On the use of the maximum likelihood estimation for analysis of vibrato tones

Characteristics of frequency vibrato are usually explained by the frequency center, rate, and extent. For analysis of vibrato tones, we apply the maximum likelihood estimation to the enhanced fundamental frequency track obtained from a short-time Fourier transform. We have also suggested the values of the parameters in the estimation and the averaging process to reduce the estimation error. Experiments show that the proposed method can be used for accurate and reliable analysis of vibrato tones.     

Modal Distribution Analysis, Synthesis, and Perception of a Soprano''s Sung Vowels

A high-resolution time-frequency analysis technique, the modal distribution, is applied to sung vowels from a soprano singer. Parameters are estimated for each partial component of notes analyzed with the modal distribution. These estimates are used in an additive synthesis model to generate replicates of the original recording, using a series of time-varying sinusoids. Additionally, a source-filter model is applied to create synthetic signals, where pitch- and vowel-specific filters and driving functions are constructed from the amplitude and frequency estimates obtained. Different driving functions, which sample the range of this singer's rate and excursion variation, are transposed and filtered to create synthetic signals. The perceptual salience of the different rates and excursions is then determined via a paired-comparison listening experiment. It is found that listeners are sensitive to small variations in both average vibrato rate and average vibrato excursion. However, the perceived amount of vibrato excursion varies somewhat depending upon the pitch at which the vibrato is “played” synthetically. Finally, the naturalness and sound quality of these synthetic examples is determined through both paired-comparison and single-note sound quality scaling listening experiments.     

Acoustic variability in vibrato and its perceptual significance

Historically, studies of vocal vibrato have concentrated on pulse rate as being a primary factor in determining whether a given vocal movement is a good or bad vibrato or a tremolo or wobble. More recently, investigators have been studying the extent of frequency variation and amplitude variation around their respective means in order to determine their influence on the perception of vibrato. The present study is an additional attempt to understand the three parameters comprising vibrato, their interrelationship, and their relationship to perception. Samples of sustained sung tones were obtained primarily from recordings. The samples were digitized using a 16-bit A/D converter at a sampling frequency of 10 kHz. Each digitized sample was converted to a useful format for marking purposes in order to derive information on vibrato pulse rate, the mean frequency of the tone, the semitone deviation around the mean, percent frequency deviation and percent amplitude variation around the mean amplitude. Data presentation utilizes representative samples of good vibrato, tremolo and wobble and describes differences in waveforms which may impact on perception.     

The time-frequency characteristics of violin vibrato: modal distribution analysis and synthesis

A high-resolution time-frequency distribution, the modal distribution, is applied to the study of violin vibrato. The analysis indicates that the frequency modulation induced by the motion of the stopped finger on the string is accompanied by a significant amplitude variation in each partial of that note. Amplitude and frequency estimates for each partial are extracted from the modal distribution of ten pitches that span the range of the violin instrument. The frequency modulation is well-represented by a single sinusoid with a mean rate of 5.9 Hz and a mean excursion of +/- 15.2 cents. A spectral decomposition of the amplitude envelopes of the partials shows that the peaks lie primarily at integer multiples of the vibrato rate. These amplitude and frequency estimates are used in an additive synthesis model to generate synthetic replicates of violin vibrato. Simple approximations to these estimates are created, and synthesized sounds using these are evaluated perceptually by seven subjects using discrimination, nonmetric multidimensional scaling (MDS), and sound quality scoring tasks. It is found that the absence of frequency modulation has little effect on the perceptual response to violin vibrato, while the absence of amplitude modulation causes marked changes in both sound quality and MDS results. Low-order spectral decompositions of the amplitude and frequency estimates also occupy the same perceptual space as the original recording for a subset of the pitches studied.     

Vibrato and pitch transitions

There are at least two timing variables to be considered in vocal music performance. The first is the note changes associated with the meter signature (4/4, 2/4, etc.) and the second is the vibrato rate of the performer. Because the probability is not great that these two temporal variables will always be in perfect synchrony, it was the purpose of this investigation to determine what singers do when these two timing variables come into conflict during singing. Six singers recorded a series of alternating upward and downward interval shifts of a third while singing whole notes, half notes, quarter notes and eighth notes. The same notes were then recorded with intervals of a fifth. The recorded samples were converted to a visual trace and examined for interactions between vibrato and meter. Analysis of the tracings indicated that, in the majority of the cases, the singers would alter their vibrato in order to adhere to the timing of the musical line.     

