论文部分内容阅读
Traditionally, a reduction of noise can be obtained by minimizing sound pressure or sound power.Thus, the objective of design optimization of an acoustic system is to reduce physical acoustic properties, i.e., sound pressure and power.However, these parameters are not sufficient to present the relation of physical sound stimulus with human perceptual judgment.In other words, for the purpose of noise reduction considering the human subjective feelings on the sound, physical acoustic properties may not represent adequate parameters for designing acoustic devices.In this sense, it becomes challenging to consider sound quality (SQ) metrics (i.e., loudness, sharpness and roughness) as an objective function in the systematic optimization process to obtain the best possible optimized design to reduce human subjective perceptions of sound.Among these SQ metrics, Zwickers loudness has been broadly used to indicate how much louder (or softer) a sound is perceived relative to standard sound, and it has formed part of a standard metric in ISO 532B (International Organization for Standardization) and has been adopted for use in a number of ITU (International Telecommunication Union) recommendations on speech and audio quality.For the computation of Zwickers loudness, it is necessary to measure the sound pressure over wide band frequencies and it requires a filtering process in order to transform the narrow band spectral results into 1/3-octave band results which are the input data format for calculating the loudness.