The Shape of a Vowel

Music hidden in the vowels we speak

I’m sure many of us have covered and uncovered our phone’s (or even our radio’s) speakers in rapid succession to create a wah-wah sound. The music sounds dull and bass-filled, and then turns up the higher frequencies. This was my first experience with vowel modification, except I didn’t know it at the time. It just sounded like a fun thing to do, especially if I could manage to do it in sync with the beat.

When I got into college, the wah-wah pedal hit me like a revelation. The first time I remember recognising its distinctive sound in a song was in the guitar intro to Jimi Hendrix’s Voodoo Child. I was mystified by how it worked. How did this effect change the sound of any instrument to mimic something we associate with a human voice? For those who haven’t heard the wah-wah pedal in action, it modifies the underlying notes with the literal sound of wah. If you were to sing a couple of notes while constantly saying wah-wah-wah, you’d be giving the wah-wah effect. It reminded me of the times when I created this same effect with only my hand and a speaker.

It was easy to replicate with the voice too. All I needed to do was to say wah-wah, or more specifically, oo-ah-oo-ah repeatedly while singing the notes.

I thought to myself, if we can change the tone of an instrument to mimic one vocal effect, maybe we can make something that modifies a sound in any way you want and make it sound like a human voice.

I didn’t know that something like that already existed. I left the thought in the recesses of my mind for the time being.

Then I came across the vocoder. I considered it a mind-blowing invention. You could change the sound of a tone by literally making the changes with your mouth and layering the vowels you’re saying onto the music that’s playing.

If you keep your mouth really close to your phone’s speaker and change the shape of your mouth as the music plays, the vowels you’re making will be imprinted onto the music, which will follow the sounds of the vowels your mouth is making.

You don’t need to vocalise. All you need to do is keep your mouth held in specific shapes like you would say the vowel corresponding to that shape.

A vocoder could do this even without a mouth. I could just specify the vowel I wanted applied on the music and it would do it for me. Of course, using my mouth would make things very intuitive, but this wasn’t mandatory for its functioning.

I should mention here that I’m in awe of the artistic manner in which Daft Punk uses the vocoder in its songs. They make it sound so good and make it fit perfectly with the music they have in mind. From Daft Punk, I learned the importance of the vocoder and how it could be used to make any sound mimic the voice.

I didn’t know how this was happening though. I had always considered vowels to be shapes; shapes I make with my mouth that create specific sounds. I didn’t understand how something like a vocoder could change the vowel without having a mouth that could change shape. My deep association of vowels with our mouths’ shapes confused me to no end when it came to effects that played with vowels.

There must be some parameter that is independent of the need for a mouth that allows us to mimic vowel sounds, I thought.

With some reading and asking around, I found out that my habit of linking vowels with mouth shapes was the problem. While the mouth is there to facilitate the sound of the vowel, this sound can be obtained in many, many ways ways. The actual property of the sound that is being altered is its overtone intensity pattern.

A lot of us (including me at one point of time) think of pitch as identical to frequency. In reality, a pitch can consist of many frequencies, each with its own intensity. The base frequency of the pitch is called its fundamental frequency. This one frequency, along with the rest, supports and shapes the identity of the sound. Depending on the relative intensities of each frequency, the same pitch can sound very different with a change in the frequency distribution.

This is why a piano sounds different from a flute, and a violin sounds different from a guitar. Even in sounds that are not pitched, the tone varies with the frequency characteristics of the sound. A crash cymbal sounds radically different from a snare drum for this reason.

Vowels can, therefore, be altered by changing the intensities of all frequencies present in the sound for a given pitch. If we keep the fundamental frequency the same and vary the relative distribution of overtones, we can make an aa into an oo (like a wah-wah pedal). When we go from saying an aa to saying an oo at the same pitch, we’ve modified the overtone distribution of the sound.

Each vowel has a specific overtone distribution. If we can replicate the overtone peaks by any means possible, we’d have produced the sound of the intended vowel.

This was a revolutionary shift in thinking for me. I could finally think of a vowel as a specific tone and not as a shape-constricted event.

When we talk, we use consonants and vowels. The vowels carry our sound as tones to the listener and the consonants act as punctuation. We sing with our vowels and stop with our consonants, even in a simple conversation. We’re constantly modifying the overtones to deliver meaning that expresses our thought and feeling.

This is the case when we modify our overtones and our pitch remains the same, because the fundamental frequency is the same. What if we removed the focus from the fundamental frequency?

If you vocalise an oo and then change it to an aa, you can hear the difference by listening to the change in the prominent notes in the sound. Yes, there are prominent notes that you can hear.

