Technical analysis of the «statistics» library code within the Nyquist programming language in order to improve the accuracy and efficiency of the processed audio material

Taran Vasiliy V.

The article deals with general scientific analysis of engineering and technical potential of the «Statistics» library associated with Nyquist programming language. The «Statistics» library analysis performed in the article reveals the essence of the approach to statistical processing of audio data presented by the authors of the library using the Nyquist language and its standard application programming interface — NyquistIDE. Some special aspects of the useful functions and specialized operators of the Nyquist language usage bundled with the engineering and computer practice of applied audio editing are clarified separately in the article. The application of the direct (unstable) algorithm for Pearson correlation coefficient, Welch test, Levene-Brown-Forsythe tests as the component of audio data statistical analysis is reasoned. The article also addresses the issues of ultra-fine editing of audio material using statistical methods in the Audacity® program. This is ensured by the invocation of the Nyquist programming language, which allows taking into account a wide range of acoustic nuances in order to avoid errors that occur during conversion, cleaning, mixing, mastering and remastering of audio material and also other engineering and technical practices.

Keywords:«Statistics» library, Nyquist programming language, NyquistIDE, statistical components in audio data processing, mathematical statistics, Pearson correlation coefficient, Welch test, Levene-Brown-Forsythe tests.