Established in 2013, RIBAJ’s Eye Line Drawing Competition gathers the year’s best architectural drawings, inviting practitioner and student entries from around the world. Entrants can submit digital or hand-drawn works, judged by a panel composed of architects, journalists, educators and RIBAJ’s deputy editor Jan-Carlos Kucharek.
This year, Bartlett Associate Professor Thomas Parker won the second prize in the competition, with ‘Post-Lenticular Baroque’ – a complex AI-generated image developed using neural radiance field modelling, which hints at future directions of travel for architectural representation.
The image, initially generated from a text prompt, dissolves eerily from the central baroque ceiling dome to a more fractal nebula of millions of pixels, through a process where neural radiance exposes “stochastic uncertainties” in the image resolution.
Judge Bongani Muchemwa said, “There’s a skill at play here that doesn’t rely on simple ChatGPT prompts – it’s a collage not of images, but of digital interrogation techniques,” while Samantha Hardingham called it “experimental, transformative and unexpected”.
"It's always a honour to be acknowledged in any drawing prize, especially for speculative work and alternative practices that are trying to find new forms of dialogue in a broad architectural landscape."
Seven Bartlett students were also among the longlisted entries recognised in the competition, announced in June:
- Jihoon Baek
- Gabe Brown
- Latisha Chan
- Sarah Chuwa
- Gabriel Edyvean Heard
- Farinoosh Hadian Jazy
- Lizhe Zhang Zhuo
Thomas Parker is an Associate Professor and The Bartlett’s Media Director. He has been a programme director for Architecture MSci (ARB/RIBA Part I & 2), and also teaches at the Architectural Association. He completed his Part 2 at The Bartlett, and is also an Architectural Design MPhil/PhD candidate.
All of the winning entries will be published in the Eye Line showcase in the July/August issue of RIBAJ.
More information
Image: A Post-Lenticular Baroque - Study Model #5. 3D scanned model, large language model image alterations, neural radiance field reconstruction and rendering; 420mm x 290mm. Credit: Thomas Parker