How to play the first bar of Rhapsody in Blue

To play this bar, players learn to tune a strong resonance of their vocal tract to the desired note—as well as unusual finger actions. This resonance of the air in the vocal tract may be stronger than the resonances in the bore of the instrument. This is a brief, non-technical introduction (see this scientific paper). It was part of the project of PhD student Chen Jer Ming.
sheet music for Rhapsody in Blue

spectrogram of first bar of Rhapsody in BlueThe two-and-a-half octave solo clarinet portamento opening Gershwin’s Rhapsody in Blue is one of the great icons of 20th century music and one of the best known bars in music. Commissioned for an “experimental concert” by Paul Whiteman and his jazz orchestra, Rhapsody premiered with Gershwin on the piano. It was not until rehearsals of the Rhapsody began that the portamento (often called a glissando) unintentionally came into being: “ a joke on Gershwin, Gorman (Whiteman's virtuoso clarinettist) played the opening measure with a noticeable glissando, adding what he considered a humorous touch to the passage. Reacting favourably to Gorman’s whimsy, Gershwin asked him to perform the opening measure that way at the concert and to add as much of a ‘wail’ as possible.” At its première (Feb 12 1924 at New York’s Aeolian Hall), “Ross Gorman began his glissando and electrified the house”. This performance tradition has continued to delight audiences ever since.

Replicating Gorman’s ‘wail’ is now standard performance practice, but is a difficult act to reproduce. To achieve this effect, expert players combine unusual fingerings with unusual configurations of their vocal tract to achieve a nearly continuous rise in playing pitch.

The spectrogram of the glissando plots frequency (vertical axis) against time, and sound pressure as colour. The lowest line on the graph is the fundamental frequency f, which gives the playing pitch. The opening trill on the note G3 is executed from 0 to 2 seconds and a scale-like run from 2.5 to 3.5 seconds. Thereafter, the pitch rises smoothly over an octave from C5 (3.5 seconds) to the sustained note at C6 (beginning at 5.7 seconds). (Note names are in clarinet written pitch.)

(An interesting feature of clarinet sound is that, for some of the lowest notes, the lowest even harmonics are weaker than the odd. Over most of the range in which the glissando is played, however, there is little difference between even and odd harmonics. An explanation is here.)

We first measured the clarinet’s resonances for various fingering positions used in the glissando. This was followed with a novel technique to study the player’s vocal tract non-invasively: a sound source and a microphone were incorporated into a clarinet mouthpiece, allowing us to “look” into the player’s vocal tract without disturbing the player and so measure the tract resonances as the player performs (see below). We then compared the vocal tract resonances with the corresponding clarinet resonances for the fingering used at that playing pitch. See the scientific paper for details.


Clarinet Resonance and Fingers

    Seven of the tone-holes on the clarinet are covered directly by the fingers rather than key pads. Thus the player can gradually uncover the holes by progressively sliding the fingers off the clarinet. Measurements of the clarinet’s resonance at various stages of finger-slide positions clearly show the intermediate resonances created as the finger gradually slides off the holes. Over part of the instrument’s range, this contributes to the smoothly increasing playing pitch (instead of jumping in discrete ‘steps’ like a normal musical scale). But it is only part of the story.

Clarinettist’s Vocal Tract Resonance

    The frequency at which the clarinet reed vibrates determines the pitch of the note played. That frequency is determined by a combination of the resonances in the bore of the instrument and the resonances in the vocal tract of the player. Usually, the clarinet’s resonances dominate and the player’s tract has only a minor effect. However, the final octave of the glissando (from C5 to C6) lies in a range of the clarinet where the clarinet resonances are weaker. In this range, experienced players can produce very strong resonances in the vocal tract comparable with or larger in magnitude than those in the clarinet: the vocal tract can ‘win’ the battle to control the reed. By tuning this resonance of the vocal tract, expert players can control the playing pitch continuously.
    Francesco Celata of the SSO plays the first bar of Rhapsody in Bluespectrogram of first bar of Rhapsody in Blue
    A schematic of the measurements of vocal tract resonances during playing. Several eminent Sydney clarinettists came to the lab for this study. Here, Francesco Celata, of the Sydney Symphony Orchestra, plays one of the experimental instruments.

    The glissando in Rhapsody in Blue requires a strong vocal tract resonance and smooth control over a large pitch range. However, the playing pitch need only deviate from that of the fingered note by a semitone or so. Greater deviations are possible: in other experiments using a single fingering, we measured the resonances of the player’s tracts as they shifted the pitch by several notes, using the vocal tract alone.


    In the clarinet, the reed’s oscillation determines the sounding note. Downstream from the reed is the clarinet bore, while the player’s vocal tract is upstream. For normal clarinet playing, downstream resonances in the clarinet bore (determined by the fingering used) dominate to drive the reed to oscillate at a particular frequency. However, if the upstream resonance in the player’s vocal tract is adjusted to be strong enough and at the appropriate frequency, the vocal tract resonance will then compete with the clarinet resonance to determine the reed’s playing frequency. (Occasionally the reed resonance has a major influence, and we hear that as a squeak!)

    By skilfully coordinating the fingers to smoothly uncover the clarinet finger-holes and simultaneously tuning strong vocal tract resonances to the continuously changing pitch, expert players are able to facilitate a smooth trombone-like glissando, particularly in the final octave of the run.

Links for more information

The scientific paper reporting this study and a brief conference paper are: Here are some references to our related work. Other scientific papers are at this link.
This project is part of the doctoral research of Jer Ming Chen, whose supervisors are John Smith and Joe Wolfe. Yamaha logo


Our research work on saxophones is supported by the Australian Research Council. The Yamaha Corporation provided instruments and Légère the synthetic reeds.

[Basics | Research | Publications | Flutes | Clarinet | Saxophone | Brass | Didjeridu | Guitar | Violin | Voice | Cochlear ]
[ People | Contact Us | Home ]

© Joe Wolfe /
phone 61-2-9385 4954 (UT + 10, +11 Oct-Mar)
Joe's music site

Music Acoustics Homepage What is a decibel? Didjeridu acoustics