Pretreatment Acoustic Predictors of Gender, Femininity, and Naturalness Ratings in Individuals With Male-to-Female Gender Identity

Exploring the Relationship Between Acoustic Measurements and Self-Perception of Voice in Trans Women

OBJECTIVE

This study aimed to explore the strength and direction of the relationship between spectral cepstral-based, time-based acoustic measures and the self-perception of voice in trans women.

METHODS

Forty-eight trans women were included in the study. Analysis of the sustained vowel phonation was performed using Multidimensional Voice Profile Analysis (MDVP), and spectral-cepstral analyses of the sustained vowel phonation, all-voiced weighted sentence, and spontaneous speech were made via Analysis of Dysphonia in Speech and Voice (ADSV) software. For self-perceptual evaluations, the Trans Woman Voice Questionnaire (TWVQ) and the Self-perception of Voice Femininity Scale (SPVF) were used. The correlation between MDVP, spectral-cepstral parameters, and TWVQ and SPVF scores was calculated.

RESULTS

The present study found a positive relationship between F0, SPVF, and TWVQ. Among the perturbation parameters, the jitter was the only one found to correlate with SPVF and TWVQ. The CPPF0 parameter was found to be associated with a more feminine voice perception and a higher voice-related quality of life in all speech samples in the present study. In addition, higher CPP values achieved from vowel phonation were associated with less feminine voice perception and lower voice-related quality of life. The present study also suggests a weak correlation with the SPVF and Cepstral Peak Prominence Standard Deviation (CPPF0 SD) of the spontaneous speech sample in a negative direction.

CONCLUSIONS

This study found weak and moderate levels of correlations between F0, jitter (%), CPP, CPPF0, CPPF0 SD parameters, and self-perceptual measures. These findings suggested that such a level of relationship is attributable to the fact that these tools evaluate different aspects of voice in accordance with the International Classification of Functioning System. According to this pioneering study, it would be beneficial to incorporate spectral-cepstral measures into the objective assessment protocol for trans women's voices.

Incorporating Mobile App Technology in Voice Modification Protocol for Transgender Women

PURPOSE

Motivated by practice and feedback principles of motor learning, a hybrid clinic-home protocol for voice feminization was developed to minimize the role of SLPs to 1 of supervision and professional guidance and to maximize learning during independent practice apart from intervention sessions. The purpose was to explore the effectiveness and acceptability of the innovative service delivery.

METHOD

This single-subject changing criterion design included four transgender women who completed a 10-week hybrid clinic-home voice intervention program delivered via 30 -minute weekly in-clinic sessions and a technology-supported home program. The program was client-centered and capitalized on principles of motor learning in that it incorporated frequent practice with intermittent, knowledge-of-result feedback. Participants' desired outcomes were measured using acoustics, self and listener ratings of audio samples, and a program evaluation questionnaire.

RESULTS

Average speaking fundamental frequency of phrases and picture descriptions gradually increased into the 170-220 Hz range for all except one participant. All four transgender women were perceived to sound more feminine following treatment compared to baseline. Participants found the in-clinic sessions useful, the app easy to use, and noted limited fatigue or discomfort.

CONCLUSION

Four transwomen met their goals using this hybrid clinic-home service delivery format. Further investigations may elucidate key factors of the success achieved in the current study by designing comparison delivery models and including people from other populations.

Sequential Wendler Glottoplasty and Laser Reduction Glottoplasty for Voice Feminization

OBJECTIVE

Voice feminizing surgery is frequently needed for transgender female patients. Among several surgical options, Wendler glottoplasty (WG) and laser reduction glottoplasty (LRG) are two endoscopic procedures. However, because a single procedure may not produce sufficient benefit, the two surgeries may sometimes be sequentially performed. This study was carried out to present the voice results of such sequential surgeries.

