Dividuum
Concept / digital world

Dividuum

A generative music video built with German composer Benjamin Richter, where the visual world evolved alongside the tools themselves.

Because the project began before the current wave of image-editing models matured, the work had to evolve in dialogue with the technology rather than simply through a fixed pre-planned pipeline.

Back to services
Overview

Project at a glance

Dividuum sits closest to worldbuilding: a project where mood, character, sequence, and visual identity emerged gradually through experimentation and revision.

Role

Visual concept development, sequence-building, image generation, and editorial refinement for a music-video collaboration.

Brief

Create a visual world for Benjamin Richter’s track that could adapt as the available tools improved, without losing coherence or authorship.

Outcome

A music video shaped through iterative rebuilding, where improved character consistency and image-editing capability directly influenced the final sequences.

1 / 13
Case-study overview

Project notes

Project intent

In this project, the aim was to build a fully felt digital world for a music-video collaboration, allowing concept, sequence, and atmosphere to evolve together.

Production approach

The piece was built through a hybrid workflow combining MidJourney, NanoBanana, ComfyUI, and DaVinci Resolve. As better editing and consistency became possible, earlier sequences were revisited, rebuilt, and folded back into the final structure.

Visual language

In this project, the visual language is built through sequence logic as much as single frames. Character, mood, and continuity are developed across the piece so the world feels sustained rather than assembled from isolated images.

Outcome

The result is a music video shaped through repeated experimentation and rebuilding, where the evolving capabilities of the tools became part of the project’s final structure rather than an invisible background condition.