Applies the Voice Analysis Toolbox on all vowel samples in a directory

The Voice Analysis Toolbox (Tsanas et al. 2011; Tsanas 2012, 2013) applies a wide range of Matlab™-implemented voice analysis procedures on a single sustained vowel (usually [a::]) and computes 339 acoustic quantities. This function calls a compiled application to compute the outcome measures from each sustained vowel recording in a directory, and collect the results into a base::data.frame.

Usage

voice_analysis(directory)

Arguments

directory

The directory where sustained vowel samples are stored. The directory needs to be writable by the user, since the compiled code will also store the results of computations there.

Value

A tibble::tibble with one row for each sustained vowel sample, and with 340 columns. The first column (listOfFiles) contains the file names of recordings. The following columns (2 to 340) then contain the measurement values for the recording.

Details

The user should be aware that applying this procedure to a directory of sound files may take 9-40 times the total duration of the sound files to perform, depending on the machine and how the application was compiled. The user should therefore make sure to capture the tibble once returned (or retrieve it immidiately from the .Last.value variable once the command completes).

Under the hood, the Voice Analysis Toolbox utilizes several other Matlab™ toolboxes which are also compiled into the runtime binary that performs the procedure. The permissive open source licences of these published external toolboxes (Little et al. 2006; Little et al. 2007; Brookes 2011) are gratefully acknowledged.

References

Brookes M (2011). “Voicebox: speech processing toolbox for MATLAB [software].” Imperial College, London.

Little M, McSharry P, Moroz I, Roberts S (2006). “Nonlinear, biophysically-informed speech pathology detection.” In 2006 IEEE International Conference on Acoustics Speech and Signal Processing Proceedings, volume 2, II--II. IEEE.

Little M, Mcsharry P, Roberts S, Costello D, Moroz I (2007). “Exploiting nonlinear recurrence and fractal scaling properties for voice disorder detection.” Nature Precedings, 1--1.

Tsanas A (2012). Accurate telemonitoring of Parkinson’s disease symptom severity using nonlinear speech signal processing and statistical machine learning. Ph.D. thesis, Oxford University, UK.

Tsanas A (2013). “Automatic objective biomarkers of neurodegenerative disorders using nonlinear speech signal processing tools.” In 8th International Workshop on Models and Analysis of Vocal Emissions for Biomedical Applications (MAVEBA), 37--40.

Tsanas A, Little MA, McSharry PE, Ramig LO (2011). “Nonlinear speech analysis algorithms mapped to a standard metric achieve clinically useful quantification of average Parkinson's disease symptom severity.” Journal of the Royal Society, Interface / the Royal Society, 8(59), 842 -- 855. doi:10.1098/rsif.2010.0456 .