Hearing roughness (polyphonia)

Air leak corresponds to the stroboscopic finding of incomplete closure. Since the first question for the examiner to answer is whether the sound is clearhusky or rough (or a combination), we should determine what roughness or polyphonia corresponds to. 

Roughness in the human voice is most often diplophonia but can be other types of polyphonia if there are more than two sound sources. Diplophonia may be harmonic (slightly pleasing) when the two pitches are mathematically related. That is, they will blend and augment each other to some degree. If two sound sources are not mathematically related, the sound quality will tend to be unpleasant as the two sounds acoustically interfere with each other rather than amplify each other.

Case study #1


Candy lost her voice after a cold last year. Since then, her singing voice remains rough. She cut back on singing because it makes her voice worse. She recalls that for a number of years her voice has a sexy quality when she uses it too much. She does love to talk; she rates herself a 6 on a 7-point talkativeness scale.

Vocal capabilities

During vocal capabilities testing, her voice is clear and very energetic on a reading task. Her maximum phonation time at her comfortable pitch is 9 seconds, a short time for a singer. From spontaneous speech, reading speech and maximum phonation time tests, I can begin to graph her voice.


The ovals represent typical male & female speaking pitches. Her comfortable speaking pitch is centered about F3#, and her speaking is melodic, covering a few notes above and below this. This is in the range of a typical female’s speaking pitch, so green represents the clear oscillations in this pitch range. 

I ask her to make sounds at several pitches, working up toward higher pitches.


She has air leak in the middle of her range until C5, where there is a jump in pitch, up an octave to two simultanteous notes, C6 and a second nearby note (diplophonia). When she gets louder, there is a strong grating noise during phonation just above C5.


Even before I get out an endoscope to look at her vocal cords, I know from her vocal capabilities that something split the vocal cords into two unequal vibrating segments in her upper range.

Taking a look at her vocal cords after this much of the vocal evaluation, we see that the vocal cords physically touch at the pitch that generates roughness. The longer anterior segment vibrates at a lower pitch than the posterior segment. It is also the pitch being detected by the stroboscope’s input. We can watch this segment move in apparent slow motion. The posterior segment moves more irregularly on stroboscopy because the strobe is not firing in sync with the pitch it is producing. 


View during stroboscopy: two simultanteous notes, C6 and a second nearby note (diplophonia). The anterior segment is longer than the posterior segment.

Case study #2


A female lost her singing voice 2 years ago.

Vocal Capabilities

When I had her glide from high to low, her voice was clear at high pitch, then alternated between diplophonic and air leak as she glided downward in pitch.


She starts with a clear sound at G4. Gliding lower in pitch there is air leak, then flutter. The air leak and flutter alternate as she descends to her lowest note.


She has a paresis of the left thyroarytenoid muscle. The cricothyroid muscle, intact on both sides, adds tension in her uppermost range, keeping both vocal cords relatively straight and at nearly the same tension yielding clarity. However, as she drops lower in pitch by relaxing the cricothyroid muscles, the paretic, loose left vocal cord oscillates further and further laterally, leaking air.


Flutter of the left vocal cord.

Intermittently the looser, left vocal cord would oscillate at a completely different pitch from the right. At those moments, while the right vocal cord was producing a single pitch, the left vocal cord was vibrating very irregularly and she would produce a sound much like a flag, flapping in the breeze. The weaker left vocal cord was oscillating at a very low pitch and the oscillations were very irregular. Flutter is a finding occurring when there is insufficient tension on a vocal cord to maintain regular oscillations.


We can diagram roughness on a pitch vs. volume plot. This correlates with a finding of two sound sources on stroboscopy.

In these example cases we have:

  1. Used a few vocal capabilities (reading and high pitch and glide from high to low) to hear a specific vocal impairment – nonharmonic diplophonia.
  2. Diagrammed the impairments on a pitch vs. volume plot.
  3. Correlated the audible finding of diplophonia at various pitches with the endoscopic finding of 2 visible sound sources. The strobe light could only track one of the two segments. The other segment appears to flutter.
  4. The two sound sources may be two short segments, created by a central glottic swelling dividing the membranous vocal cord into anterior and posterior sound sources, two short segments.
  5. The two sound sources may be the left and right vocal cords if they are different from each other in some way (mass, tension or length). 
  6. However, two segments do not always generate diplophonia. Two segments of the same length (and tension and mass) will tend to vibrate at the same pitch.
  7. Even at some small difference (in length, tension or mass), the sound energy from one cord may temporarily drive the other cord to vibrate at the same pitch.

There were other audible findings in these individuals.