singing The voice
Why study singing in this class?
More evidence for vocal plasticity Interesting extension of many concepts weve studied so far
Resonance F0/pitch Patterns of vocal fold vibration
An appreciation for how talented singers really are
Biomedical applications
A few issues in the study of singing
Conflicting terminology Demands of teaching vs. scientific study
Describing the way singing "feels" vs. what is actually happening Imagery, metaphor
Classical western singing much more studied than popular or nonwestern singing
Part I: Classical style Western singing (operatic voice)
Kinds of singers
Classification is partly based on F0 range
Soprano, alto, tenor, bass...
Range for most trained singers is about 2.5 octaves BUT... Singers in all classifications have overlapping ranges, so classification also depends on where in range best quality is produced
Kinds of singers
Formant frequencies are also important in voice classification Lower pitched voices naturally tend to coincide with longer vocal tracts/lower resonant frequencies. Interesting vocal effects may occur when vocal fold size and vocal tract length do not covary as expected. For example, a singer who has a long neck but short vocal folds may produce a voice that is pitched high but sounds "dark". Some subclassifications of voices in opera (lyric baritone, e.g.) may be based on this kind of hybrid case.
Resonances and voice type
How does singing voice differ from speaking voice?
For classically-trained singers, large differences occur between singing and speech. A good operatic singing voice should have:
a large overall intensity range stable vibrato an efficient voice source, so that H2 is larger than H1 (low frequency resonance) a prominent high frequency resonance, or singers formant
Operatic singers also need the ability to keep quality constant while varying pitch and loudness, and to keep pitch constant while varying loudness.
Characteristic 2: Vibrato
Usually described as nearly-periodic modulations of the fundamental frequency about its mean value, so that the pitch of the voice varies slowly but regularly around the singers target pitch. Rate is typically 4-7 Hz The pitch differential in vibrato usually extends about 6% above and below the mean F0. Amplitude changes may occur as well, but seemingly occur because of the F0 changes.
Where does vibrato come from?
Vibrato occurs naturally, but it is much less frequent in untrained than in trained voices Vibrato develops quasi-automatically during singing training, and occurs in popular singing as well as in classically trained voices. Inhibiting vibrato is difficult for most trained singers.
Where does vibrato come from?
Vibrato is more conspicuously present in trained womens than mens voices. Children sing less frequently with vibrato, and with aging, vibrato rate tends to become slower, becoming a less than pleasant "wobble". The range of normal vibrato rates is very similar to that found in pathological vocal tremors, suggesting that the rate derives from a naturalistic central nervous system process.
Why have vibrato?
Not all styles of singing require vibrato.
Singers of medieval and renaissance music often use straight tone almost exclusively probably the standard in the historical period, due to the frequent use of young males as singers, in whom this sort of intonation is more natural
Vibrato is not more efficient than straight tone
10% higher airflow in vibrato, in all three registers
Presence of vibrato does not imply that the singer can be less precise in hitting the target note.
The perceived pitch of tones with vibrato corresponds almost perfectly to the average fundamental frequency during phonation
Why have vibrato?
Part of a warm, human tone
Many listeners find vibrato more exciting, while straight tone might signal coldness and monotony, or mystery and foreboding.
Vibratory movements may prevent laryngeal fatigue, since in "pendulum like movements," the musculature alternately works and rests. It may enhance vowel intelligibility, by providing more information about where the vowel formants are.
Why have vibrato?
Vibrato may allows the singing voice to achieve a pattern distinct from accompanying instruments, such as piano, violin, or orchestra. Vibrato may help the singing tone convey emotion effectively ???
Characteristic 3: Efficient source
Shorter open phase, longer closed phase Quicker closing than opening Result: More high frequency energy in the voice
Brighter sound Less wasted air
Characteristic 4: The singers formant
A singers formant is a resonance that appears at approximately 3 kHz regardless of the vowel being sung.
The weakest part of the orchestral spectrum.
Allows the singer to increase loudness so much that he can be heard over a symphony orchestra without amplification, and without damaging his vocal folds.
Why do we need a singers formant?