METHODS

This is an individual retrospective cohort study, performed at a tertiary referral center, that is a university hospital. 18 transgender patients were treated with WG initially and then underwent LRG; 17 had LRG first then WG. All 35 cases were performed during a 15-year period and followed for at least 1 year postoperatively. Voice Handicap Index (VHI-30), transsexual voice questionnaire (TVQ), and acoustic analysis with /a/ and running speech were obtained pre- and postoperatively.

RESULTS

VHI and TVQ improved significantly postoperatively (p < 0.05). Their preoperative, first, and second postoperative mean sF0 were 146, 175, and 215 Hz, respectively; these differences were statistically significant (p < 0.001). Their postoperative mean jitter percent, shimmer percent, noise to harmonic ratio (NHR), cepstral peak prominence (CPP), and cepstral spectral index of dysphonia (CSID) worsened significantly compared to preop values (p < 0.05); however, mean postoperative acoustic results were still within normal limits. Patients' self-ratings of their postsurgery voices revealed all feminine, leading to a patient gratification score of 100%.

CONCLUSION

If transgender female patients are unsatisfied with their voice after WG or LRG, the addition of the alternative procedure may significantly feminize their voice. Sequential WG and LRG is a successful surgical option for voice feminization.

LEVEL OF EVIDENCE

4 Laryngoscope, 134:1133-1138, 2024.

Auditory-Perceptual Assessment and Acoustic Analysis of Gender Expression in the Voice

OBJECTIVE

Determine if acoustic measurements exist that are predictive of Auditory-Perceptual Assessment (APA) of gender expression in the voice of transgender, nonbinary, and cisgender Brazilian speakers by transgender, nonbinary, and cisgender judges, as well as speech-language pathologists in the area of voice studies.

METHODS

Cross-sectional study. Clips of speech (automatic speech and expressive reading of poetry) and sustained vowel emission of people of different genders were recorded and underwent APA for gender expression in the voice using a visual analog scale across 100 points, ranging from very masculine to very feminine. Sixteen acoustic measurements were extracted (noise, perturbation, spectral, and cepstral measurements). A descriptive and inferential analysis was performed using interclass coefficients of correlation and stepwise multiple linear regression, considering P < 0.05 for statistical significance.

RESULTS

Forty-seven people of different genders had their voices recorded. The perceived gender of these voices was judged by 236 people (65 speech-language pathologists, 101 cisgender people, and 70 transgender and nonbinary people). The perceptions and measurements that were predictive of gender perception in the voice differed according to the task (vowel or speech) and the group of judges. The predictive acoustic measurements that were common in all groups were: speech-median F0, harmonic-to-noise ratio (HNR), F0 standard deviation (F0sd), average width between F0 peaks, and spectral emphasis (Emph); vowels-median F0, HNR, F0sd, and average width between F0 peaks. Divergent measurements between groups were: speech-coefficient of variation of intensity, speech rate (Sr), minimum and maximum F0, jitter, and shimmer; vowels-coefficient of variation of intensity, Emph, Sr, and minimum and maximum F0.

CONCLUSION

There are acoustic measures that may predict APA; however, each group of judges considers different measures to evaluate gender, revealing an important influence of context on the evaluator in gender assessment through the voice.

Auditory free classification of gender diverse speakersa)

Auditory attribution of speaker gender has historically been assumed to operate within a binary framework. The prevalence of gender diversity and its associated sociophonetic variability motivates an examination of how listeners perceptually represent these diverse voices. Utterances from 30 transgender (1 agender individual, 15 non-binary individuals, 7 transgender men, and 7 transgender women) and 30 cisgender (15 men and 15 women) speakers were used in an auditory free classification paradigm, in which cisgender listeners classified the speakers on perceived general similarity and gender identity. Multidimensional scaling of listeners' classifications revealed two-dimensional solutions as the best fit for general similarity classifications. The first dimension was interpreted as masculinity/femininity, where listeners organized speakers from high to low fundamental frequency and first formant frequency. The second was interpreted as gender prototypicality, where listeners separated speakers with fundamental frequency and first formant frequency at upper and lower extreme values from more intermediate values. Listeners' classifications for gender identity collapsed into a one-dimensional space interpreted as masculinity/femininity. Results suggest that listeners engage in fine-grained analysis of speaker gender that cannot be adequately captured by a gender dichotomy. Further, varying terminology used in instructions may bias listeners' gender judgements.

