Frequency Analysis of Sound Broadcasting Signal Parameters

The purpose of the work is to develop algorithms for frequency analysis of spectral, rhythmic, cepstral and spectral parameters of a sound broadcasting signal (SBS), which provide consistent, with the perception characteristics of a peripheral human auditory analyzer, results and allow assessing the quality of transmission of the sound-broadcasting signal by a real signal. Traditional methods of spectral analysis of the sound-broadcasting signal, at short time intervals commensurate with individual sounds, do not provide acceptable accuracy and resolution capability. The requirements for accuracy and resolution capability of spectral analysis consistent with the capabilities of the human peripheral auditory analyzer are justified. Algorithm of signal analysis based on discrete cosine transformation (DCT) of complex represented SBS is proposed with simultaneous generation of additional compensating signal, which allows to eliminate side petals of estimation. Subsequent analysis of multiple transposed SBS allows us to bring the score closer to the accuracy of the human auditory analyzer. The resolution capability of the analysis is twice of the discrete Fourier transform (DFT). Examples of practical use of developed algorithms are given.