On the dynamics of the clavichord: from tangent motion to sound

An experimental study of variations in the sound of clavichord notes at different dynamic levels is described. Radiated acoustic signal, tangent velocity and two tangent-string contact signals are synchronously measured for all 51 notes of an unfretted instrument. More than ten repeated measures are recorded in order to obtain as much variation in dynamic level as possible. The tangent motion, expressed in terms of velocity, is studied in the time and frequency domains. A model of the tangent-string contact point velocity is proposed. Then, three aspects of the sounded tones are analyzed: SPL and its relationship to tangent velocity, spectral slope, and pitch variations. These results indicate a linear relationship between sound pressure level and tangent peak log velocity. Spectral slope seems almost constant independent of tangent velocity and dynamic level. Both tangent velocity and finger pressure are shown to influence the fundamental frequency. In conclusion, controlling both finger velocity and finger pressure may prove challenging for the player, and this may explain why the sound quality of the clavichord depends so much on the players ability.     

Effects of Professional Singing Education on Vocal Vibrato—A Longitudinal Study

Vocal vibrato is regarded as one of the essential characteristics of voice quality in classical singing. Professional singers seem to develop vibrato automatically, without actively striving to acquire it. In this longitudinal investigation, the vocal vibrato of 22 singing students was examined at the beginning of and after 3 years of professional singing education. Subjects sang an ascending-descending triad pattern in slow tempo on vowel [a:] at a comfortable pitch level twice at soft (piano) and twice at medium (mezzoforte) loudness. The top note of the triad pattern was sustained for approximately 5s. The mean and the standard deviation (SD) of the vibrato rate were measured for this note. Results revealed that after 3 years of training, voices with vibrato slower than 5.2 Hz were found to have a faster vibrato, and voices with vibrato faster than 5.8 Hz were found to have a slower vibrato. Standard deviation of vibrato rate was higher in soft than in medium loudness, particularly before the education. Also high values of SD of vibrato rate, exceeding 0.65 Hz, had decreased after the education. These findings confirm that vibrato characteristics can be affected by singing education.     

Auditory stream segregation in monkey auditory cortex: effects of frequency separation, presentation rate, and tone duration

Auditory stream segregation refers to the organization of sequential sounds into "perceptual streams" reflecting individual environmental sound sources. In the present study, sequences of alternating high and low tones, "...ABAB...," similar to those used in psychoacoustic experiments on stream segregation, were presented to awake monkeys while neural activity was recorded in primary auditory cortex (A1). Tone frequency separation (AF), tone presentation rate (PR), and tone duration (TD) were systematically varied to examine whether neural responses correlate with effects of these variables on perceptual stream segregation. "A" tones were fixed at the best frequency of the recording site, while "B" tones were displaced in frequency from "A" tones by an amount = delta F. As PR increased, "B" tone responses decreased in amplitude to a greater extent than "A" tone responses, yielding neural response patterns dominated by "A" tone responses occurring at half the alternation rate. Increasing TD facilitated the differential attenuation of "B" tone responses. These findings parallel psychoacoustic data and suggest a physiological model of stream segregation whereby increasing delta F, PR, or TD enhances spatial differentiation of "A" tone and "B" tone responses along the tonotopic map in A1.     

Acoustical Comparison Between Samples of Good and Poor Vibrato in Singers

The purpose of this research was to analyze samples of frequency vibrato taken from recordings of eight different singers, which were classified as examples of good or poor singing. The samples were analyzed by a software package, which makes use of the linear prediction coding (LPC) method to determine the time varying rate and extent of the frequency vibrato wave. Four parameters, which relate to the periodicity of the samples, were extracted from the time varying rate and extent and investigated in order to verify or reject the hypothesis that the best vibrato samples were the most symmetric ones. Ten samples per singer were analyzed, 5 good and 5 poor, for a total of 80 samples. The results show that the samples judged as good were the most periodic ones.     

