Entwicklung eines Modulkonzeptes für die psychoakustische Geräuschenalayse unter MatLab

Diplomarbeit (Acrobat Reader) Dezember 1998

This thesis is concerned with the evaluation and scaling of psychoacoustic perceptions. These parameters are necessary for the specification of the subjective impression of noise. They are digitally computed from calibrated data stored in sound files.
Complex algorithms are used to obtain these objective parameters. The algorithms based on different hearing models are described. A new compuation model for the sensation of roughness and fluctuation strength as well as a new method for impulse detection are presented.
The developed software-package "PAAS" (PsychoAcoustic A nalysis Software) contains the computation of the psychoacoustic parameters and a calibration tool for sound files. Furthermore, additional parameters frequently used in noise measurement applications (e.g. dB(A)) can be calculated. The algorithms are implemented in MatLab 5.1.