I had earlier associated vowels with shapes but once I listened closely, I could hear the change in the higher frequencies when moving from one vowel to another.

The prominent frequency peaks that allow us to differentiate vowels are called formants. There can be more than one formant in a sound. Most of the time, 2 formants are enough to differentiate two vowels but sometimes you need at least 3.

Hopefully, you’re not alarmed by the jargon. Formants are the specific frequency peaks that help us identify a vowel. You could say that each vowel has a formant signature that identifies it as a unique tone.

If we were to remove our focus from the fundamental frequency and pay attention only to the formant frequencies while changing them, we would start to perceive vowels as a whistled tune instead of the usual fundamental-connected overtones.

I found this incredible when I first heard it. Changing vowels is like whistling, without even trying to whistle. That means that if we could somehow not think about the fundamental frequency of the sound we made, all we’d be left with was this beautiful, sonorous whistle.

Enter Tuvan throat singing.

There are many styles of overtone singing used in the world. Tuvan throat singing (or Khoomei) is done by producing a fundamental frequency and simultaneously controlling another pitch over it. Yes, you’re producing two pitches at once from your throat! The prominence of the overtone is increased to an extent that it competes with the fundamental as the perceived pitch.

Please listen to this beautiful style of singing to know how prodigiously the masters have explored it artistically. There are so many variations on this technique, but they all involve the perception of simultaneous multiple pitches. Some styles like Kargyraa are practically growling and produce an overtone that is actually below the fundamental. This is also called sub-harmonic singing. It really helps you hit those ridiculously low notes.

Overtone singing gives us to a way of changing vowels that is almost completely dissociated from the way we usually think of them. The intentional modification of overtones is used to create melodies. These overtones are harmonic overtones that are very closely related to the fundamental.

The fundamental itself stays the same so as to not attract attention to it and the overtones are used as the notes of the melody. If the fundamental is changed as well, harmonic overtones related to the new fundamental can now be used. This gives us a complex system of harmony using one person’s voice, where we can change at least two lines of melody at once. More overtones can be brought into the picture as well to create more lines of melody.

Outside the world of overtone singing, where singers focus their efforts on one pitch at a time, vowels still hold immense importance. Singers have to be very clear in their articulation of vowels to make the lyrics understandable. Some vowels give more problems in the higher pitches, so singers slightly modify vowels to go easier on their throats. These modifications can only be done in a permissible range so that the vowel can still be associated to the original vowel.

The language in use puts limits on the creative choices that are available for vowel modifications. If modifying the vowel changes it to another vowel that is heavily used in the language, and this causes the meaning of the word to change, then the singer has to try her best to walk the middle path between these two vowels so that she pronounces the correct word. This is not always an easy task. Praise to the singer who always pulls it off!

This conception of vowels as determiners of tone has changed the way I perceive music. Whether I’m mixing tracks or just listening to music, I think of the overall tone in terms of vowels. It makes it easier for me to think of the song’s sound in its entirety. For example, when changing room characteristics that I’m trying to simulate in a recording, I think of the whole music as having a voice of its own. The room is a mouth that changes its shape to say the vowel I’m imagining.

Those days in my childhood, when I was putting my mouth to the speaker and modifying the vowels to make the music change its sound, I was doing it without having a specific tone in my mind that I wanted to get. Now, I try to think of the vowel that I would want the music to have. I imagine how I would shape my mouth to get the desired tone and I try to achieve the same effect in my mix.

Just like we can sing while changing vowels with our vocal tract, I love that we can also make a song sing by changing the overall vowel sound. Of course, if overdone, it can be distracting, but using it subtly can have a powerful effect on the listener.

While composing, I think about the emotional effect that a vowel could have at a certain part and I try to make the tonality of the music fit the vowel that I feel is just right. It allows me to think of songwriting and mixing as a way of singing as well, and if anything opens my ears to a new way of listening, more often than not — I adopt it wholeheartedly.

_/\_

Around Sound turns my personal experiences with music, both as a musician and as a listener, into stories.

Improve your sense of rhythm (How I improved my sense of rhythm: Part 1, Part 2, Part 3, Part 4) as you read about my journey through the world of rhythm. How’s that for combining a lesson and a story into one? :D

Get a better grasp on notes with my 3-part How I learned to speak with notes series: Melody, Harmony and Connection

You might even find these interesting:
The Sound of Water, The Mirror in the Music and The Voice of a Story

You can have a look at all my articles here: Anirudh Venkatesh