Are Praat's default settings optimal for infant cry analysis?

Abstract

In recent years, the number of studies investigating possible noninvasive health screening techniques for infants have increased exponentially. Amongst those, one of the most prominent is health screening based on the acoustic investigation of infant cry. Clinicians involved in the field moved from visual inspection of the audible spectrum to automatized analysis of cry samples using computer software. A software that has been more widely adopted in recent years is Praat, a free software designed for speech analysis. Unfortunately, the software’s default settings are not suitable for investigation of cry samples, yet rarely used settings are reported in final manuscripts. In this article, we tested 4 different computer generated signals, with frequency features comparable to cry frequencies, and 3 real cry samples using both Praat’s standards and tuned settings. Our results highlight the importance of properly tuning software’s parameters when expanding their field of usage, and provide a starting point for the development of optimal Praat algorithm’s parameters selection for cry analysis