Recognition through Geometric Kaleidoscopes

Across several projects our work has involved taking something familiar, breaking it apart, and building it back together again. Similar to the constant processes happening inside every cell in your body right now.
Deconstruction and reconstruction aren’t just methods here; they’ve become part of the identity of the work itself.

At first this emerged through observation.
Later it became methodology.
Now it feels closer to a way of thinking.

This current work sits at the proposal stage, so I can’t speak directly about the final project yet. What I can talk about is the direction it has pushed the images toward: recognition built from fragments.

The focus returns to faces.

Not because they are portraits in the traditional sense, but because the face is the easiest thing for the brain to find. Even when information is incomplete, even when detail is missing, the mind insists on resolving a person.
We are designed to detect each other.

We recognise structure before accuracy.

The Halftone Idea

Rather than carefully drawing a likeness, we approached the images through what are essentially halftones.

A halftone works because of distance.
Up close you see marks.
Step back and the marks agree with each other.

The image appears almost instantly.

I’m interested in that moment when separate pieces suddenly become a face. When the marks connect, it mirrors how we process unfamiliar places and experiences: we assemble understanding quickly from partial information, and only later realise how much we invented.

A Tool, Not The Subject

To explore this, I built a small image generator.
It analyses a photograph, divides it into a grid, and rebuilds it using groups of simple shapes.

Each area is translated rather than copied:
lighter areas carry fewer marks, darker areas carry more weight.

Instead of a single dot system, the image is constructed from a shifting vocabulary; geometric forms, softer organic shapes, and irregular marks. From nearby it reads as texture. From a distance it resolves into a person.

The homemade software is not the artwork, only part of the studio process to gets whats in here (Nathan points to head) to out there, to you, and in to your head. It’s no different to adjusting levels in Photoshop or assembling elements in Illustrator. It allows repetition, variation, and small decisions to accumulate.

Multiple versions can exist from the same source image.
None are corrections.
They are parallel readings.