Open AccessAcoustic Features of the Surgut Khanty Consonants
Author(s) -
Timofey V. Timkin
Publication year - 2021
Publication title -
vestnik novosibirskogo gosudarstvennogo universiteta. seriâ: lingvistika i mežkulʹturnaâ kommunikaciâ
Language(s) - English
Resource type - Journals
ISSN - 1818-7935
DOI - 10.25205/1818-7935-2021-19-1-106-116
Subject(s) - obstruent , formant , consonant , noise (video) , locus (genetics) , acoustics , speech recognition , mathematics , computer science , physics , vowel , artificial intelligence , image (mathematics) , gene , biochemistry , chemistry
This paper deals with the acoustic features of the Surgut Khanty consonants. The research is based on the data gathered during fieldwork in Kogalym town (2018) and the Ugut village (2019). The audio samples are provided by three native speakers of the Tromjegan, Malyi Yugan, and Bolshoi Yugan idioms. The total size of the sample database numbers more than six thousand isolated consonant pronouncements. The data for the research was obtained using oscillographic and spectrographic methods, formant locus analysis, spectral moment analysis. The analysis was performed via Praat and Emu-SDMS software. Oscillograghy and spectrography methods reveal that voiceless fricative phonemes may be voiced in intervocalic distribution. It is common for the sonants to become devoiced in the final and preconsonantal positions. Moreover, due to devoicing, different phonemes may acquire low-obstruent and obstruent consonant features. For the fricative, lateral-fricative consonants, affricates spectral moment analysis has been carried out. The spectral moments technique gives an opportunity to represent complex noise data as a relatively small set of numbers that can be processed statistically. According to the data on spectral moments, four types of noise have been defined: high-frequency low-dispersion noise resembling /s/, medium-frequency low-dispersion noise resembling /ʃ/, /tʃ/, /cc/, low-frequency medium-dispersed noise for phonemes /ɫ/, / /, low-frequency dispersed noise for phonemes /w/, /γ/. The forman analysis is used o es ima e onsonan resonan frequen ies. As shown by he formant locus analysis, the smallest values of the second formant locus are associated with the labial and velar phonemes. Larger values are associated with the coronal phonemes. The largest ones are specific to the palatal phonemes. At the same time, the acoustic features make it possible to stably distinguish the nasal /n/ - /ɲ/, wherein the opposition of the middle and fron lingual ar i ula ions is observed only in some speakers’ re ordings for the pairs /ɫ/ - / /, /tʃ/ - /cc/.