How Tuvan vocalists sing two notes at once

Allison Parshall: In southern Siberia, just north of Mongolia, there’s a republic about the size of North Dakota that has been capturing the attention of music lovers across the world for decades. 

Sean Quirk: So you go through these massive Alpine mountains, and you make this turn and there’s a big sign that says “Republic of Tuva”.

Allison Parshall: This is one of those music lovers — Sean Quirk, a Milwaukee native who has been living in the Republic of Tuva since 2003.

Sean Quirk: And at the instant we turn past that sign, lying before you is this beautiful spectacular golden plane with all of these mountains rising up in the distance, like the crests of waves. 

Allison Parshall: Quirk is the producer, road manager, and sound engineer for the Tuvan music ensemble Alash. I actually saw Alash perform in high school and I was shocked at the end of the concert to find out it had been nearly three hours long. I didn’t feel any of that time, because the music just… sucks you in.

(Music fades in. A vocalist sings a low note with a high, whistling melody overtop)

Allison Parshall: Now, to be clear, that is one person singing — both that low, buzzing drone and the whistling melody dancing above are coming from the same person’s mouth, at the same time. Tuvan musicians, like Alash’s three band members, are masters of a vocal technique called khöömei, sometimes referred to as throat singing. Quirk got hooked on khöömei in college with a CD his friend bought him, by a Tuvan folklore ensemble called Huun-Huur-Tu, which he refers to as Alash’s musical forefathers. 

Sean Quirk: I wouldn’t let anybody put another CD in the player for like the first month. Just like, put that Tuvan stuff on, man, that’s what I want to hear.

Allison Parshall: Tuvan singers take advantage of the fundamental principles of sound to create two notes at once. It’s often called “overtone singing”, and in theory, anyone can do it.

Sean Quirk: Trying to do it, I was like, pretty painful on the ear. So looking back on it, you know, I have no ill will towards the roommates who kicked me out or the girlfriend who dumped me. I get it. I lost my job because of it, because I was always trying to do it in the break room of the music store I worked at.

Allison Parshall: That one CD changed his life in plenty of good ways, too. He went to Tuva on a Fulbright scholarship and now lives there with his wife and five trilingual children. And his khöömei is a lot better these days, though he still leaves that to the professionals. To understand how these Tuvan master vocalists — and Sean — sing two notes at once, let’s look at what happens when you or I sing a regular old note.

Pantelis Vassilakis: At the end of the day, you have a sound that is made out of a collection of frequencies at different strengths. It’s a perceptual illusion, the fact that all of them sound as one to start with.

Allison Parshall: This is Pantelis Vassilakis, who teaches sound perception at Loyola Marymount University. He explains that all sounds in nature are made up of many frequencies, or pitches, combined together. It’s what gives sounds their color, their character.

Pantelis Vassilakis: Actually I cannot think of a single sound created in nature that would contain only one frequency inside.

Allison Parshall: Now, we can create sounds of just one frequency with computers. Here’s what that sounds like.

(An ear-numbing, computer-generated note)

Allison Parshall: I personally find it kind of unpleasant, so to make it sound a bit more like a sound you’d find in nature, we can layer other frequencies on top. 

(More computer-generated notes layer together, each a higher frequency than the last)

Allison Parshall: There are many different frequencies here, but your brain still perceives them together as one note. 

(The same computer-generated notes play all at once, melding together into one)

Allison Parshall: These higher frequencies are called overtones, and they have particular mathematical relationships to each other. Different instruments and voices accentuate certain ranges of overtones over others — which is part of what makes a guitar, which is heavy-handed with its low overtones…

(Single note on a guitar)

Allison Parshall: …sound so different from a sitar, which has a lot of bright, high overtones.

(Single note on a sitar)

Allison Parshall: In the end, the overtones you get all come down to the shape and material of the instrument, and how the sound resonates within it. And then there’s the instrument of the human voice. Tuvan throat singers are able to vary the shape of their vocal tract, from deep in their throats to the front of their mouths, so that the sound resonates in very particular ways. 

They start with a low drone. Then they shape their throats so precisely that you hear one of those overtones resonating louder than others. And by switching up which overtone you hear, they create a melody. 

Christopher Bergevin: Because they have such precise motor control over the vocal tract that they can almost pick out just one of those overtones and accentuate it while all others are basically just almost… squashed.

Allison Parshall: Christopher Bergevin is a physicist at York University in Toronto and a fan of the legendary Tuvan ensemble, Huun-Huur-Tu, that first hooked Quirk on Tuvan music. Back in 2019, Huun-Huur-Tu came to Toronto, and Bergevin’s colleague had a lucky connection.

Bergevin: We were able to get the band to come visit us here at York and do a bunch of recordings, including some MRI.

Allison Parshall: The researchers wanted to use magnetic resonance imaging, or MRI, to see how the singers’ vocal tracts changed as they sculpted their overtone melodies. There were a couple of challenges to tackle. 

Christopher Bergevin: He had never tried to do this sort of singing laying down before, let alone with his head just completely wrapped up with this coil in the bore of an MRI tube. 

Allison Parshall: In case you were curious, here’s what Tuvan throat singing sounds like in chorus with the racket of an MRI scanner.  

(A voice throat singing over the sounds of an MRI scanner)

Christopher Bergevin:  What we were able to come up with is a relatively simple classical sort of model for how the vocal tract works. But you know, ultimately what they’re doing is more complicated than just what we modeled. 

Allison Parshall: As cool as this model is, Tuvan khöömei singers don’t need — and have never needed — a mathematical model to teach and master their art. Here’s Sean Quirk.

Sean Quirk: Every teacher I’ve talked to of khöömei says, generally, there’s a different approach for people who are Tuvans and people who aren’t. People who aren’t Tuvans really like to have some kind of anatomical explanation. And Tuvan learners are generally just acquainted enough with the sound that it’s just this very imitative kind of process. And that doesn’t mean that there’s not like all kinds of complexity and depth in how it’s taught.

(Alash’s music fades in, complete with instruments and multiple vocalists singing in Tuvan)

Allison Parshall: And despite not being Tuvan, that music seems to have stuck with Quirk, and with me, for a reason. 

Sean Quirk: Even though I’ve been here for like 18 years and I don’t like to romanticize stuff, there’s just this feeling that arises within you wherever you are when you hear Tuvan music. It’s this reflection. Everything about nature informs the music and everything about the music reflects the nature around. 

(Alash’s music rises again, then fades)

Allison Parshall: For Scienceline, I’m Allison Parshall.