CamSounder - The Movie

A huge thank you to Hege Hammersland for fantastic energy and the chance to work together again — on a project that lit up every creative spark we had. An incredibly fun deliverable that felt great from start to finish.

CamSounder is an underwater camera and AI system from Scantrol Deep Vision, mounted directly in the trawl — it photographs and identifies the catch as it passes through the net, automatically measures size, and sends catch data straight to the skipper on the bridge while the tow is still underway. Set to be shown live for the first time at Nor-Fishing, Scantrol Deep Vision wanted to tell the story from an unexpected angle: through two shrimp and a mackerel who, in the middle of the trawl, suddenly discover they're being watched. We've had an incredibly fun and close collaboration with Hege throughout the process, with quick feedback and a real openness to letting the concept grow along the way.

Bycatch and imprecise sorting are a real problem in trawl fishing, and a complex AI product had to be communicated in under a week — to an audience meeting it on a trade show floor, not in front of a screen with sound on. The film therefore had to work entirely visually, without a single word of voice-over, while still conveying a genuinely technical product in a way that's both funny and credible.

We didn't start in an AI tool, but with classic design work: moodboarding and stylescaping to find the right visual expression, and a series of direction studies to identify a look that struck the right balance — humorous and warm, yet still clear enough to highlight the actual capabilities of the technology Scantrol Deep Vision has developed. Only once this style was locked — soft, cinematic 3D renders, expressive characters — did we bring in AI as a production tool, building out each character as consistent reference sheets, ready for reuse throughout production.

With the style locked, we built a complete storyboard scene by scene, in close dialogue with Hege — including deliberate adjustments to the original script along the way, letting the humor and timing find their own shape rather than being bound to a fixed script.

Each scene was built as a series of locked still frames before we touched video — including clean base plates for the on-screen graphics, composited separately afterward in Affinity Designer and Affinity Photo to guarantee sharp, precise graphics. Video was then generated and tested across several AI video models to find the right tool for the right type of motion.

We composed an original instrumental score for the film, and designed distinct, characterful voices for each figure — so every character is instantly recognizable by voice alone, regardless of visible mouth movement.

All layers — dialogue, music, ambient sound and on-screen graphics — were assembled and polished in Final Cut Pro, with a clear distinction in soundscape and lighting between the above- and below-water scenes.