engleski

The Sensitivity Of Audio Distortion Measurement Methods

Distortion measurements should give us objective evaluation and comparison of different electroacoustical devices. Ordinary methods for distortion measurements give us partial information about certain devices, because used test signals don't have characteristics of natural signals &#8211 ; ; ; speech and music. Proposed combined audio test signal (CATS) consists of three sawtooth signals, and has characteristics of natural signals. Frequency components of the CATS spectrum are unevenly distributed, so intermodulation components fall between components of the original spectrum at the input of the measured system. Using peak search method intermodulation components can be extracted form original signal, and their comparison can be made. Proposed audio pseudo-random test signal is compared with other audio distortion measurement signals and methods. Comparison is made with investigations of the sensitivity by experimental measurements on electrical circuits, which simulate basic distortion mechanisms. Electrical circuits are made with combination of two operational amplifiers, resistors and zener diodes. The results show that the ordinary methods for measurements of total harmonic distortion are not relevant for dynamic distortions. Every used method has low sensitivity for at least one distortion mechanism. Correlation functions between different methods are obtained in order to predict measurement results obtained with one method from another one.

audio; distortion; measurement; methods

nije evidentirano