Acoustic characteristics of vocal oscillations: Vibrato, exaggerated vibrato, trill, and trillo

Acoustic analysis was performed on recordings of 10 early music singers producing examples of vibrato, exaggerated vibrato, whole-tone trill, and trillo to obtain measures of oscillation rate, frequency extent, and jitter. Oscillation rates for vibrato, exaggerated vibrato, and trill were similar, but trillo rate was much faster. Average frequency extent of oscillation was 1 semitone (st) for vibrato, 2.21 st for exaggerated vibrato, 2.71 st for whole-tone trill, and 1.64 st for trillo. Jitter measures indicated that exaggerated vibrato had the most stable oscillations and trillo the least.     

Vocal warm-up produces acoustic change in singers' vibrato rate

Vibrato rate and vibrato extent were acoustically assessed in 12 classically trained female singers before and after 25 minutes of vocal warm-up exercises. Vocal warm-up produced three notable changes in vibrato rate: (1)?more regularity in the cyclic undulations comprising the vibrato rate of a note, (2) more stability in mean vibrato rates from one sustained note to the next, and (3) a moderating of excessively fast and excessively slow mean vibrato rates. No significant change was found for vibrato extent. The findings indicate that vocal warm-up may regulate vibrato rate. Thus tone quality, which is strongly linked to vibrato characteristics, may undergo positive change as a result of vocal warm-up.     

The role of auditory feedback in sustaining vocal vibrato

Vocal vibrato and tremor are characterized by oscillations in voice fundamental frequency (F0). These oscillations may be sustained by a control loop within the auditory system. One component of the control loop is the pitch-shift reflex (PSR). The PSR is a closed loop negative feedback reflex that is triggered in response to discrepancies between intended and perceived pitch with a latency of approximately 100 ms. Consecutive compensatory reflexive responses lead to oscillations in pitch every approximately 200 ms, resulting in approximately 5-Hz modulation of F0. Pitch-shift reflexes were elicited experimentally in six subjects while they sustained /u/ vowels at a comfortable pitch and loudness. Auditory feedback was sinusoidally modulated at discrete integer frequencies (1 to 10 Hz) with +/- 25 cents amplitude. Modulated auditory feedback induced oscillations in voice F0 output of all subjects at rates consistent with vocal vibrato and tremor. Transfer functions revealed peak gains at 4 to 7 Hz in all subjects, with an average peak gain at 5 Hz. These gains occurred in the modulation frequency region where the voice output and auditory feedback signals were in phase. A control loop in the auditory system may sustain vocal vibrato and tremorlike oscillations in voice F0.     

Vibrato rate and extent in soprano voice: a survey on one century of singing

This work presents a statistical study of vibrato parameters in soprano voices. More than one hundred recordings of the same tone sung by 75 artists have been analyzed. Vibrato rate and extent, tone length and intonation, together with their correlations are the main parameters under examination. The study shows a clear decrease of the mean vibrato rate during the last century (-1.8±0.3 Hz/century), together with an increase of vibrato extent (56.4±0.3cent/century). Vibrato rate and extent show a statistically significant negative correlation (r=-0.62). Vibrato rate increase near the end of the tone has been observed too, in agreement with previous measurements, together with a mean increase of the pitch of the tone. A small positive correlation has been also found among note duration and vibrato extent.     

Vibrato Rate Adjustment

The goal of the present study was to document the acoustic changes that occur as singers attempt to increase or decrease their vibrato rate to match target stimuli. Eight advanced singing students produced vowels with vibrato in three registers, both naturally and while attempting to match faster or slower rate stimuli. Slower rates were associated with lower intensity and less steady vibrato. Faster rates involved increased vibrato extent in the chest register and increased intensity in the head register. Singers whose spontaneous vibrato rates were naturally either slower or faster tended to also be relatively slower or faster when matching target rates. This ability to modify rate may have beneficial effects on the artistic quality of the voice for performance.     

