The purpose of physically-based image synthesis is to predict the natural appearance of a scenario by simulating the distribution of light using radiometric quantities. A set of spectral measurements on the film plane of the virtual camera is computed to generate a two-dimensional projection of the scene. To guarantee photometric and colorimetric consistency, the measured incident radiance distributions are reproduced exactly on the display device by reconstructing spectra with the same luminance and chromaticity. Although a significant number of predictive rendering systems have been presented in the past, there is still a lack of a comprehensive model incorporating local reflection, light transport, measurement, and reproduction. Furthermore, despite the fact that some of these frameworks have been validated in the last decades, trustworthy reproduction of color and luminance is a challenging field of research. The mathematical model to simulate and reproduce physically-based images while preserving photometric and colorimetric consistency is summarized. As a proof of concept, we integrated the model into a ray tracing system to generate photorealistic images. To validate our approach, a virtual scene and a real model of a welldefined box scenario were built and evaluated by photometric and colorimetric measurements. The scenario included a GretagMacbeth ColorChecker to allow an accurate verification of both luminance and chromaticity values. The illumination was simulated using spectral path tracing and radiometric quantities for the light source and the materials. The projected image was reproduced on a colorimetrically characterized LC device. Our results show that the 1976 CIELAB differences between the measurements of the ColorChecker patches in the real world scene and on the monitor displaying the reproduction of our simulation are well below the Just Noticeable Difference threshold.