Synthesizing choir singing

Analysis by synthesis is a method that has been successfully applied in many areas of scientific research. In speech research, it has proven to be an excellent tool for identifying perceptually relevant acoustical properties of sounds. This paper reports on some first attempts at synthesizing choir singing, the aim being to elucidate the importance of factors such as the frequency scatter in the fundamental and the formants. The presentation relies heavily on sound examples.     

Overtone singing: Productive mechanisms and acoustic data

Overtone singing is where one person sings in two voices, the first voice represented by the fundamental and the second by an enhanced harmonic. Overtone singing is performed in chest register. Tuning of the first or second formant and a reduction of the formant bandwidth down to 20 Hz make harmonics prominent. Narrowing the pharynx, velar constriction, variation of the small mouth opening, and a tension of the walls of the mouth cavity are used. Changing prominent harmonics has the effect of creating an overtone melody with sustained tones, tone steps, and trillos.     

Feedback from spectrum analysis applied to the singing voice

Despite the general availability in the past two decades of the spectrum analyzer and prominent pedagogical theories concerning the conscious tuning of vowel formants to enhance the singing voice, there has been little reported use of spectrum analysis to track formant frequencies in singing. An important exception is Sundberg's work (1) on the soprano voice. The reasons for this neglect are considered: in the singing range where information on formant tuning would be most helpful, the wide spacing of the harmonics renders the formants difficult to locate by spectrum analysis. Methods are described for obtaining continuous spectrograms with the vocal tract in the varied articulations of singing by using sweep tones and nonharmonic voice sources, and thus locating quickly and accurately the frequencies of the first five formants.     

Singer's Formant in Sopranos: Fact or Fiction?

Although it is generally agreed that the singer's formant (F_S) is a prerequisite for successful stage performance, the results of this research do not support its presence in the soprano voices of trained female singers. Results are based on a recent investigation testing 10 advanced/professional sopranos in two groups singing sustained vowels at three frequencies: high (932 Hz), mid (622 Hz), and low (261 Hz). Spectrographic analysis shows that the nature of harmonic energy varies in relation to pitch. A resonance band somewhat resembling the tenor F_S was usually evident in vowels sung at low and mid pitch. However, unlike the F_S of typically less than 1 kHz bandwidth associated with tenors, sopranos singing similar pitches produced corresponding bandwidths which were significantly broader, usually at least 2-kHz wide. Vowels sung by sopranos at high-pitch levels exhibited strong fundamental frequency production with strong reinforcement of adjacent harmonics extending to 5 kHz and beyond. This type of production in essence nullifies the necessity for a typical F_S. Absence of the F_S in strong soprano voices might also imply the adaptation of a sufficiently different overall vocal tract configuration, so that techniques geared to developing maximal projection should not be the same as those developed to maximize the F_S in other voices.     

Vocal Tract Resonance Analysis of Aging Voice Using Long-Term Average Spectra

This study is the first to use long-term average spectra (LTAS) to investigate resonance characteristics of dynamic speech in young adulthood and old age. A total of 80 speakers participated, divided equally by age group and gender. All elderly speakers were healthy, active members of the community. Measurement of the first three spectral peaks in LTAS from the first paragraph of the Rainbow Passage revealed significant lowering of peak 1 from young adulthood to old age in both men and women. Peaks 2 and 3 also lowered significantly across the adult lifespan in women and showed a tendency to lower in men. These acoustic findings are consistent with anatomic data suggesting that aging results in lengthening of the supraglottic vocal tract. Findings that women demonstrate more substantial lowering of spectral peaks with aging than men suggest that women may undergo more pronounced age-related lengthening of the supraglottic vocal tract. Alternatively, it is possible that elderly men systematically alter tongue position during vowel articulation while elderly women are less inclined to do so. Taken in conjunction with previous research, these findings suggest a "mixed model" of vocal tract resonance changes with aging in which an interaction exists between gender, the resonance effects of laryngeal lowering, and vowel articulatory patterns.     

Open and covered singing as studied by means of fiberoptics, inverse filtering, and spectral analysis

Covering the voice near the passaggio is a technique mainly used in male classic Western singing. The purpose is said to be to smooth the register transition and avoid register breaks. The physiological and acoustical differences between open and covered singing near the so-called passaggio were investigated by means of fiberoptics, inverse filtering, and spectral analysis. The study corroborated previous findings that covering physiologically means a widening and lengthening of the pharynx. Acoustically, covering implies a change of formant frequencies, and an elevated sound pressure level of the lowest source spectrum partial, i.e., the fundamental, resulting from an increased transglottal air flow. The findings indicate that covered singing also is associated with larynx lowering. Covered singing near the passaggio shows similarities to so-called phonation, and is probably preferable from the point of view of vocal hygiene.     

