Traditionally, concert hall acoustics is evaluated by listening to live concerts, which makes a direct comparison challenging. This thesis presents new tools and methods in the domain of the room acoustics evaluation, studies, and auralization. Auralization stands for the process of rendering an existing or modeled acoustic space in a way that it can be presented to the listener as he/she was listening to a sound inside the space under study. An essential topic in this thesis is a framework for studying room acoustics with a wide-area loudspeaker array. The proposed loudspeaker orchestra consists of a number of loudspeakers that are positioned in the shape resembling a symphony orchestra on a stage. The acoustics can be evaluated in-situ by playing back anechoic signals, or in laboratory conditions via convolution of the impulse responses measured from the loudspeaker orchestra. The presented method enables a direct comparison of concert halls and it has been successfully applied in practice in several research articles. The principal requirement for such a loudspeaker orchestra is anechoic signals of high quality. For this purpose, a method and implementation of a system for recording the symphony orchestra instruments individually is presented. As the result, a selection of anechoic orchestral music is obtained with perfect channel separation. The recordings, intended for advancing the research on acoustics and auralization, are published for academic use. Directivity of the orchestra instruments in performance situation is investigated with anechoic measurements. The results for different instruments can be compared against each other or applied directly into auralizations. Data from the directivity measurements is also applied in the objective analysis of the presented loudspeaker orchestra. Furthermore, the implemented measurement system is utilized in investigating the sound radiation of the balloons, which are often used in room acoustic measurements. Related to the anechoic recordings, a novel approach to creating an impression of a group of musicians from a single recorded player is proposed. The method is mainly based on the video and audio analysis of the temporal differences between orchestra string players. The method is particularly beneficial with the anechoic recordings, where recording an instrument section is not possible, and recording a large number of musicians individually is time-consuming. The listening test results show that the presented method provides a plausible simulation of an instrument section sound in comparison to an industry-standard method.