would eventually pave the way for Dragonframe, now the global industry standard used on virtually every major stop-motion production worldwide.
"We hadn't just finished a movie. We had begun the standardisation of the industry."
Image — The Legacy
The digital capture workflow built on Corpse Bride transferred directly to LAIKA's subsequent productions. By ParaNorman (2012), when LAIKA adopted Dragon Stop Motion as the studio's capture system and worked closely with its developers, the foundational architecture was already in place. ©LAIKA LLC
Citations & Primary Sources
Definition Magazine — "Marriage Made in HD: Corpse Bride." Technical deep-dive covering the Canon EOS-1D Mark II pipeline, dcraw colour processing, and FileMaker Pro editorial database.
Post Magazine, September 2005 — "The Corpse Bride." Production coverage including the digital workflow, conform pipeline, and the decision to switch from film to digital SLR cameras six weeks before principal photography.
AWN / VFXWorld, September 2005 — Bill Desowitz, "Corpse Bride: Stop Motion Goes Digital." On-set report covering the digital capture system, puppet fabrication, and the collaborative relationship between production and the digital pipeline team.
dcraw — Open-source RAW image decoder written by Dave Coffin, used as the foundation of the Corpse Bride digital colour pipeline to decode Canon EOS raw files into usable Cineon images.
Dragonframe — Adopted by LAIKA from ParaNorman (2012). Developed in close partnership with the studio into the global industry standard for stop-motion animation capture. Used on Isle of Dogs, Missing Link, Pinocchio (del Toro), and Wallace & Gromit: Vengeance Most Fowl.
Knowledge transfer that compounds — the digital pipeline built on Corpse Bride became the foundation LAIKA's subsequent productions, and ultimately the industry standard, were built on
Large-scale production management: 200+ person company, multi-million dollar technology budget
Finding the human dimension of a technical problem — understanding that a flat, scientifically correct image drained something essential from the production, and that the solution had to preserve craft alongside data
Change management inside a creative organisation: building the case for technological transition with a sceptical, skilled crew whose expertise was real and whose concerns were legitimate
First-principles technical problem solving in a production environment with no existing roadmap