Comparison of vocal tract formants in singing andnonperiodic phonation

The skilled use of nonperiodic phonation techniques in combination with spectrum analysis has been proposed here as a practical method for locating formant frequencies in the singing voice. The study addresses the question of the degree of similarity between sung phonations and their nonperiodic imitations, with respect to both frequency of the first two formants as well as posture of the vocal tract. Using magnetic resonance imaging (MRI), linear predictive coding (LPC), and spectrum analysis, two types of nonperiodic phonation (ingressive and vocal fry) are compared with singing phonations to determine the degree of similarity/difference in acoustic and spatial dimensions of the vocal tract when these phonation types are used to approximate the postures of singing. In comparing phonation types, the close similarity in acoustic data in combination with the relative dissimilarity in spatial data indicates that the accurate imitations are not primarily the result of imitating the singing postures, but have instead an aural basis.     

Perceived pitch of whispered vowels-relationship with formant frequencies: A preliminary study

To clarify the role of formant frequency in the perception of pitch in whispering, we conducted a preliminary experiment to determine (1.) whether speakers change their pitch during whispering; (2.) whether listeners can perceive differences in pitch; and (3.) what the acoustical features are when speakers change their pitch. The listening test of whispered Japanese speech demonstrates that one can determine the perceived pitch of vowel /a/ as ordinary, high, or low. Acoustical analysis revealed that the perception of pitch corresponds to some formant frequencies. Further data with synthesized whispered voice are necessary to confirm the importance of the formant frequencies in detail for perceived pitch of whispered vowels.     

The singing/acting voice in the mature adult

With years of training and performance, the mature vocal performer experiences less vocal changes with aging than does his/her age peer who is not a performer. We have considered, some physical problems that may adversely influence the voice of the older performer. With some awareness and effective management of these possible problems, the negative effects on the older performer's voice can be minimized.     

Vocal Tract Area Function for Vowels Using Three-Dimensional Magnetic Resonance Imaging. A Preliminary Study

OBJECTIVE: To assess whether magnetic resonance imaging (MRI) allows the vocal tract (VT) area function to be determined for a normal male speaker. METHOD: VT shapes were acquired using MRI during sustained production of French points vowels: /i/, /a/, /u/. Cross-sectional areas were measured from a series of planes spaced at intervals of 1 cm along the length of the VT and were used as input in a previously described VT model to simulate the vowels. The first three formant frequencies, F1, F2, and F3, computed from the MRI-measured VT model were compared with subject's natural formant frequencies. RESULTS: Including piriform sinuses, calculated formants differed from measured formants F1, F2, and F3, respectively, for /i/ by -3.5%, +7.7%, and +27.5%; for /a/ by +11% +19.5%, and -4.3%; and for /u/ by +.9%, +23.4%, and +9.6%. Excluding piriform sinuses, calculated formants differed from measured formants F1, F2, and F3, respectively, for /i/ by -3.5%, +12%, and +28%, and for /u/ by +10.1%, +26.8%, and +13.7% The piriform sinuses were not visualized for /a/ on MRI. CONCLUSIONS: MRI is a noninvasive technique that allows VT imaging and determination of VT area function for a normal male speaker. Several possible sources of discrepancies are as follows: variability of the articulation, difficulties in assessment of VT wall boundaries, role of the piriform sinuses, and VT length.     

Vocal fold strain and vocal pitch in singing:Radiographic observations of singers and nonsingers

The relationship between vocal fold strain and vocal pitch in singersand nonsingers singing a rising pitch series has been indirectly investigated by means of lateral radiographs. Nonsingers tend to exhibit more strain than singers. To standardize the degree of strain, an index of strain per semitone is proposed. The semitone strain indicates the average amount of strain per 1 semitone of pitch increase or decrease. The index has been shown to be affected by several factors: gender, singing training, singing technique, voice class, age, and status of muscle function. Observations suggest that similar groups of individuals occupy different positions on the stress-strain curve, indicated by their semitone strain values.     

Formant frequency tuning in singing

It is sometimes claimed that some singers tune their two lowest formant frequencies to harmonic partials in order to increase the audibility of the voice. Voice acoustics predicts that such tuning of formants should cause vowel quality to change. Using a newly constructed digital singing machine, the authors have explored the perceptual consequences of such tuning. Four different cases were represented, in which the two lowest formant frequencies were either constant or adapted to the fundamental frequency according to either of three different strategies. The resulting voice timbres were judged by an expert panel of singing teachers in a listening test consisting of descending chromatic scales. Constant formant frequencies were clearly preferred, presumably because formant tuning entails formant frequency shifts between adjacent tones so substantial that salient vowel quality shifts occur.     

