Dawn KLCC
Dawn KLCC represents my entry into the field of Architectural Visualization. As my first project, my objective was to bypass traditional rendering limitations and explore a modern, hybrid pipeline that blends physically-based light transport with generative AI
The Search for Realism
One of the greatest challenges in ArchViz is the 'Uncanny Valley' of 3D characters. For Dawn KLCC, we implemented a first-of-its-kind workflow for our studio, using Stable Diffusion to synthesize human elements.
Unlike standard 3D assets, this generative approach allowed for hyper-realistic skin textures and clothing that react naturally to the environment's specific lighting. It bridges the gap between a 3D render and a lifestyle photograph, bringing a soul to the architectural frame.
First-Time Mastery of Unreal Engine 5
This was my first time utilizing Unreal Engine 5, and I chose to push the engine to its limits by using the Path Tracer. I prioritized light accuracy over real-time speed to capture the complex reflections of the Dawn KLCC building
Cinematic Execution
The walkthrough serves as the final realization of my storyboard. I planned the movement to create a dialogue between the real photographic elements of Malaysia and the digital architecture, using After Effect and UE5 to grade the final frames into a cohesive.
Production Credits
Creative Director: Alven Lim
Art Director: Khor Shing Nam
Render Engine: Unreal Engine 5 (Path Tracing)
Human Synthesis: Stable Diffusion (AI Integration)
Post-Production: Adobe Photoshop
Location: Kuala Lumpur, Malaysia
