Ben Carter Townsend's profile

Project # 29 NeRF scanning Luma.ai

Cast court at the V&A, London; it is my favourite room by far and this cast in particular is full of drama and narrative, it is a cast of 'Christ bearing His cross' by Hans Brüggemann.

What I love about it is that the cast is like a physical storyboard, it is a filmmakers dream; as a whole it works as a dramatic scene, but dive into the details and each character tells a story. Your eye is passed from character to character thanks to the artists masterful blocking (a filmmaking term used to describe spatial arrangement, geometry and compositional lines).

I wanted to use technology to push a camera behind the glass and get into this scene, following the sequence of 'handoffs' between characters and reveal the whole scene. 

Personally, I naturally associate these sculptures with religion and religious spaces and have become desensitised to them. I wanted to avoid that association and appreciate the scene afresh, show the scene through my personal lens, show my recognition of the artistry and relevance to my work. So I don't show christ and I don't use ecclesiastical music.

I used Luma.ai to convert my basic video into a model, allowing me to push a virtual camera through the glass and create a smooth camera move where a camera would otherwise not fit. 

Luma uses a neural radiance field (NeRF) is a neural network to reconstruct complex three-dimensional scenes from a partial set of two-dimensional images; in this case a video I took. 

On top that Luma uses a beautiful technique called gaussian splatting which allows points in space to look different from different angles; in simple terms the light adapts from different angles as you move the camera.

Here is a link to more information on the sculpture
https://collections.vam.ac.uk/item/O127720/christ-bearing-his-cross-altarpiece-sauermann-h/

Here is a link to the model
https://lumalabs.ai/capture/90F83268-AD5B-4DC8-9674-2E41214F6FE9
Project # 29 NeRF scanning Luma.ai
Published:

Project # 29 NeRF scanning Luma.ai

Published: