Singer Identity Research Articles

Singing and Social Identity in Young Children.

A range of studies suggest that singing activities with young children can have a beneficial impact on other aspects of their development. However, there is little research examining the relationship between young children's singing and their developing social identity. In the current study, data were captured of young children's singing and social identity as part of a larger-scale, longitudinal evaluation of the nationwide Sing Up programme in England. Participants were 720 children aged 5-8 years old. The assessment of young children's singing ability employed an established measure and was undertaken individually. With adult support, the children were also asked to complete a simple questionnaire that focused on selected aspects of their social identity, both in general terms and also related to singing. Key themes embraced their attitudes to singing (at home, in school and in informal settings), singer identity (emotional engagement with singing and self-concept), and perceptions of self (self-efficacy, self-esteem, social integration). Comparative data were collected from young children of a similar age outside the programme. Findings suggested that the programme had a positive impact on children's singing ability, both overall and including the youngest children. The data analyses suggest that children could be identified as either “pupils with positive singing identity” or “pupils with less positive, or still developing singing identity.” Overall, pupils with a more positive singer identity—irrespective of Sing Up-related experience—tended to report more positive attitudes toward singing at school and other settings, had higher perceived levels of self-esteem and social integration, as well as more positive evaluations of their singing ability. Furthermore, the research suggests that successful participation in high-quality singing activities is likely to have a positive impact on young children's singing ability and, by implication, such positive singing development will also be associated with aspects of self that are related to contexualised singer identity and their sense of social inclusion.

The song, not the singer: Infants prefer to listen to familiar songs, regardless of singer identity.

Parent's infant-directed vocalizations are highly dynamic and emotive compared to their adult-directed counterparts, and correspondingly, more effectively capture infants' attention. Infant-directed singing is a specific type of vocalization that is common throughout the world. Parents tend to sing a small handful of songs in a stereotyped way, and a number of recent studies have highlighted the significance of familiar songs in young children's social behaviors and evaluations. To date, no studies have examined whether infants' responses to familiar versus unfamiliar songs are modulated by singer identity (i.e., whether the singer is their own parent). In the present study, we investigated 9- to 12-month-old infants' (N=29) behavioral and electrodermal responses to relatively familiar and unfamiliar songs sung by either their own mother or another infant's mother. Familiar songs recruited more attention and rhythmic movement, and lower electrodermal levels relative to unfamiliar songs. Moreover, these responses were robust regardless of whether the singer was their mother or a stranger, even when the stranger's rendition differed greatly from their mothers' in mean fundamental frequency and tempo. Results indicate that infants' interest in familiar songs is not limited to idiosyncratic characteristics of their parents' song renditions, and points to the potential for song as an effective early signifier of group membership.

Semantic Tagging of Singing Voices in Popular Music Recordings

Singing voice is a key sound source in popular music. As recent music streaming and entertainment services call for more intelligent solutions to retrieve songs or evaluate musical characteristics, automatic analysis of popular music targeted to singing voice has been a significant research subject. The majority of studies have focused on quantitative or objective information of singing voice such as pitch, lyrics or singer identity. However, singing voice has a wide variety of dimensions that are somewhat difficult to quantify and therefore we often describe by words. In this article, we address the qualitative analysis of singing voice as a music auto-tagging task that annotates songs with a set of tag words. To this end, we build a music tag dataset dedicated to singing voice. Specifically, we define a vocabulary that describes timbre and singing styles of K-pop vocalists and collect human annotations for individual tracks. We then conduct statistical analysis to understand the global and temporal characteristics of the tag words. Using the dataset, we train a deep neural network model to automatically predict the voice-specific tags from popular music recordings and evaluate the model in different conditions. We discuss the results by comparing them to the statistical analysis of tag words. Finally, we show potential applications of the vocal tagging system in music retrieval, music thumbnailing and singing evaluation.

Voice Change and Singing Experiences of Adolescent Females

The purpose of this qualitative study was to understand how undergraduate- and graduate-level choristers perceived the influence of adolescent female voice change on their singing experiences since the age of 11 years old. Data included individual and focus group interviews with 17 female collegiate choral singers enrolled in one of two women’s choirs at a large midwestern university. Three overarching themes emerged: perceptions of vocal development, teacher influence, and emotion. Findings revealed that participants’ perceptions of singing experiences since age 11 largely revolved around vocal challenges, fear, and insecurity and have remained mostly negative even after the passing of time. Experiences in choir were not necessarily positive, and participants perceived their individual vocal needs as secondary to the needs of the larger choral ensemble. Out of fear of hurting the larger choir, participants remained passive and did not self-advocate, resulting in additional challenges with self-confidence and struggles pertaining to singer identity development during adolescence and into adulthood.

