RANDY LY, MBSc
Investigating The Predictability Gap Of High Dynamic Range Imaging In Environmental Simulation
Read the full Master Research Paper HERE.
This study investigates the luminance predictability gap in daylighting simulations. Building performance simulations such as energy and daylighting modeling are an integral component in designing sustainable buildings. This study investigates the difference in predicted and measured luminance to comprehend the margin's possible error using daylighting simulations. Using traditional sky models such as CIE Clear Sky can be problematic because their predictions are based on mathematical calculations based on location rather than actual sky condition. Radiance maps from high dynamic range images were compared to simulated renderings. Using RMSE (root mean square error) and MAE (Mean absolute error) were used to measure simulations' performance. One case study showed a reasonable RMSE of 26% and MAE of 15%, signifying that simulations can produce accurate predictions. However, the second case study showed an RMSE and MAE exceeding 80% due to the presence of artificial shadows.
Investigating The Predictability Gap Of High Dynamic Range Imaging In Environmental Simulation
Read the full Master Research Paper HERE.
This study investigates the luminance predictability gap in daylighting simulations. Building performance simulations such as energy and daylighting modeling are an integral component in designing sustainable buildings. This study investigates the difference in predicted and measured luminance to comprehend the margin's possible error using daylighting simulations. Using traditional sky models such as CIE Clear Sky can be problematic because their predictions are based on mathematical calculations based on location rather than actual sky condition. Radiance maps from high dynamic range images were compared to simulated renderings. Using RMSE (root mean square error) and MAE (Mean absolute error) were used to measure simulations' performance. One case study showed a reasonable RMSE of 26% and MAE of 15%, signifying that simulations can produce accurate predictions. However, the second case study showed an RMSE and MAE exceeding 80% due to the presence of artificial shadows.