Effect of Endoscopic Glottoplasty on Acoustic Measures and Quality of Voice: A Systematic Review and Meta-Analysis

INTRODUCTION

The voice is an important marker of the transition process to the new gender identity of the transgender person. For 20% of patients seeking voice feminization, voice therapy is not completely satisfactory, and surgery should be considered, with endoscopic glottoplasty being the current practice. While the increase in fundamental frequency (F0) after glottoplasty has been well described, no systematic review of the literature or meta-analysis regarding other acoustic parameters and quality of voice has been performed yet.

OBJECTIVE

To define the effect of endoscopic glottoplasty on acoustic measures and quality of voice by assessing F0, maximum phonation time (MPT), frequency range, and grade of dysphonia.

METHODS

A literature review was performed in Medline/PubMed, Cochrane, Science Direct, LILACS, and Google Scholar, following PRISMA guidelines, with no constraints on publication date. We included studies in English, Portuguese or Spanish that assessed transgender women who underwent endoscopic glottoplasty. All the included articles were measured in terms of their methodological quality.

RESULTS

The PRISMA approach yielded 14 studies, totaling 566 patients. There was significant heterogeneity between studies regarding follow-up time, surgical technique and perioperative care. Thirteen studies were submitted to meta-analysis. The results showed significant changes in pre- to postglottoplasty mean differences of F0 = 78.49 Hz (95%CI: 75.69-81.30), MPT = -1.11 seconds (95%CI -1.67 to -0.54), frequency range = -3.55 semitones (95%CI -5.74 to -1.36) and grade of dysphonia on the GRBAS scale = 0.44 (95%CI 0.27-0.61).

CONCLUSION

Glottoplasty is effective in significantly increasing fundamental frequency, but slightly decreases MPT, frequency range and vocal quality measured by the grade of dysphonia on the GRBAS scale.

Effect of Anchor Term on Auditory-Perceptual Ratings of Feminine and Masculine Speakers

BACKGROUND

Studies investigating auditory perception of gender expression vary greatly in the specific terms applied to gender expression in rating scales.

PURPOSE

This study examined the effects of different anchor terms on listeners' auditory perceptions of gender expression in phonated and whispered speech. Additionally, token and speaker cues were examined to identify predictors of the auditory-perceptual ratings.

METHOD

Inexperienced listeners (n = 105) completed an online rating study in which they were asked to use one of five visual analog scales (VASs) to rate cis men, cis women, and transfeminine speakers in both phonated and whispered speech. The VASs varied by anchor term (very female/very male, feminine/masculine, feminine female/masculine male, very feminine/not at all feminine, and not at all masculine/very masculine).

RESULTS

Linear mixed-effects models revealed significant two-way interactions of gender expression by anchor term and gender expression by condition. In general, the feminine female/masculine male scale resulted in the most extreme ratings (closest to the end points), and the feminine/masculine scale resulted in the most central ratings. As expected, for all speakers, whispered speech was rated more centrally than phonated speech. Additionally, ratings of phonated speech were predicted by mean fundamental frequency (f _o) within each speaker group and by smoothed cepstral peak prominence in cisgender speakers. In contrast, ratings of whispered speech, which lacks an f _o, were predicted by indicators of vocal tract resonance (second formant and speaker height).

CONCLUSIONS

The current results indicate that differences in the terms applied to rating scales limit generalization of results across studies. Identifying the patterns across listener ratings of gender expression provide a rationale for researchers and clinicians when making choices about terms. Additionally, beyond f _o and vocal tract resonance, predictors of listener ratings vary based on the anchor terms used to describe gender expression.