The loudest harmonics in an orchestra tend to be at about the same frequency as the loudest harmonics in the normal human voice, so one would normally expect an orchestra to drown out a voice. Increasing loudness enough to be heard by normal means (increasing subglottal pressure and glottal resistance) can result in damage to the vocal folds.
Singers formant: Male voices
In males, this boost is achieved by lowering the larynx, and by widening the pharynx, laryngeal ventricle, and piriform sinuses.
Changes the frequencies of the third, fourth, and fifth formants so that they cluster, boosting the spectrum in the 3 kHz region. Increases loudness in this critical area of acoustic orchestral weakness Higher harmonics of the voice also radiate forward, so that more of the voice goes toward the audience when a singers formant is present
Because these effects are achieved by simple articulatory changes, rather than by "singing louder," the increase in loudness is achieved without any increased vocal effort or risk of damage to the vocal folds.
The singers formant
The advantage of a singers formant
What about female singers?
Female singers must also boost loudness, but for additional reasons. F0 for a soprano may climb to over 1400 Hz, producing harmonics at 2800 Hz, 4200 Hz, and so on.
What about female singers?
Typical values for the first and second formants for the vowel /a/ spoken by a female are about 936 Hz and 1551 Hz, respectively
The first formant may not be excited at all, and the second formant may be excited only by the fundamental or the second harmonic when F0 is high. Harmonics are widely spaced when F0 is high, so individual harmonics may not line up well with the center frequencies of the vocal tract resonances, even in the cases when F0 is lower than F1.
Thus, the part of the sound spectrum that is most efficiently radiated does not appear in the output spectrum at all, because no harmonics are produced in that frequency range.
How do female singers do this?
Despite this apparently built-in inefficiency, singing is much louder than speech for trained female singers, as it is for males. While males typically lower the larynx to adjust the frequencies of the higher formants, female singers open their jaws more as F0 increases, regardless of the vowel being sung. Female singers may also shorten the vocal tract somewhat by retracting the corners of the mouth, which also increases the frequencies of the formants, moving them toward the high F0.
How do female singers do this?
Opening the jaw in this way "tunes" F1 to a value near F0, increasing the sound energy to a level like that present in a singers formant (a nearly 1000-fold increase in energy. As with the singers formant, this adjustment increases loudness as a passive of effect the facts of vocal resonance. It does not require much muscular energy or any change to the voice source, thus decreasing the likelihood of injury or vocal strain.
Singers formants: Female voices
Another view of the same thing
Effects on intelligibility
These increases in loudness often come at the cost of large decreases in intelligibility. The frequencies of the first two or three formants are strong cues to the identity of the vowel being spoken, and formant frequencies depend largely on the shape and size of the vocal tract. The articulatory manipulations needed to generate vocal loudness restrict the ability of a singer to produce a target vowel accurately.
Intelligibility and operatic singing
Intelligibility of the singing voice
Fortunately, other cues to vowel identity are available from continuous song
Information from the transitions between sounds Semantic context of the lyrics Supertitles in many opera houses
Listeners need not be completely baffled when trying to decipher operatic plots.
A few notes on F0, pitch, and register in classical singing
Pitch is primarily determined by rate of vocal fold vibration, BUT... Perceived pitch is affected by resonance and spectral shape (TA contraction)
Too much TA contraction damps high frequencies: Sound may sound flat even when it isnt Too little TA contraction: Sound may sound sharp
Register
Register in singing is commonly defined as the phonation frequency range in which all tones are perceived as being produced in a similar way and possess similar vocal timbre Authors do not agree on whether or not registers exist, the number that should be distinguished (if they do exist), or how to define individual registers. Named by singers in terms of sensations they produce or "where the voice resonates in the body"
Chest, head, mixed; heavy and light voice; vocal fry, modal, falsetto, whistle...
Register
The perception of a register change probably reflects a change in the way the vocal folds vibrate Chest register
more or less the same as modal phonation often defined as a typical quality in speaking and low pitched singing (for which vibrations are supposedly felt in the chest and trachea)
Head register
sometimes described as mixture of chest and falsetto vibrations felt in the head or face
Whistle register
the very highest part of the range of a few sopranos
Register
Register boundaries do not correspond precisely to particular F0 ranges. Depending on talent and training, professional singers may be able to sing most of their vocal range in a single register without a break. Transitions between registers can cause problems in singing, especially when this happens involuntarily. Such register breaks are cultivated in yodel and in Hawaiian music, while operatic singers are trained to avoid abrupt register transitions.
