Resonance Tuning in Singing (Formant tuning)

Transcrito por TurboScribe.ai. Atualize para Ilimitado para remover esta mensagem.

A review of research on resonance tuning by singers. This work was done on Bedegal Country
and we respect and acknowledge elders past and present. Hello, I'm Joe Wolfe.

This talk is about how singers tune a resonance of their vocal tracts to match a harmonic of the
note they're singing. Resonance tuning is useful because it gives more output sound for the
same input effort. It might also help stabilise the voice.

The talk mainly reviews the work done in our lab at UNSW Sydney. I'll identify various students
and researchers as we go, beginning with my co-author John Smith. This graphic relates
frequencies, measured in hertz or vibrations per second, to pitches.

Basses can sing with fundamental frequencies below 100 hertz and sopranos sometimes sing
above 1000 hertz. The voice has harmonics, meaning that vocal folds vibrating at 100 hertz also
generate frequency components at 200 hertz, 300, 400, etc. Here the dashed lines show the
harmonics.

The pink shading shows the range of the first vocal tract resonance, R1. The blue shows the
second resonance, R2. Their overlaps with the fundamental frequency tell us that resonance
tuning is more important for high voices.

The voice uses resonances in a way that's very different from other musical instruments.
Supposing I sing at 100 hertz. To play that note using a simple cylindrical resonator I need a
pipe 85 centimetres long.

About 100 hertz. But my vocal tract, the pipe from larynx to lips, is about 17 centimetres long
like this pipe. A pipe as long as my vocal tract has a first resonance at about 500 hertz and a
second resonance at about 1500.

Both are higher than I can sing. Wow, I've got a trombone pitch voice in a piccolo-sized
resonator. But that's for a rigid cylinder.

With a flexible pipe I can change the resonant frequencies by changing the shape. Here's a
sound with many frequencies present. I put it through a 17 centimetre cylinder and already it
starts to sound more like a human vowel.

Now if I change the shape I can change the resonant frequencies. Can you hear different
vowels like To explain what voice resonances do we'll use an oversimplified model called source
and filter. In singing the larynx is a source with lots of equally spaced harmonics, e.g. for 100
hertz the spectrum has harmonics at 100, 200, 300 etc.

The vocal tract transmits this sound but the tract boosts the harmonics that lie near its
resonances. In this sketch the tract has resonances at 200 and 1100 hertz. So here with a 100
hertz voice the output spectrum would show strong 2nd and 11th harmonics.

These strong bands of frequencies in the sound are called formants. Remember these terms.
The shape of the vocal tract determines the resonances which are properties of the tract.

The resonances produce formants which are properties of the output sound. Formants
determine what vowel we hear. For more about that see our website.

Here's a map of the vowels of Australian English described using words made of H vowel D.
Mouth opening mainly controls the vertical direction. Tongue position the horizontal. Try it
yourself without thinking of what vowel you're making.

Mouth opening, tongue constriction. The axes here are the resonance or formant frequencies.
Without changing the pitch you can vary the first resonance by opening and closing the mouth
and vary the second with the tongue.

The resonances produce the formants in your voice spectrum and they largely determine which
vowel a listener hears. Here's a different demonstration of how mouth opening controls the
first tract resonance. The eminent voice researcher Johann Schoenberg observed that sopranos
usually open their mouths wider for high notes.

He argued that they were tuning the first resonance R1 to the fundamental FO of their voice.
What we now call R1 FO tuning. Our lab developed a technique to measure the vocal tract
resonances precisely and without much disturbing the singer.

We carefully synthesize and calibrate a sound containing hundreds of different frequencies and
deliver this via a flexible tube to the singer's lower lip while he or she is singing. Here's an
example measurement. The sharp peaks are the harmonics of the voice.

FO up to 7 FO here. The quasi-continuous line is the response of the vocal tract to our signal.
The peaks in that line indicate the resonances.

In this example the first resonance R1 lies between the second and third harmonics so it
doesn't produce much of a first formant. But the second resonance R2 lies near the fifth
harmonic giving a strong second formant. Here's an example of a soprano singing OO at two
pitches.

For the note C6 she has raised R1 to match FO at about a thousand hertz giving her a very
strong fundamental. Here a soprano sings up the scale on four different vowels and we plot the
first resonance versus the pitch frequency FO. The grey shadings show the normal values for R1
for those vowels.

At low pitch she puts R1 roughly where we'd expect for each vowel but at high pitch she
increases R1 to keep it close to the fundamental so that R1 gives her voice a helpful volume
boost. Repeating for low pitch the first resonance R1 has roughly its normal value no tuning.
For high pitch whether consciously or not she's tuning R1 near the pitch frequency FO that's R1
FO tuning.

Tuning R1 means that the vowel is largely determined by the pitch not the lyrics. To illustrate
this we asked classical soprano Kristen Buchanski to sing scales on five different vowels and so
on. Then I cut the recordings up.

Here are the first notes in order and here are the last notes in a different order. Listen carefully
and identify the order. When I posted this I invited any soprano who could differentiate vowels
on the high note to send me a recording.

Nobody has. At that high pitch there simply aren't enough harmonics to identify vowels. I stress
this is not a complaint about sopranos.

It's a physical constraint on the instrument. You can't ask a trombonist to play pizzicato. You
can't ask a soprano to distinguish vowels at high pitch.

