Acoustic Measures of Articulation and Vocal Quality in Transgender People Completing Vocal Feminization Therapy

Acoustic outcomes documented following gender-affirming voice and communication therapy (GAVT) for vocal feminization often include measures of vocal pitch and resonance; however, it is likely that other aspects of speech and voice change following successful voice therapy. Therefore, we investigated how additional articulatory and vocal quality (spectral, cepstral) measures changed pre- to post-feminizing GAVT.

A retrospective chart review yielded 16 clients (15 transgender women, one nonbinary and gender queer; age = 31.0 ± 9.4 years) who completed vocal feminization therapy. Acoustic recordings of sustained vowels and sentences (eg, rainbow passage) were captured pre and post therapy. Acoustic measures related to pitch (fundamental frequency [f_o]), articulation (formant [F] 1-4, vocal tract length, vowel articulation index, and voice onset time), and vocal quality (cepstral peak prominence [CPP], low-to-high [L/H] spectral ratio, and harmonics-to-noise ratio [HNR]), were extracted and compared pre/post therapy.

As expected, mean f_o increased an average of 5.4 semitones following therapy. F2 also significantly increased, indicating a more forward tongue position during vowel productions. L/H ratio significantly increased, but there was no change in HNR or CPP, indicating a shift in high-frequency harmonic content but no increase in vocal breathiness. No other vocal quality or articulatory changes were found.

Besides an increase in F2, the present study did not support articulatory changes following successful GAVT. The study did endorse spectral shifts toward higher frequencies, but more research is needed to determine how those relate to perceived vocal quality changes.