Aurally-adequate sound measurement technology makes use of both present psychoacoustic knowledge, e.g. loudness, roughness, fluctuation, sharpness and so forth, and Artificial Head measurement technology with transmission characteristics comparable to human hearing. By taking into account psychoacoustic evaluation parameters very often good results for judging sound events with regard to sound quality, as compared to subjective impressions, can be achieved. If a sound situation is relatively complex, i.e. if it consists of various single sound sources at different spatial positions, significant level and phase differences between the left and right ear occur which - in comparison with a monaural evaluation - can yield different results. Such effects have been observed for some time already. Speech intelligibility in a noisy environment, e.g., depends on the positions of the sound sources. Furthermore, investigations into noise in workplaces showed that binaural recordings as compared to monaural recordings caused significantly different physiological reactions of the test subjects. Moreover, it became evident that noise annoyance of single sound sources in a complex mixture of sounds does not only depend on absolute parameters such as A-weighted SPL, loudness etc., but also on their localization. In the interior of a car, for instance, the individual engine orders can produce significant amplitude and phase differences between the left and the right ear, due to wave and multipath propagation, influencing the subjective evaluation. There are no objective measurement procedures available at present. In the last decades a lot of scientific basic research was done on binaural signal processing. Due to the complexity of signal processing in human hearing there have been no complete models yet which simulate binaural signal processing in a simple way and thus constitute a basis for the objective determination of sizes, derived from a binaural measurement to describe subjectively perceived sound quality. Therefore, investigations aiming at the selection of certain signal components from a complex mixture of sounds were carried out, based on additional information such as directional sound incidence of the sound sources under test or engine r.p.m. for selecting individual engine orders. On the basis of known models to predict speech intelligibility in a noisy environment, binaural models for the objective determination of sound quality, based on these simple application examples and Artificial Head recordings, were developed.
