An effect of source-filter interaction on amplitudes of source spectrum partials
2017 (English)In: Proceedings and Report - 10th International Workshop on Models and Analysis of Vocal Emissions for Biomedical Applications, MAVEBA 2017, Firenze University Press , 2017, p. 95-98Conference paper, Published paper (Refereed)
Abstract [en]
The timbral properties of the voice are partly determined by the voice source, i.e., the pulsating glottal airflow, the properties of which are controlled by the combination of subglottal pressure, glottal adduction and other laryngeal adjustments. Its waveform, the flow glottogram, mainly reflects the amplitudes of the lowest partials. Due to source-filter interaction the lowest formants can affect the periodicity of vocal fold vibration, particularly when the first or second formant coincides with a partial. The aim of the present experimental study was to study associated spectrum effects. Glide tones performed by male singers on /ae/ or /a/ were analyzed by inverse filtering, using ripple-free closed phase as criterion. Partials coinciding with the first formant were observed to have amplitudes causing a dip in the source spectrum envelope. The sound level of a vowel is determined mainly by the strongest spectrum partial, typically the partial closest to the first formant. Glide tones obtained from the formant synthesizer MADDE, which is void of source-filter interaction, showed a much stronger sound level variation with fundamental frequency than the singer subjects. The findings thus seem relevant to the understanding of voice range profiles which show sound level versus fundamental frequency.
Place, publisher, year, edition, pages
Firenze University Press , 2017. p. 95-98
Keywords [en]
Medical applications, Natural frequencies, Formant synthesizers, Fundamental frequencies, Inverse Filtering, Range profiles, Source spectrum, Spectrum effects, Subglottal pressure, Vocal fold vibration, Spectrum analysis
National Category
Otorhinolaryngology Fluid Mechanics Musicology
Identifiers
URN: urn:nbn:se:kth:diva-302186Scopus ID: 2-s2.0-85054159625OAI: oai:DiVA.org:kth-302186DiVA, id: diva2:1597715
Conference
10th International Workshop on Models and Analysis of Vocal Emissions for Biomedical Applications, MAVEBA 2017, 13 December 2017 through 15 December 2017
Note
QC 20210927
2021-09-272021-09-272025-02-09Bibliographically approved