Register
Vocal register occurs in many species other than humans.
Dogs bark in modal voice, and howl and whine in a falsetto-like register. Cats meow in a higher register, but growl in a lower register.
II. Non-classical singing
Kinds of singing
Significant differences between voices may also occur within a single singer across styles of singing. Many styles have been named:
belt, mix, blues, legit, croon, grunge, twang, yodel, croon...
Singing in these styles is learned mostly by listening and by trial and error Although singers seem to know what such terms mean and may produce these different qualities at will, clear definitions have not been agreed upon, and even singing teachers may not understand precisely how these qualities are produced.
Kinds of popular singing
Differences between popular singing and speaking
Differences in vocal function between speaking and singing are quite small for singers without classical training, BUT... Not much data out there (descriptions of only a few singers and singing styles.
How do styles differ?
Some consistent differences in voice source characteristics may occur across styles. Blues singing required more vocal compression and higher subglottal pressures, resulting in a decrease in the energy in the fundamental frequency and an increase in the energy in the higher harmonics (sometimes called "pressed" phonation). Pop style singing uses a more neutral glottal configuration Jazz singing is characterized by lower subglottal pressures and less glottal compression, adding energy to the fundamental frequency.
Unusual singing styles in Western music
Within Western music, heavy metal vocalists and contemporary vocal improvisers use a variety of interesting vocal effects that differ markedly from those heard in classical music (even when it is contemporary). In rock singing, "dist" tones containing subharmonics and/or broadband noise are employed as ornaments for high or loud tones Dist tones seem to be produced by using very high subglottal pressures and flow levels to set the mucosa above the vocal folds into vibration.
These vibrations may be periodic and/or aperiodic; When they are periodic, they need not be harmonically related to F0 (unlike the high tones in throat singing).
Some techniques of vocal improvisers
Similar phenomena have been described in the production of vocal improvisers, who intentionally use nonlinear phenomena like period doubling, biphonation, and broadband noise in a reproducible manner during their performances.
Ingressive airflow Asymmetrical vocal fold vibration Falsetto And many, many others
Non-western singing
Overtone singing (also called [Tuvan] throat singing, double-voice singing, biphonic singing, and Xmij)
used in music from Tibet, China, and Mongolia, Flamenco singing, and in traditional music of the Andes
Combines a high melody pitch with a low drone pitch, both produced simultaneously by the same person and always in a harmonic relationship.
Overtone singing
Two explanations have been suggested (1) A second sound source (possibly vibrations of the ventricular folds) provides one of the tones. (2) Overtone singing is produced by exploiting resonance.
Singers use articulatory adjustments to tune the formants to one of the higher harmonics of the voice. At the same time, they narrow the bandwidth of the tuned resonance, so that the higher harmonic is increased in prominence They also lower the amplitude of the first harmonic by increasing the amount of time the vocal folds remain closed during each glottal cycle (which increases the high frequency energy in the source relative to the fundamental)
Overtone singing
To tune formants and bandwidths:
narrow the pharynx add a very narrow tongue constriction keep the mouth opening very small (to reduce radiation effects on the spectrum) increase the tension of the walls of the mouth cavity to reduce damping losses.
Singers may also add nasalization to create a spectral zero that acoustically isolates the overtone resonance from the other harmonics of the voice. Perceptual data confirm that overtone singing is perceived as more nasal, stressed, and pressed than normal phonation, consistent with resonance and articulatory explanations for how it is produced.
Another kind of overtone singing
The case of a woman who produced "double voice." High-speed imaging confirmed that the singer was able to control the vibratory frequency of each vocal fold separately, so that she could sing in harmony with herself This differs from throat singers, who produce a varying overtone and a fairly background constant droning tone. Although this kind of phonation (sometimes called ,,diplophonia or ,,biphonation) occurs sometimes in vocal pathology (when it cannot be controlled), reports in normal speakers are uncommon, and the degree of control this speaker had is particularly noteworthy.