Vocal Tract Length Normalization Features for Audio Search

This paper presents speaker normalization approaches for audio search task. Conventional state-of-the-art feature set, viz., Mel Frequency Cepstral Coefficients (MFCC) is known to contain speaker-specific and linguistic information implicitly. This might create problem for speaker-independent audio search task. In this paper, universal warping-based approach is used for vocal tract length normalization in audio search. In particular, features such as scale transform and warped linear prediction are used to compensate speaker variability in audio matching. The advantage of these features over conventional feature set is that they apply universal frequency warping for both the templates to be matched during audio search. The performance of Scale Transform Cepstral Coefficients (STCC) and Warped Linear Prediction Cepstral Coefficients (WLPCC) are about 3% higher than the state-of-the-art MFCC feature sets on TIMIT database.

Presbítero Manuel Maximiliano Alberti (1763-1811) : párroco de San Nicolás de Bari y vocal de la Primera Junta. En el bicentenario de su muerte

Resumen: El artículo presenta las sucesivas etapas que configuraron el derrotero del sacerdote patriota en las primeras horas del movimiento independentista. Se destacan tanto los valores de su desempeño ministerial como su ascendencia social y política en el ambiente porteño, con la perspectiva propia de los ideales ilustrados de su tiempo y en particular durante el curato en San Nicolás de Bari, la segunda parroquia más importante de la ciudad. Con información bien documentada da cuenta de su nacimiento y de su formación desde sus estudios humanísticos, se reconstruyen los años siguientes dedicados al oficio pastoral de regreso en Buenos Aires y en Maldonado (Uruguay) en tiempos de las invasiones inglesas, su ministerio en San Nicolás de Bari cuando acontece la Revolución de Mayo y su tarea a favor de la causa patriótica desde su participación en el Cabildo Abierto del 22 y su incorporación como vocal de la Primera Junta. Se despliegan los aspectos sobresalientes de esta última misión y los aspectos que se destacan en este recorrido por la vida y la obra de este cura patriota son: compromiso honesto, integridad y actitud de servicio sin reservas desde la función pública.

Numero vocal, exemplar, catholico, e politico, proposto ao mayor entre os santos, o glorioso S. Joam Baptista : para imitaçam do mayor entre os principes, o serenissimo D. Joam V., Nosso Senhor : offerece por Mão do Excelentissimo D. Francisco Xavier de Meneses, Conde de Eyriceyra, do Conselho de sua Majestade, & C. Sebastiam Pacheco Varella

Sebastião Pacheco Varella nasceu em Aveiro, Portuga em 1671 e morreu na mesma cidade , em 170 6. Notável pregador e muito hábil nas línguas latina, espanhola, francesa e italiana, o autor publicou ‘Numero vocal’. O ‘Sermão da seráfica Madre Santa Teresa, na manhã da sua festa, pregado em Aveiro no ano de 1700’, em Coimbra, em 1701; ‘Sermão da seráfica Madre Santa Teresa, na tarde da sua festa, pregado em Aveiro, no dito anno’, em Coimbra, em 1701; e ‘Sermão da bem-aventuraria Santa Joana, Princesa de Portugal e Senhora de Aveiro, pregado no mosteiro da mesma vila em que viveu e morreu na ultima tarde do seu tríduo’, em Lisboa, em 1702. Essa edição é a primeira da obra, que é um conjunto de princípios inspirados na fé católica e no exemplo de São João Batista, como objetivo de orientar os monarcas católicos em suas ações. Segundo Inocêncio, ‘Varella reúne em seu favor os votos de alguns críticos ilustrados’. Morais o "cita" no Dicionário muitas vezes, autorizando com ele o uso de vários termos, e o P. Francisco José Freyre não duvidou qualificá-lo de autor de bastante propriedade na locução e de linguagem corrente’

Orígenes indoeuropeos (y no indoeuropeos) de la vocal [y] eslava

[ES] La vocal 'y' de las lenguas eslavas procede esencialmente de i.e. *ū. Pero en el dominio de la flexión tanto nominal como verbal se puede encontrar asimismo una vocal 'y' de origen intraeslavo y derivada de diversas secuencias en final de palabra. Al estudio de estas evoluciones así como al análisis de los distintos estratos de préstamos léxicos que incrementan la presencia de la vocal y en el léxico eslavo se dedica buena parte de este trabajo. En él se rastrean, además, los datos comparativos e internos que permiten detectar otro origen para la vocal eslava 'y' los diptongos i.e. *ou y *au.