An improved model for the acoustic radiation impedance of the mouth based on an equivalent electrical network

This paper proposes a model for acoustic radiation impedance of the mouth in the form of the equivalent electrical network. Five known models of radiation impedance are compared: radiation of a circular piston set in a spherical baffle: radiation of a circular piston set in an infinite baffle, the Flanagan model, the Wakita and Fant model, and the Stevens, Kasowski and Fant model. The proposed model most accurately approximates the radiation impedance of a circular piston set in a spherical baffle. Differences between the acoustic resistance and reactance calculated by the proposed model and the piston set in a spherical baffle of 9 cm radius are relatively small in the kr < 2 region. The deviations in calculated values of the acoustic resistance and the reactance are within ±0.023 × ρc/A_m and ±0.008 × ρc/A_m, respectively, where A_m denotes the area of the mouth aperture. The accuracy of the proposed model is demonstrated by vowel formant frequency calculations. Differences in formant frequencies calculated by applying the proposed model and the piston set in a spherical baffle model are less than 0.3%.     

Comparisons of pharynx, source, formant, and pressure characteristics in operatic and musical theatre singing

Belting, a vocal technique typically cultivated in musical theatre singing, differs timbrally from operatic singing in many interesting respects. The underlying phonatory differences have not been previously investigated in detail. Yet, belting is frequently associated with disturbances of voice function. Articulatory and phonatory characteristics are investigated in a female subject who is a professional singer (co-author JL) trained in both the operatic and belting styles and in an intermediate vocal technique (“mixed”). This article presents data obtained from this subject by video-fiberoptic observation of the pharynx, inverse filtering of airflow, and measurement of subglottal pressure. The results reveal that belting was characterized by very high subglottal pressures and sound levels, and apparently also by a comparatively high degree of glottal adduction. Comparisons with other investigations of related aspects of belting and operatic singing support the assumption that the data obtained from our subject are representative for these vocal techniques.     

Large scale data acquisition of simultaneous MRI and speech

We describe an arrangement for simultaneous recording of speech and vocal tract geometry in patients undergoing surgery involving this area. Experimental design is considered from an articulatory phonetic point of view. The speech signals are recorded with an acoustic-electrical arrangement. The vocal tract is simultaneously imaged with MRI. A MATLAB-based system controls the timing of speech recording and MR image acquisition. The speech signals are cleaned from acoustic MRI noise by an adaptive signal processing algorithm. Finally, a vowel data set from pilot experiments is qualitatively compared both with validation data from the anechoic chamber and with Helmholtz resonances of the vocal tract volume, obtained using FEM.     

Effects of Tonsillectomy on Speech Spectrum

Changes in the speech spectrum of vowels and consonants before and after tonsillectomy were investigated to find out the impact of the operation on speech quality. Speech recordings obtained from patients were analyzed using the Kay Elemetrics, Multi-Dimensional Voice Processing (MDVP Advanced) software. Examination of the time-course changes after the operation revealed that certain speech parameters changed. These changes were mainly F3 (formant center frequency) and B3 (formant bandwidth) for the vowel /o/ and a slight decrease in B1 and B2 for the vowel /a/. The noise-to-harmonic ratio (NHR) also decreased slightly, suggesting less nasalized vowels. It was also observed that the fricative, glottal consonant /h/ has been affected. The larger the tonsil had been, the more changes were seen in the speech spectrum. The changes in the speech characteristics (except F3 and B3 for the vowel /o/) tended to recover, suggesting an involvement of auditory feedback and/or replacement of a new soft tissue with the tonsils. Although the changes were minimal and, therefore, have little effect on the extracted acoustic parameters, they cannot be disregarded for those relying on their voice for professional reasons, that is, singers, professional speakers, and so forth.     

Acoustic analysis of the singing and speaking voice in singing students

The singing power ratio (SPR) is an objective means of quantifying the singer's formant. SPR has been shown to differentiate trained singers from nonsingers and sung from spoken tones. This study was designed to evaluate SPR and acoustic parameters in singing students to determine if the singer-intraining has an identifiable difference between sung and spoken voices. Digital audio recordings were made of both sung and spoken vowel sounds in 55 singing students for acoustic analysis. SPR values were not significantly different between the sung and spoken samples. Shimmer and noise-to-harmonic ratio were significantly higher in spoken samples. SPR analysis may provide an objective tool for monitoring the student's progress.