Details
Year: 2026
Software:
This project explores a new way of presenting architecture using static 3D renderings, transforming them into an audiovisual piece through artificial intelligence.
The video develops a narrative centered on outdoor spaces and swimming pools, incorporating movement in the water, vegetation, people, changes in light, and small everyday gestures that help convey how each space might feel and be experienced.
The goal is not to replace architectural renderings, but to expand their possibilities: to move from a still image to a visual experience with a greater capacity to convey atmosphere, lifestyle, and emotion.
A combination of 3D architectural visualization, art direction, and AI-generated video applied to architecture, real estate development, and visual communication.
Translated with DeepL.com (free version)
After initial tests focused on generating lifestyle scenes, we continued to explore a new goal: building narrative sequences while maintaining visual consistency between shots.
The research has focused on two key aspects. The first is the creation of micro-sequences by combining wide, medium, and close-up shots. The second is character continuity, maintaining their identity as the scene evolves.
Maintaining character consistency remains one of the main challenges of AI-generated video. Changes in framing, camera distance, or interaction with the environment can alter the characters’ appearance and disrupt the continuity of the sequence.
These tests are part of a research process aimed at developing new workflows for architectural visualization. The goal is to combine 3D rendering, audiovisual storytelling, and artificial intelligence to create more immersive content capable of conveying not only the aesthetics of a space but also the experience of inhabiting it.
The integration of architectural visualization, animation, and artificial intelligence opens up new possibilities for creating presentations, promotional content, and visual experiences applied to architecture and design.
Translated with DeepL.com (free version)