FEEDING AND VOCAL BEHAVIORS OF BLACK GIBBONS (HYLOBATES-CONCOLOR) IN YUNNAN - A PRELIMINARY-STUDY

Feeding and vocal behaviours of wild black gibbons (Hylobates concolor) were observed from 1987 to 1989 in south-western Yunnan, notably H. concolor jingdongensis at Mt. Wuliang (24-degrees 18-42'N, 100-degrees 30-50'E) in the early spring of 1989. 12 plant species were observed to have been eaten by the gibbons; these included tree species, lianas and epiphytes. Approximately 21 % of feeding time was devoted to eating fruits, 61 % to leaf buds and shoots, 7 % to flowers and 11 % to leaves. The gibbons preferred fruits to leaves even though they commonly ate leaves. In this study, the morning songs (duet and solo), reponsive (territorial) songs, alarm calls and communication calls were recorded. The gibbons sang their morning songs mainly in the early morning, with a single bout lasting more than 10 min on average. The singing of a group would trigger other groups, and all groups in an area tended to sing sequentially. The morning duet song bout was dominated by an adult male. The male emitted booms, aa notes, early multimodulated figures, intermediate multimodulated figures and codas, the latter occurring only in duets following the female's great call. The female uttered great calls and abortive great calls. The subadults or juveniles also took part in the morning songs. Lone males were heard to utter solos which lasted longer than the duets of the pairs.

Mixed source model and its adapted vocal tract filter estimate for voice transformation and synthesis

In current methods for voice transformation and speech synthesis, the vocal tract filter is usually assumed to be excited by a flat amplitude spectrum. In this article, we present a method using a mixed source model defined as a mixture of the Liljencrants-Fant (LF) model and Gaussian noise. Using the LF model, the base approach used in this presented work is therefore close to a vocoder using exogenous input like ARX-based methods or the Glottal Spectral Separation (GSS) method. Such approaches are therefore dedicated to voice processing promising an improved naturalness compared to generic signal models. To estimate the Vocal Tract Filter (VTF), using spectral division like in GSS, we show that a glottal source model can be used with any envelope estimation method conversely to ARX approach where a least square AR solution is used. We therefore derive a VTF estimate which takes into account the amplitude spectra of both deterministic and random components of the glottal source. The proposed mixed source model is controlled by a small set of intuitive and independent parameters. The relevance of this voice production model is evaluated, through listening tests, in the context of resynthesis, HMM-based speech synthesis, breathiness modification and pitch transposition. © 2012 Elsevier B.V. All rights reserved.

The Net and the Self: Colliding Views of Individuality and Nationhood in the Pre-Devolutionary plays of Mark Jenkins and Ed Thomas

Owen, Roger, 'The Net and the Self: Colliding Views of Individuality and Nationhood in the Pre-Devolutionary plays of Mark Jenkins and Ed Thomas', In: 'Cool Britannia: British Political Drama in the 1990s', Rebecca D'Mont? and Graham Saunders (eds), (Basingstoke: Palgrave Macmillan), pp.158-175, 2007 RAE2008

The feminine soul: love, courage, and sacrifice in three centuries of vocal repertoire

Music is a unique form of communication. It sings from soul to soul, touching each of us in uniquely different ways. While researching the life of Clara Schumann, I became intrigued with the idea that feminine qualities, as expressed by her music, could be found in other composers work for the soprano voice. I explored this possibility in a lecture recital and followed up this work by performing two operatic roles and a recital.

CIRQUE DE LA VOIX: VOCAL PERFORMANCES FOR THE 21st CENTURY AUDIENCE

This dissertation addresses the growing need to entice people to attend a classical solo vocal recital by incorporating thematic programming, multi-media presentations, collaborations and innovative marketing. It comprises four programs that use the above tactics, creating live performances of classical vocal music that appeal to the attention deficient 21st-century audience. Each program focuses on repertoire appropriate for the male alto voice and includes elements of spoken word, visual imagery and for movement through collaborations with actors, singers, dancers, designers and visual artists. Program one (March 1, 2004), La Voix Humaine: The Life of an Englishwoman in Music, Poetry, & Art, outlines the life of a fictitious Englishwoman through a self-composed narration, spoken by an actress, a Power Point presentation of visual art by 20th-century English artists and musical commentary provided by the collaboration of a vocalist and a pianist. Program two (October 15, 2004), La Voix Thfrmatique: Anima - Music that Moves, is a program of pieces ranging from the 14th- to the 20th-centuries of which half are choreographed by members of the University of Maryland Dance Department. Program three is a lecture recital entitled L 'Haute Voix: Identifying the High Male Voice and Appropriate Repertoire which is presented in collaboration with three singers, a pianist, a harpsichordist and a cellist. Program four, La Voix Dramatique: Opera Roles for the Countertenor Voice, comprises performances of George Frederic Handel's Giulio Cesare in Egitto (1724) in collaboration with the Maryland Opera Studio and the Clarice Smith Performing Arts Center (Leon Major, director; Kenneth Merrill, conductor). There are two performances each of the title role, Cesare (April 15 & 17, 2005), and his nemesis, Tolomeo (April 21 & 23,2005). All programs are documented in a digital audio format available on compact disc and are accompanied by program notes also available in digital format. Programs two and four are also documented in digital video format available on digital video disc.

