Talker-to-listener distance effects on speech production and perception

摘要

Simulating talker-to-listener distance (TLD) in virtual audio environments requires mimickingnatural changes in vocal effort. Studies have identified several acoustic parameters manipulated bytalkers when varying vocal effort. However, no systematic study has investigated vocal effortvariations due to TLD, under natural conditions, and their perceptual consequences. This workexamined the feasibility of varying the vocal effort cues for TLD in synthesized speech and realspeech by (a) recording and analyzing single word tokens spoken at 1 m≤TLD ≤ 32 m, (b)creating synthetic and modified speech tokens that vary in one or more acoustic parametersassociated with vocal effort, and (c) conducting perceptual tests on the reference, synthetic, andmodified tokens to identify salient cues for TLD perception. Measured changes in fundamentalfrequency, intensity, and formant frequencies of the reference tokens across TLD were similar toother reports in the literature. Perceptual experiments that asked listeners to estimate TLD showedthat TLD estimation is most accurate with real speech; however, large standard deviations in theresponses suggest that reliable judgments can only be made for gross changes in TLD.