Acoustical properties of speech produced in noise presented through supra-aural earphones

This study determined the acoustical properties of speech known as Lombard Speech produced in background noise. Tape recordings were made for ten normally hearing adults (5 women, 5 men) reading connected speech (131 word passage "My Grandfather") at their most comfortable level in quiet and in wideband, traffic, and multitalker noise delivered through earphones at 70 and 90 dB SPL. Spectral analysis of the recordings revealed that, compared with speech in quiet, Lombard speech was characterized by: (1) an increase in overall SPL; (2) smaller vocal pitch shifts for female than male subjects; (3) shifts in spectral distributions of speech energy; and (4) the same spectral slope above 630 Hz regardless of subject gender, noise level, or noise type. Overall, the results of this study do not provide support for the theory that acoustical properties of Lombard speech are identical with loud speech produced in quiet.