Reclaiming the Vocal High Ground: Performing Baroque Repertoire in a Modern World

The early music revival of the late twentieth century revolutionized music with the birth of historically lnformed performance. With this revolution came a stereotype of the "early music singing voice" as small, bright, straight-toned, and unfortunately, often inferior techrucally to the mainstream opera singer. An assessment of the validity of this stereotype was made though readings of treatises and modern manuals of performance practice, and through listening to recordings. Sources on ornamentation, recitative, dance rhythm, and baroque gesture were examined, resulting in the finding that these issues are far more important to historical accuracy than are voice timbre and size. This dissertation is comprised of three historically informed performances intended to satisfy both the early music specialist and the mainstream voice teacher. Program One (May 15, 2004) is a performance of The "Peasant" Cantata, BWV 212, by J.S. Bach, with The Bach Sinfonia at the Washington Conservatory of Music. Program Two (January 29, 2005) is the role of Eurilla in a staged production of Antonio Vivaldi's serenata, Eurilla e Alcindo. The performance is a collaboration with the Baltimore-based ensemble, La Rocinante, and is conducted from the keyboard by Joseph Gascho. Program Three (March 14, 2005) is a solo recital entitled, Fairest Isle: Music of Baroque London. All three programs are documented in a digital audio format available on compact disc, with accompanying programs and notes also available in digital format.

Imaging Learned Song Representations in Populations of Sensorimotor Neurons Essential to Vocal Communication

Perceiving or producing complex vocalizations such as speech and birdsongs require the coordinated activity of neuronal populations, and these activity patterns can vary over space and time. How learned communication signals are represented by populations of sensorimotor neurons essential to vocal perception and production remains poorly understood. Using a combination of two-photon calcium imaging, intracellular electrophysiological recording and retrograde tracing methods in anesthetized adult male zebra finches (Taeniopygia guttata), I addressed how the bird's own song and its component syllables are represented by the spatiotemporal patterns of activity of two spatially intermingled populations of projection neurons (PNs) in HVC, a sensorimotor area required for song perception and production. These experiments revealed that neighboring PNs can respond at markedly different times to song playback and that different syllables activate spatially intermingled HVC PNs within a small region. Moreover, noise correlation analysis reveals enhanced functional connectivity between PNs that respond most strongly to the same syllable and also provides evidence of a spatial gradient of functional connectivity specific to PNs that project to song motor nucleus (i.e. HVC_RA cells). These findings support a model in which syllabic and temporal features of song are represented by spatially intermingled PNs functionally organized into cell- and syllable-type networks.

Convergent differential regulation of parvalbumin in the brains of vocal learners.

Spoken language and learned song are complex communication behaviors found in only a few species, including humans and three groups of distantly related birds--songbirds, parrots, and hummingbirds. Despite their large phylogenetic distances, these vocal learners show convergent behaviors and associated brain pathways for vocal communication. However, it is not clear whether this behavioral and anatomical convergence is associated with molecular convergence. Here we used oligo microarrays to screen for genes differentially regulated in brain nuclei necessary for producing learned vocalizations relative to adjacent brain areas that control other behaviors in avian vocal learners versus vocal non-learners. A top candidate gene in our screen was a calcium-binding protein, parvalbumin (PV). In situ hybridization verification revealed that PV was expressed significantly higher throughout the song motor pathway, including brainstem vocal motor neurons relative to the surrounding brain regions of all distantly related avian vocal learners. This differential expression was specific to PV and vocal learners, as it was not found in avian vocal non-learners nor for control genes in learners and non-learners. Similar to the vocal learning birds, higher PV up-regulation was found in the brainstem tongue motor neurons used for speech production in humans relative to a non-human primate, macaques. These results suggest repeated convergent evolution of differential PV up-regulation in the brains of vocal learners separated by more than 65-300 million years from a common ancestor and that the specialized behaviors of learned song and speech may require extra calcium buffering and signaling.