Low delay statistical singing voice conversion with direct waveform modification based on spectral differential considering global variance

This paper presents a low-delay statistical singing voice conversion (SVC) method with direct waveform modification based on spectral differential (DIFFSVC) considering global variance (GV). Statistical SVC based on a Gaussian mixture model (GMM) is a technique to convert singer identity of a source singer's voice into that of a target singer converting several acoustic features. We have successfully improved quality of converted singing voices by proposing the DIFFSVC considering the GV to avoid parameterization errors caused by vocoding process and alleviate over-smoothing effects. On the other hand, this method is based on batch-type conversion processing, and therefore, it is difficult to directly apply it to real-time conversion processing that needs a low delay conversion method. In this paper, to develop a high-quality and real-time SVC system, we propose a low-delay DIFFSVC method using GV-based post-filtering process. The experimental results have demonstrated that the proposed method makes it po...

“We ARE our instrument!”: Forming a singer identity

Preparation for the operatic stage means that for some students their voice will undergo significant transformation during training. For most operatic singers, voice type will determine future roles, the music they perform and potential career trajectory. Voice type becomes a facet of identity and position within the operatic world. This article presents a narrative thematic analysis of a young woman’s account of her teaching and learning experiences in preparation for a career on the operatic stage. Two semi-structured interviews were conducted, and beliefs and values about her singing and voice type were elicited. This article discusses how significant changes to her voice have affected her self-concept, singer identity and potential career in opera. Also discussed is the impact of conservatorium environment, peers and mentors on her feelings of self-belief.

Cross-modal signatures in maternal speech and singing.

We explored the possibility of a unique cross-modal signature in maternal speech and singing that enables adults and infants to link unfamiliar speaking or singing voices with subsequently viewed silent videos of the talkers or singers. In Experiment 1, adults listened to 30-s excerpts of speech followed by successively presented 7-s silent video clips, one from the previously heard speaker (different speech content) and the other from a different speaker. They successfully identified the previously heard speaker. In Experiment 2, adults heard comparable excerpts of singing followed by silent video clips from the previously heard singer (different song) and another singer. They failed to identify the previously heard singer. In Experiment 3, the videos of talkers and singers were blurred to obscure mouth movements. Adults successfully identified the talkers and they also identified the singers from videos of different portions of the song previously heard. In Experiment 4, 6− to 8-month-old infants listened to 30-s excerpts of the same maternal speech or singing followed by exposure to the silent videos on alternating trials. They looked longer at the silent videos of previously heard talkers and singers. The findings confirm the individuality of maternal speech and singing performance as well as adults' and infants' ability to discern the unique cross-modal signatures. The cues that enable cross-modal matching of talker and singer identity remain to be determined.

Voice, place and genre in popular song performance1

Earlier sociolinguistic studies have conceptualised popular song as a field of phonological variation where singers do or do not maintain features of their national or regional accents in singing. The present paper explores a wider agenda for the sociolinguistics of popular song, theorised as a diverse field of performance organised according to genre. Following initiatives in the sociology of popular music (particularly Simon Frith's research), voice is interpreted as the repertoire of meaning‐making options available to performers. Voice subsumes dialect indexicality, but also the management of singer identity and singer‐audience relations through the performance of lyrics, rhythmic and bodily modalities. Place is understood as the specific socio‐cultural contexts that are explicitly or implicitly voiced, including contexts of performance and reception. By performing within or against generically structured stylistic norms, performers construct and disseminate different vernacular values and identities. Live tracks from three different, broadly‐defined genres are considered in detail – classic rock and roll (Chuck Berry'sMaybellene), folk/country (James Taylor'sCopperline), and punk rock (the Sex Pistols’Johnny B. Goode).

Classification of the Classical Male Singing Voice Using Long-Term Average Spectrum

Singing-voice classification is often considered the cornerstone of a classical singer's identity. Traditionally, classification has been a highly subjective, nonstandardized process. As a result, misclassification of the singing voice is thought to be common, especially in young singers. Long-term average spectrum (LTAS) average is an objective measurement that could be used to classify a singer's voice. The purpose of this study was to determine the relationship of LTAS with singing-voice classification. Descriptive between-subject study. Nine professional classical male singers performed the "Star Spangled Banner" in a comfortable key of their choice. LTAS was calculated for the first two phrases, the remainder of the song, and the entire song. The overall LTAS averages of each sample as well as the physiological and singing ranges were compared with self-reported singing-voice classification. Voice classification and overall LTAS average were moderately correlated, but the strength of the correlation varied with each sample. The strongest correlation was with the entire song. Voice classification and singing range were strongly correlated. LTAS remains a promising tool to aid in singing-voice classification. However, how to best use LTAS in classification remains unclear because of the influence of sample length and phonetic and pitch content on LTAS.

