Appreciation of Proportion in Architecture: A Comparison Between Facades Primed in Virtual Reality and on Paper

Abstract

Virtual reality (VR) is a rapidly emerging tool in design and architecture. VR allows to define and evaluate design equations at full scale providing the user a sense of depth, volume and distances, as opposed to drawings, plans and CAD designs displayed on screen. Currently, VR enables to provide subjects with a preliminary experience of their own body and relative position towards the virtually designed space. That real time virtual body experience is expected to become more realistic the forthcoming decade, given the uprising of motion tracking and posture generating algorithms, and the expected breakthrough of augmented reality. These evolutions will make VR a tool that is complementary to design and simulation of body near products with 3D anthropometry and rapid prototyping such as 3D printing. Whereas the latter allows simulation of interaction with artifacts at scale of the human body or smaller, VR will enable to simulate and evaluate structures that surrounds the human body: rooms, spaces, buildings and urban implants. In particular, evaluation of scale with relation to human proportion is expected to benefit from VR. The present research is a first step in comparing classical drawing techniques with VR. This paper is confined to evaluate the aesthetic preference of subjects when exposed to facades in a virtual urban planning environment. The preference for a certain proportion is evaluated by exposing subjects to the same facades in both mediums. In this first pilot study, we isolated one variable: the length to width ratio of windows in the facade of a house with a flat roof. Subjects were primed with three length to width ratios: the square root of 2, the golden (in approximation 1.618) section and a value that was intuitively designed to be most pleasing, turned out to be 1.80. A total of n = 30 subjects were enrolled in this study. Only descriptive statistics were used due to small sample size. Results were interpreted as tendencies. There were no differences found in preferences when exposed to drawings compared to when exposed to the same models in VR. These results indicate that-restricted to our simple and controlled setting- evaluation in VR coincides with regular evaluation techniques, and thus could extend the latter, not only to relative proportion, but also in relation to subject’s body dimensions as a sense of absolute scale. The test setting presented in this pilot can be used to plan large scale-controlled studies to further investigate this hypothesis.