Machined surfaces are judged to determine machining conditions. Quantitative parameters like surface roughness and sensory parameters, such as glossiness, cloudiness, sense of discomfort and so on, of the tester, are used as the criteria. Sensory parameters can be dominant especially in judging mirror surfaces with low surface roughness. Such parameters and their representations, however, depend on each tester’s experience. Consequently, we cannot quantitatively evaluate mirror surfaces considering human sensory parameters. In this research, we aim to classify and quantify sensory parameters, and evaluate mirror surfaces quantitatively. In this report, We analyzed the appearance of the turned surface of an aluminum alloy (A2017) under different cutting conditions, focusing on ‘glossy’ and ‘cloudy’ surfaces. Two testers visually inspected the surfaces. The surface shape was measured with a surface roughness measuring device. Additionally, images of the surfaces were obtained with two different illumination configurations. Consequently, it was expected that fine irregularities on the surface influenced the surface appearance and images. Furthermore, the distribution of the light reflection coefficient was calculated from the measured shape data, based on the Beckmann’s theory on the scattering of electromagnetic waves. Consequently, the sensory evaluation, the surface images and the calculation of the light reflection coefficient provided the same trends and conclusions.