Modeling vocal sounds in polyphonic musical audio signals

This paper describes our research aimed at modeling vocal sounds (singing voices) in available music recordings and its applications to singer identification, singer similarity, and lyrics synchronization. Our predominant‐F0 estimation method, PreFEst, can obtain the melody line by modeling the input sound mixture as a weighted mixture of harmonic‐structure tone models (probability density functions) of all possible F0s and estimating their weights and the tone models by MAP estimation. Since the PreFEst was designed for general melodies, we extended it to specialize in vocal melodies by using vocal timbre models ‐‐‐ vocal and nonvocal GMMs. Those GMMs are trained beforehand and used to evaluate the vocal probability. The GMMs are also used to identify vocal regions, but its strategy should depend on applications. For singer identification and singer‐similarity calculation, since the purpose is to model singer's identity by training each singer's vocal GMM, certainly reliable vocal regions should be ident...

I have a voice but I just can't sing: a narrative investigation of singing and social anxiety

The purpose of this study was to investigate adult singing anxieties arising within the context of a music methods course. Participants were three female elementary education majors who reported suffering from anxiety related to singing. Anxiety toward singing was evidenced through observations and participants’ descriptions of cognitive, somatic, behavioral, and affective symptoms. Several themes emerged from data collected over a 10-week period through participant journals, interviews, and field texts. Singing anxiety only seemed to arise within a social context, where participants seemed concerned with the possibility that they would be personally evaluated by others. The self-presentational theory of social anxiety was supported by these cases. All participants traced the roots of their anxiety to negative experiences in a school music program. Music teachers and family members seemed to have the greatest influence in shaping their beliefs about singing ability and singer identity.

Algorithm for separating vocals from polyphonic music

An algorithm for the separation of vocals from polyphonic music is described. The algorithm consists of two stages, which first estimate the predominant melody line, and then the sinusoidal modeling parameters corresponding to the melody line. The melody line is estimated using a hidden Markov model where the output of a multiple fundamental frequency estimator is used as a feature set. The states of the hidden Markov model correspond to musical notes having different fundamental frequencies, and the state transition probabilities are determined by a musicological model. The sinusoidal modeling stage estimates the frequency, amplitude, and phase of each overtone of the predominant melody line in each frame. The sinusoidal modeling stage can also include mechanisms which reduce the effect of interfering sound sources on the estimated parameters. The resulting separation algorithm is independent of singer identity, musical genre, or instrumentation. Simulation experiments on real polyphonic music show that the algorithm enables separation of vocals from the accompaniment, providing robust results on various musical genres.

How do they duet? Sexually dimorphic behavioural mechanisms structure duet songs in the black-bellied wren

In many bird species, pairmates sing together, forming duet songs. Studies on duetting songbirds consistently report nonrandom associations between the male and female song types that constitute duet songs. Previous research has demonstrated that when female black-bellied wrens, Thryothorus fasciatoventris , answer male song to form a duet, they abide by individually specific, temporally stable duet codes linking discrete acoustic stimuli (male song types) to discrete vocal responses (female song types). The present study builds on this finding with the goal of understanding the behavioural mechanisms in both males and females that result in nonrandom song type associations. I recorded duets from free-living pairs, and isolated four transitions from male song type to female song type in the duets of each pair. These transitions served as predicted links in the females' duet codes. The male and female song types from the transitions were played to the pair from which they had been recorded. Both males and females ‘duetted’ with playbacks of their mates' songs. Females' answers adhered very closely to the predictions, supporting previous findings. Males also answered according to the predictions at a level exceeding chance, demonstrating that they abide by the same duet codes as do their mates. Stimulus type was a strong predictor of response type in females, but a relatively weak predictor of response type in males. When presented with chimerical male song stimuli, females typically answered with one of the predicted responses for the songs composing the chimera. The song fragments constituting the beginning and end of chimerical stimuli were equally influential in determining female responses, but longer fragments were more influential than shorter fragments. Overall, the findings are consistent with the hypothesis that answering according to a unique duet code communicates the singer's identity to the mate, facilitating cooperative defence of the pair's shared resources.

