PRACTICAL APPLICATIONS OF DIGITAL SIGNAL PROCESSING THEORY

摘要

The objective of this paper is to illustrate software-based methods for post-processing of acoustics and vibration data. The author explores Parseval's Theorem as it pertains to the equivalency of energy in time- and frequency-domain data, the application of signal coherence to improved estimates of transfer functions, and the synthesis of proportional bandwidth data from fixed-resolution data. Firstly, about Parseval's Theorem, the author has noted that this theory is necessary but not sufficient to ensure that frequency-domain estimates of amplitudes/logarithmic levels for WAV files are accurately determined. A little-known correction factor has been utilized to improve the agreement with time-domain data. Secondly, the author explores the improvement of the accuracy of transfer functions through a system by the mining of coherence data in a manner that accounts for the different proportions of coherent and incoherent data that can result from signal analysis. This is of practical importance when estimating the coherent signals that pass through a sound or vibration system and allows for more accurate estimates of the transfer functions that manifest themselves in practice. Thirdly, the author explores methods for synthesizing proportional bandwidth data from narrowband, fixed resolution frequency information using a lever law assessment of the frequency-domain energy. Even in the absence of time-domain data, such post-processing allows for the alternate reporting of data in the frequency domain in either narrowband, fixed resolution or proportional bandwidth methods of analysis.