The relationship between measured vibrato characteristics and perception in Western operatic singing

This study examined the association between acoustic and perceptual data related to vibrato in Western operatic singing using recordings of performances by internationally famous opera singers. Three related studies were conducted. Study 1 used commercial recordings of the same five singers and the same cadenza examined by Siegwart and Scherer(1), measured vibrato rate and extent in each singer's performance of the cadenza and tested possible associations between these vibrato attributes and judges' preference for singers. Studies 2 and 3, using recordings of different internationally famous singers and a different cadenza, measured vibrato onset, rate, and extent in each singer's performance of the cadenza, required judges to rank the singers in order of personal preference, to identify the emotion expressed, and to assess the degree of success in communicating emotion achieved by the singer. The findings showed that the perception of the singers' vibrato did not always agree with acoustic measurements. However, a comparison of the acoustic measurements with the preference and emotion judgments suggest that some elements of vibrato may affect listeners' perception of the voice, their preference for a particular singer, and assist the communication of emotion between singer and audience.     

Recognizing transient low-frequency whale sounds by spectrogram correlation

A method is described for the automatic recognition of transient animal sounds. Automatic recognition can be used in wild animal research, including studies of behavior, population, and impact of anthropogenic noise. The method described here, spectrogram correlation, is well-suited to recognition of animal sounds consisting of tones and frequency sweeps. For a sound type of interest, a two-dimensional synthetic kernel is constructed and cross-correlated with a spectrogram of a recording, producing a recognition function--the likelihood at each point in time that the sound type was present. A threshold is applied to this function to obtain discrete detection events, instants at which the sound type of interest was likely to be present. An extension of this method handles the temporal variation commonly present in animal sounds. Spectrogram correlation was compared to three other methods that have been used for automatic call recognition: matched filters, neural networks, and hidden Markov models. The test data set consisted of bowhead whale (Balaena mysticetus) end notes from songs recorded in Alaska in 1986 and 1988. The method had a success rate of about 97.5% on this problem, and the comparison indicated that it could be especially useful for detecting a call type when relatively few (5-200) instances of the call type are known.     

Vibrato of saxophones

Several alto saxophone players' vibratos have been recorded. The signals are analyzed using time-frequency methods in order to estimate the frequency modulation (vibrato rate) and the amplitude modulation (vibrato extent) of each vibrato sample. Some parameters are derived from the results in order to separate the two ways of vibrato playing: vibrato "à la machoire" and vibrato "sur l'air." Moreover, time domain simulations of single-reed instrument vibratos are created. The model is controlled by two parameters: the mouth overpressure and a parameter characterizing the reed-mouthpiece system. Preliminary comments and comparisons between the simulated vibratos and recorded vibratos results are made.     

Comparative measures of vocal tremor and vocal vibrato

Acoustic analyses were carried out on vocal vibrato produced by nine opera singers and vocal tremor accompanying the sustained phonation of patients with the following diagnoses: Parkinson's disease, amyotrophic lateral sclerosis, spinal muscular atrophy, essential tremor, and adductor spastic dysphonia. While vocal tremor on average had a faster oscillatory rate and greater amplitude extent when compared to vocal vibrato, only the cycle to cycle measures of shimmer and jitter differed significantly between these groups. However, these differences existed even when the effect of the oscillation was removed. These data are consistent with the hypothesis that vocal vibrato in singers and vocal tremor in patients may be part of the same continuum.     

Effects of Topical Anesthetic and Flexible Fiberoptic Laryngoscopy on Professional Sopranos

This study examined the acoustic and perceptual effects of topical anesthetic and flexible fiberoptic laryngoscopy (FFL) against a control condition on the singing voices of ten professional sopranos. Recordings of a section of an aria, various scales, and a messa di voce exercise were obtained in the three experimental conditions. Acoustic analyses of the same aria section recorded during the three conditions were similar with respect to the distribution of energy across the spectrum (LTAS) and vibrato rate and extent. The ability of the participants to achieve their highest and lowest notes or to complete the messa di voce was also not affected by the anesthetic or FFL. Perceptual ratings of a variety of parameters by experienced singing teachers also revealed little difference across conditions with only "appropriate velopharyngeal closure" found to differ in one comparison. These results indicate that highly experienced operatic sopranos are either not affected by or appear to have the ability to compensate for the presence of anesthetic and the FFL. The most likely explanation is that this group of singers relied on a solid vocal technique. Results will need to be replicated on less accomplished singers before concluding that this medical procedure does not affect the operatic singing voice.