Chickadee Song Structure is Individually Distinctive Over Long Broadcast Distances

AbstractThe two-note fee-bee song of male black-capped chickadees functions during the dawn chorus, in part, as a sexual signal across large distances. How song structure might encode information about male quality, however, remains unclear. We studied the availability of cues to male social rank (a proxy indicator of male quality), within the acoustic structure of dawn chorus songs of male chickadees whose flock dominance status we determined the previous winter. We used analysis of variance and discriminant function analysis to demonstrate that five temporal, frequency or relative amplitude features of song can predict individual identity but not the category of social rank (dominant versus subordinate) to which individuals belong. After transmitting chickadee songs through the forest and re-recording them at four broadcast distances, we found that song structure continued to effectively predict singer identity by our statistical methods despite significant acoustic degradation for as long as songs remained audible (up to 80 m). In particular, the relative frequency interval between the two notes is both the most invariant between-male measure and among the most individually distinctive. We conclude the structure of dawn chorus songs could function across large distances to signal the identity of familiar singing males whose relative quality is known to the listener from other interactions (such as encounters within winter flocks).

Analysis and enhancement of country singing

The study of human singing has focused extensively on the analysis of voice characteristics. At the same time, a substantial body of work has been under study aimed at modeling and synthesizing the human voice. The work on which we report brings together some key analysis and synthesis principles to create a new model for digitally improving the perceived quality of an average singing voice. The model presented employs an analysis‐by‐synthesis overlap‐add (ABS‐OLA) sinusoidal model, which in the past has been used for the analysis and synthesis of speech, in combination with a spectral model of the vocal tract. The ABS‐OLA sinusoidal model for speech has been shown to be a flexible, accurate, and computationally efficient representation capable of producing a natural‐sounding singing voice [E. B. George and M. J. T. Smith, Trans. Speech Audio Processing 5, 389–406 (1997)]. A spectral model infused in the ABS‐OLA uses Generalized Gaussian functions to provide a simple framework which enables the precise modification of spectral characteristics while maintaining the quality and naturalness of the original voice. Furthermore, it is shown that the parameters of the new ABS‐OLA can accommodate pitch corrections and vocal quality enhancements while preserving naturalness and singer identity. Examples of enhanced country singing will be presented.

Vowel representation and synthesis of singer identity

Professional singers shape the acoustical attributes of their vocal production to create the perceptual impression of a single instrument spanning a variety of vowel timbres over a wide range of pitch. Samples of the five Italian vowels over a two-octave range were analyzed using the modal distribution to yield a time-frequency image of the signal. From such images (high-order) representations of the composite transfer function of the signal were extracted. These representations provide little evidence for the existence of an acoustic ‘‘signature’’ that is invariant over the singer’s entire range of production. Rather, the data, along with results from a perceptual study, suggest that singers create the impression of a single instrument by smoothly transitioning from one local region of invariance to the next. By identifying these regions, low-order, perceptually robust approximations are constructed which bear the identity of the singer when used to synthesize new instances of sung vowels. These low-order representations also serve as endpoints for interpolating between two different singers, as well as boundary points for extrapolating a singer from chest to head modes of production. [Work supported by the MusEn Project at the University of Michigan.]

Singing is based on two markedly different kinds of signaling

Singing is a complex form of communicating, incorporating two classes of signaling specializations with distinctive properties: primary signal units, and formal rules governing sequential order in sustained performances. The properties of each class enable it to provide its own characteristic kinds of information. As a fundamental procedure used in a wide range of circumstances, singing has many functions and is a major part of the communication of an enormous spectrum of animals. Yet recognition of singing's breadth and informative potential is commonly obscured by involved and inadequate definitions and use of one term, “song”, to cover both the components of singing and the formal sequencing rules. A simple, operationally defined distinction between sustained performances and their components is proposed herein, and its implications are examined. Sustained orderly singing performances are particularly useful for communicating at a distance, often obviating closer encounters. None the less, because singing provides information continuously it can also operate in close, nearly stable interactions (rapidly changing events disrupt or preclude singing). Each component signal of a performance usually provides information about a singer's identity, location and behavior, and different components can provide different information about behavior. The pattern in which the components are sung can reveal relative probabilities of different actions and the rates and directions of changes in those probabilities. The elaboration of singing that provides such information cannot be explained simply as a result of sexual selection or other postulates that assume almost total redundancy of components. Contrary to many existing definitions, singing is not limited by complexity of component form, the gender or age of performers, or season. It functions much more widely than is generally understood, for example, in parent-offspring interactions, mobbing, and vigilance behavior. Much singing that goes unrecognized because it incorporates signal units that are not “songs” is revealed by distinguishing between the properties of sustained performances and the very different properties of their components. Even non-vocal signaling sequences can be formally similar to singing.