. Author manuscript; available in PMC: 2015 Feb 1.

Published in final edited form as: J Speech Lang Hear Res. 2014 Feb 1;57(1):26–45. doi: 10.1044/1092-4388(2013/12-0103)

TABLE 3.

Settings for formant or spectrogram analyses in the four softwares.

	Praat	Wavesurfer	TF32	CSL
File types accepted	WAV, AIFF, AIFC, NeXT/Sun AU, NIST, FLAC, MP3	WAV, Sun AU, AIFF, MP3, CSL, SD	CSpeech, NCVS92 (new CSpeech), WAV, NSP, TIMIT database (CMU and SPHERE), UW Microbeam Database compressed (ACC)	NSP, WAV, RAW
Window type with default listed first other choices in parentheses	Gaussian	Hamming (others: Hanning, Bartlett, Blackman, Rectangular)	Hamming (other: Rectangular)	Blackman (others: Rectangular, Triangular, Hamming, Hanning)
Window length	25 ms for formants 5 ms for spectrograms	49 ms	6.7 ms for wide- band 22 ms for narrow-band	10 ms
Frame interval	3 ms	10 ms	Frame = 1 pitch period (no frame overlap)	10 ms
LPC analysis method	Burg	Modified Burg	Weighted least- squares	Autocorrelation
LPC order (number of coefficients) with default and range	10 (1 – 400+)	12 (1 – 40)	26 (5 – 240)	12 (2 – 36)
Smoothing	None	Automatic	Optional	Optional