SUPPLEMENTAL MATERIAL

https://doi.org/10.23641/asha.19617564.

Revisiting the acoustics of speaker gender perception: A gender expansive perspective

Examinations of speaker gender perception have primarily focused on the roles of fundamental frequency (f_o) and formant frequencies from structured speech tasks using cisgender speakers. Yet, there is evidence to suggest that f_o and formants do not fully account for listeners' perceptual judgements of gender, particularly from connected speech. This study investigated the perceptual importance of f_o, formant frequencies, articulation, and intonation in listeners' judgements of gender identity and masculinity/femininity from spontaneous speech from cisgender male and female speakers as well as transfeminine and transmasculine speakers. Stimuli were spontaneous speech samples from 12 speakers who are cisgender (6 female and 6 male) and 12 speakers who are transgender (6 transfeminine and 6 transmasculine). Listeners performed a two-alternative forced choice (2AFC) gender identification task and masculinity/femininity rating task in two experiments that manipulated which acoustic cues were available. Experiment 1 confirmed that f_o and formant frequency manipulations were insufficient to alter listener judgements across all speakers. Experiment 2 demonstrated that articulatory cues had greater weighting than intonation cues on the listeners' judgements when the f_o and formant frequencies were in a gender ambiguous range. These findings counter the assumptions that f_o and formant manipulations are sufficient to effectively alter perceived speaker gender.

Associations Between Voice and Gestural Characteristics of Transgender Women and Self-Rated Femininity, Satisfaction, and Quality of Life

Purpose Client-based subjective ratings of treatment and outcomes are becoming increasingly important as speech-language pathologists embrace client-centered care practices. Of particular interest is the value in understanding how these ratings are related to aspects of gender-affirming voice and communication training programs for transgender and gender-diverse individuals. The purpose of this observational study was to explore relationships between acoustic and gestural communication variables and communicator-rated subjective measures of femininity, communication satisfaction, and quality of life (QoL) among transfeminine communicators. Method Twelve acoustic and gestural variables were measured from high-fidelity audio and motion capture recordings of transgender women (n = 20) retelling the story of a short cartoon. The participants also completed a set of subjective ratings using a series of Likert-type rating scales, a generic QoL questionnaire, and a population-specific voice-related QoL questionnaire. Correlational analyses were used to identify relationships between the communication measures and subjective ratings. Results A significant negative relationship was identified between the use of palm-up hand gestures and self-rated satisfaction with overall communication. The acoustic variable of average semitone range was positively correlated with overall QoL. No acoustic measures were significantly correlated with voice-related QoL, and unlike previous studies, speaking fundamental frequency was not associated with any of the subjective ratings. Conclusions The results from this study suggest that voice characteristics may have limited association with communicator-rated subjective measures of communication satisfaction or QoL for this population. Results also provide preliminary evidence for the importance of nonverbal communication targets in gender-affirming voice and communication training programs.

Laser Reduction Glottoplasty for Voice Feminization: Experience on 28 Patients

OBJECTIVES

Laser reduction glottoplasty is a relatively new surgical procedure for voice feminization on transgender women. This study aims to determine long-term voice results of glottoplasty on transwomen.

METHODS

Nonrandomized, retrospective, cohort. Tertiary referral center. Endoscopic laser reduction glottoplasty was performed on 28 transwomen. Voice Handicap Index (VHI-30), Transsexual Voice Questionnaire (TVQ), acoustic analysis with /a/ for F₀, jitter, shimmer, noise to harmonic ratio and acoustic analysis for speaking F₀ were measured before and after surgery. Patients self-evaluated their postoperative voices and medical students and 2 voice experts scored patients' pre- and postsurgery voice samples as masculine, feminine or neither.

