المفاهيم الأساسية
Voices in French audiovisual media have shown a tendency to lower in pitch over time, independent of gender, while female voices exhibit a decrease in pitch with age that is not observed in male voices.
الملخص
This study presents a diachronic acoustic analysis of 1,023 speakers from French media archives, spanning four time periods (1955/56, 1975/76, 1995/96, 2015/16) and four age groups (20-35, 36-50, 51-65, >65) for both genders.
The key findings are:
Fundamental frequency (F0) and vocal tract length (VTL) were estimated from the speech samples using a multi-algorithm approach to ensure robust measurements on the heterogeneous data.
Base-F0, calculated as the 7th decile of the F0 distribution, showed a major difference between genders (about 9 semitones) and a decrease with age for female but not male voices.
The Age:Period interaction revealed that while base-F0 was increasing with age in the 1950s, it showed a tendency to decrease with age in the later periods (1995-96 and 2015-16), independent of gender.
VTL, estimated from the first four formants, showed a significant effect of gender (1.7 cm difference between females and males) and a modest but significant increase over the time periods, with a small increase with age.
These results suggest a lowering of voice pitch in French media over time, potentially reflecting changes in social norms or vocal habits, which warrants further investigation. The methodology used to extract robust acoustic features from heterogeneous archival data provides a framework for similar diachronic studies of voice characteristics.
الإحصائيات
The fundamental frequency (F0) and the first four formants (F1-4) were estimated from the speech samples.
The base-F0 value was calculated to estimate the vocal register.
The average vocal tract length was estimated from the formant frequencies.
اقتباسات
"Vocal characteristics are an important part of identity, with our voices indicating our gender and other social constructs."
"Some studies also claimed that voices may have changed across time (for populations of comparable age); this was claimed for female voices, for example in Australia."