Authenticity of the simulation of room acoustics is evaluated by comparing auralizations with real recordings. Impulse responses are recorded in two concert halls with 34 loudspeakers positioned on the stage in the shape of an orchestra and a 3D microphone probe for spatial reproduction. The acoustics of the same concert halls are simulated with acoustics modeling software. B-format impulse responses are calculated by using the identical source and receiver positions as in the real halls. Additionally, two processing methods are applied to the simulated responses in order to decrease the difference in acoustical impression. Objective room acoustic parameters between the real and simulated halls are compared, and a listening test utilizing convolutions with anechoic orchestral music is organized. The results suggest that the subjects can be categorized in two groups having preference of brightness or naturalness. Depending on the music style, auralizations with processed responses are assessed equal or better than the real hall in terms of instrument balance and brightness.