In the context of scales every soprano we've studied whether trained or not uses R1 FO tuning
above about the middle of the staff. Lower for close vowels like U. Altos and sometimes tenors
also use consistent R1 FO tuning in their highest ranges. Resonance tuning is harder for leaps
than for scales and it may require training in that context.

Sopranos usually tune R1 to the fundamental but lower voices sometimes tune R1 to a higher
harmonic. One spectacular example is Bulgarian folk singing where altos tune R1 to the second
harmonic. This is called R1 2 FO tuning.

Mara Ki demonstrates. Live it is very loud. R1 2 FO tuning is also used in belting.

Tenors and baritones sometimes tune R1 to the first second or third harmonic especially for
high notes. For the lower range low voices have less need of resonance tuning because low
pitch means the harmonics are closely spaced and that means that one of the harmonics
usually falls close to a resonance anyway. However male singers sometimes tune R1 to a higher
harmonic for important notes.

An extreme case of tuning resonances to high harmonics of low notes is diphonic or overtone
singing. Here the singer usually holds a single low note then uses a very strong tract resonance
to produce a single very prominent high harmonic. The fundamental is low and falls in a range
where our hearing is not very sensitive.

For that reason and also because the fundamental is unchanging we don't focus on the
fundamental and we notice instead the single strong high harmonic as a nearly pure tone. By
varying the tract and retuning the tract resonance overtone singers can produce melodies
using notes from the harmonic series a bit like a bugle. Mal Webb demonstrates.
Oh An interesting topic for when we have more time. In classical singing male singers use a
different resonance effect. They lower the larynx again lower the larynx.

This enhances the third and fourth resonances which together produce the singer's formant
which means a characteristic boost to any harmonics lying roughly around three kilohertz. The
singer's formant is another interesting story but it doesn't require tuning of the resonances so
I'll say no more about it here. Instead I'll talk about resonance tuning in the high soprano
range.

Relatively few sopranos sing much above the note C6 or high C which has a frequency of
roughly one kilohertz. Further relatively few sopranos are able to tune R1 much above C6.
Among the participants in our studies a few sopranos could tune R1 a few notes higher but this
required either a short vocal tract, a very large mouth aperture or both.

For many sopranos the upper limit to R1FO tuning is close to the upper limit of their singing
range. Could these limits be connected? Some of the singers we studied use a different tuning
strategy to sing substantially higher. The second resonance R2 has a convenient range for the
octave above high C and in this range some sopranos tune R2 to the fundamental.

They practice R2FO tuning. That's illustrated in the lower graph. So the very high soprano range
sometimes involves R2FO tuning.

It also sometimes involves the transition from the head register to the whistle register or the
second vocal mechanism M2 to the third mechanism M3. We don't see any simple relation
between these two transitions. A soprano can make the transition from head to whistle register
while in the R1FO tuning range or she can switch from R1FO to R2FO while still in the head
register then make the whistle transition at a higher pitch.

These options contribute to the complicated taxonomy of the high soprano range. However
most sopranos don't use the R2FO tuning. Further it's interesting that for most singers who
don't use R2FO tuning the upper limit of their singing range coincides roughly with the limit of
their R1FO tuning range.

Could it be that the limit of their tuning range limits their singing range and if so why don't they
use R2FO tuning? Well it seems to be difficult to learn R2FO tuning in part because it's
counterintuitive. Singing up a scale at the point of transition the soprano must suddenly reduce
the mouth aperture rather than increase it. So we wondered whether we could teach R2FO
tuning and other types of resonance tuning using our measurement technique to give visual
feedback.

Here's a photo of Noel Hanna using the system injecting our calibrated sound source at the lips
and here's a screen grab of what the user sees. Again the sharp peaks on the black curve show
the voice harmonics. On the musical staves at the top the harmonics are black dots and the
horizontal red bars show the resonances R1, R2 and R3.
In an experiment eight sopranos came separately to our lab to learn the difficult R2FO tuning.
In a one-hour session they received visual feedback to tune R2 first while miming then while
singing. After the one-hour session most participants could use the visual feedback to tune R2
to a desired pitch.

However after only one hour they were not able to do the R2FO tuning consistently without the
visual feedback. They often said that it was hard to reduce the mouth opening for higher notes.
It's the opposite of what they normally do.

Nevertheless two participants did spontaneously switch to R2FO tuning for single notes at the
Marie Geneteau is a soprano and a physics student who worked on this project in our lab for
four months. She used the feedback system on herself for about half an hour a day. After four
months she could do R2FO tuning without needing feedback.

Further her upper comfortable singing limit extended by a fifth from A5 up to E6. That is from
the A below high C to the E above high C. Let's review. Tuning a vocal track resonance to a
harmonic gives singers more sound for the same effort and may stabilise the voice.

The easy case is tuning the first resonance R1 to the fundamental. Sopranos usually do this
without training and other voices sometimes use it in their high range. Tuning R1 to the second
harmonic is done systematically by some altos especially in belting and in some ethnic music
styles.

Adjusting R1 to higher harmonics is sometimes used by lower voices especially from important
notes and in the extreme high range where R1FO tuning is impossible some sopranos tune the
second resonance to the fundamental. There's much more on our website search voice
acoustics. Link is below and you can also leave questions below.

Transcrito por TurboScribe.ai. Atualize para Ilimitado para remover esta mensagem.