RESULTS

Mean total VHI and TVQ scores improved significantly postoperatively (P < .001). Pre- and postsurgery mean F₀ were 132 and 198 Hz and mean speaking F₀ were 123 and 194 Hz, respectively; these variations were found statistically significant (P < .001). Postoperative mean jitter, shimmer and NHR increased significantly compared to preoperative values (P < .05). Nine patients (32%) were not happy with their postsurgery voice result and were offered anterior glottic web formation as secondary procedure. MFT women's self-ratings of their postsurgery voices showed 3 masculine, 19 feminine and 6 neither outcomes, leading to patient gratification score of 68%. Medical students evaluated 79% of postsurgery voice specimens as feminine. Voice experts evaluated 75% of postsurgery voice specimens as feminine.

CONCLUSIONS

Laser reduction glottoplasty is an accomplished and satisfying operation for feminizing voice of transwomen. Its voice outcome appears to be durable for 5 years. However, secondary operation may be needed to further gratify transwomen.

Perceptual Evaluation of Speech Naturalness in Speakers of Varying Gender Identities

Purpose The purpose of this study was to investigate how speech naturalness relates to masculinity-femininity and gender identification (accuracy and reaction time) for cisgender male and female speakers as well as transmasculine and transfeminine speakers. Method Stimuli included spontaneous speech samples from 20 speakers who are transgender (10 transmasculine and 10 transfeminine) and 20 speakers who are cisgender (10 male and 10 female). Fifty-two listeners completed three tasks: a two-alternative forced-choice gender identification task, a speech naturalness rating task, and a masculinity/femininity rating task. Results Transfeminine and transmasculine speakers were rated as significantly less natural sounding than cisgender speakers. Speakers rated as less natural took longer to identify and were identified less accurately in the gender identification task; furthermore, they were rated as less prototypically masculine/feminine. Conclusions Perceptual speech naturalness for both transfeminine and transmasculine speakers is strongly associated with gender cues in spontaneous speech. Training to align a speaker's voice with their gender identity may concurrently improve perceptual speech naturalness. Supplemental Material https://doi.org/10.23641/asha.12543158.

Contributions of Voice and Nonverbal Communication to Perceived Masculinity-Femininity for Cisgender and Transgender Communicators

Purpose The purpose of this study was twofold: (a) to identify a set of communication-based predictors (including both acoustic and gestural variables) of masculinity-femininity ratings and (b) to explore differences in ratings between audio and audiovisual presentation modes for transgender and cisgender communicators. Method The voices and gestures of a group of cisgender men and women (n = 10 of each) and transgender women (n = 20) communicators were recorded while they recounted the story of a cartoon using acoustic and motion capture recording systems. A total of 17 acoustic and gestural variables were measured from these recordings. A group of observers (n = 20) rated each communicator's masculinity-femininity based on 30- to 45-s samples of the cartoon description presented in three modes: audio, visual, and audio visual. Visual and audiovisual stimuli contained point light displays standardized for size. Ratings were made using a direct magnitude estimation scale without modulus. Communication-based predictors of masculinity-femininity ratings were identified using multiple regression, and analysis of variance was used to determine the effect of presentation mode on perceptual ratings. Results Fundamental frequency, average vowel formant, and sound pressure level were identified as significant predictors of masculinity-femininity ratings for these communicators. Communicators were rated significantly more feminine in the audio than the audiovisual mode and unreliably in the visual-only mode. Conclusions Both study purposes were met. Results support continued emphasis on fundamental frequency and vocal tract resonance in voice and communication modification training with transgender individuals and provide evidence for the potential benefit of modifying sound pressure level, especially when a masculine presentation is desired.

Acoustic Features of Transfeminine Voices and Perceptions of Voice Femininity

The purpose of this study was to evaluate the relationships between acoustic measures of transfeminine voices and both self- and listener ratings of voice femininity. Connected speech samples were collected from 12 transfeminine individuals (M = 36.3 years, SD = 10.6 years) and a control group of five cisgender (cis) women and five cis men (M = 35.3 years, SD = 13.3 years). The acoustic measures of fundamental frequency (f_o), f_o variation, formant frequencies, and vocal intensity were calculated from these samples. Transfeminine speakers rated their own voices on a five-point scale of voice femininity. Twenty inexperienced listeners heard an excerpt of each speech sample and rated the voices on the same five-point scale of voice femininity. Spearman's rank-order correlation coefficients were calculated to measure the relationships between the acoustic variables and ratings of voice femininity. Significant positive correlations were found between f_o and both self-ratings (r = 0.712, P = 0.009) and listener ratings of voice femininity (r = 0.513, P < 0.001). Significant positive correlations were found between intensity and both self-ratings (r = 0.584, P = 0.046) and listener ratings of voice femininity (r = 0.584, P = 0.046). No significant correlations were found between f_o variation or formant frequencies and perceptual ratings of voice femininity. A Pearson's chi-square test of independence showed that the distribution of self- and listener ratings differed significantly (χ² = 9.668, P = 0.046). Self- and listener ratings were also shown to be strongly correlated (r = 0.912, P < 0.001). This study provides further evidence to support the selection of training targets in voice feminization programs for transfeminine individuals and promotes the use of self-ratings of voice as an important outcome measure.

Voice, Articulation, and Prosody Contribute to Listener Perceptions of Speaker Gender: A Systematic Review and Meta-Analysis

PURPOSE

The aim of this study was to provide a systematic review of the aspects of verbal communication contributing to listener perceptions of speaker gender with a view to providing clinicians with guidance for the selection of the training goals when working with transsexual individuals.

METHOD

Preferred reporting items for systematic review and meta-analysis protocols (PRISMA-P) guidelines were adopted in this systematic review. Studies evaluating the contribution of aspects of verbal communication to listener perceptions of speaker gender were rated against a new risk of bias assessment tool. Relevant data were extracted, and narrative synthesis was then conducted. Meta-analyses were conducted when appropriate data were available.

RESULTS

Thirty-eight articles met the eligibility criteria. Meta-analysis showed speaking fundamental frequency contributing to 41.6% of the variance in gender perception. Auditory-perceptual and acoustic measures of pitch, resonance, loudness, articulation, and intonation were found to be associated with listeners' perceptions of speaker gender. Tempo and stress were not significantly associated. Mixed findings were found as to the contribution of a breathy voice quality to gender perception. Nonetheless, there exists significant risk of bias in this body of research.

CONCLUSIONS

Speech and language clinicians working with transsexual individuals may use the results of this review for goal setting. Further research is required to redress the significant risk of bias.

Trans Male Voice in the First Year of Testosterone Therapy: Make No Assumptions

PURPOSE

The purpose of this study was to prospectively examine changes in gender-related voice domain of pitch measured by fundamental frequency, function-related domains of vocal quality, range, and habitual pitch level and the self-perceptions of transmasculine people during their first year of testosterone treatment.

METHOD

Seven trans men received 2 voice assessments at baseline and 1 assessment at 3, 6, 9, and 12 months after starting treatment.

RESULTS

Vocal quality measures varied between and within participants but were generally within normal limits throughout the year. Mean fundamental frequency (MF0) during reading decreased, although to variable extents and rates. Phonation frequency range shifted down the scale, although it increased in some participants and decreased in others. Considering MF0 and phonation frequency range together in a measure of habitual pitch level revealed that the majority of participants spoke using an MF0 that was low within their range compared with cisgender norms. Although the trans men generally self-reported voice masculinization, it was not correlated with MF0, frequency range, or habitual pitch level at any time point or with MF0 note change from baseline to 1 year of testosterone treatment, but correlations should be interpreted with caution due to the heterogeneous responses of the 7 participants.

CONCLUSION

In trans men, consideration of voice deepening in the context of objective and subjective measures of voice can reveal unique profiles and inform patient care.