Kashudo (歌手道): Second formant distraction: a point of identity

There are times when an apparent vocal imbalance isn’t a vocal imbalance at all. Such was the experience with one of my students recently. The top voice was always secure but the middle, as is the case with many female voices was inconsistent. I worked with her on balancing the two muscular registers and the problem was minimized. She could sing through her range with relative ease, but I was not satisfied. There was a quality about the sound in the middle that seemed inappropriate to her particular voice. I knew that she sang musical theater in her youth, so I asked her to take out one of the pieces she sang early in her development before she had any real training. Low and behold, a perfectly balanced middle range was achieved within minutes. The moment she let go of her “operatic voice” the true voice emerged.

It also explained much about this relatively young spinto who had been developed as a dramatic soprano before she was thirty years old. Singers who develop spontaneously, that is without much training, must be instructed carefully. This particular singer had been in search of her “real” voice for so long, but seduced by the idea of being a dramatic soprano she took a wrong turn. She could not let go of the false middle voice that was manifactured to support the dramatic soprano identity until she gave up that identity. She is in fact more likely to be able to sing dramatic repertoire (when the voice has fully matured) now that the voice is being produced efficiently. The new presence in the middle and low makes much of her spinto repertoire viable even now.
This does not mean that everyone can find their true voice by singing music that they sang in their youth. But that is certainly a helpful strategy in discovering at least part of the singer’s “nature”. Some voices, from inappropriate habits (e.g. speaking, poor training) begin their early training with imbalances. Therefore returning to that time will not fix the imbalance, but it could expose how far historically the imbalance goes. The nature of such voices is discovered through training. Such is certainly my case. I feel natural as a tenor, whereas I had to create my baritone voice over time (of course it did not feel as if I was manufacturing a sound back then, but rather looking for my real voice).
The more difficult problem for that particular singer and perhaps for many is that they perceive the operatic sound as something foreign and different from their “speaking voices”. If the speaking voice is balanced, it is very likely that the singing voice would develop spontaneously from it. However when the young singer begins training for opera, s/he often thinks s/he has to do something differently to be viably operatic.
There are real reasons why singers feel that the operatic sound is a fabricated sound. I have observed over the last 20+ years of teaching that the acoustic changes in the voice often detract the singer from his/her balanced phonation. In other words, the singer will spontaneously change his/her mode of phonation when s/he feels a change in resonance. Because of this the most common registration problems occur in the female middle voice (through and right above the first acoustic change), and in the male upper voice (through and above the only acoustic change–excluding countertenors). For that reason, a skilled “belter”, who essentially does not change formant strategy through the acoustic area around Eb4-G4 tends to have a more consistent phonation pattern than the average opera singer in training. This is because the belter is not distracted by the acoustic passaggio areas.
In truth, the second formant, which defines the operatic voice, feels foreign to the singer at first because it is different from the resonance of the speaking range, which the singer identifies as his/her natural voice. As discussed here before, spontaneous accessing of the second formant (high resonance) depends as much on acurate adjustment of the second formant as it does on maintaining viability in the first formant (low resonance). That is:
If we consider the two formants as two equally strong magnets with the sung notes between them, if the two magnets maintain their respective positions, as the notes get higher they will be pulled by the magnetism of the upper formant. Likewise, the lower formant will influence them as they go lower. If however the lower resonator follows the sung pitch (the larynx rises–we know that a high larynx raises the frequency of the firs formant), at the passaggio point the first formant will continue to dominate where the second should take over. The higher notes in such a case lose warmth and so the singer feels a need to make up for that loss of quality by altering his/her natural phonation or resonance mode.
In the case of my student, it is the resonance of the middle range that had been altered to fit her concept of a more dramatic voice type. That in turn had forced a thicker phonation mode in that part of the voice. I had begun to deal with the weight issue and had done exercises that would promote a more natural resonance strategy. Both issues (because they were related) were corrected as soon as she went back to her musical theater mode of singing. This approach restored her natural resonance mode and corrected the muscular (phonation) imbalance for the most part. Slight weight issues were quickly corrected because she could feel them clearly, since they were no longer masked by her “learned” resonance adjustment to fabricate a dramatic sound.
To summarize, this experience addresses many issues: 1) The speaking voice (formant 1) identified by the singer as the natural voice quality; 2) The foreign sensations associated with second formant resonance, which is necessary in the middle of the female operatic voice and the upper middle and top of the male operatic types; 3) Spontaneous register balancing as a result of going back to the singer’s “native” mode of singing. These are fundamental issues we need to continue discussing. A balance is always to be struck between the objective nature of scientific information and the subjective nature of the singer’s personal experience.
I therefore conclude this post with yet another axiom from my former teacher, the late Glenn Parker: “Teach the singer, not just the voice!”
© 08/15/2009

4 thoughts on “Kashudo (歌手道): Second formant distraction: a point of identity

  1. Indeed, this was one of the things I found difficult as well – learning to stop adjusting according to my own notion of what it was supposed to sound like, and just focus instead on maintaining the position.

    I will admit to some confusion from your use of the word 'format'. I know that there are different uses of the word. Personally, I've been trained in the terminology used by e.g. Prof. Johan Sundberg, who consistently uses the word 'formant' to mean resonance frequencies in (and around) the pharynx. In this context, the first and second formants form the vowels, while the third and fourth (mainly) affect the color of the voice. f0 is then typically referred to as the 'fundamental'. Even Sundberg straddled the fence when he talked about the “singer's formant”, when this phenomenon would strictly be called an amplification due to a clustering of the third and fourth formants (singer's formant is catchier though).

    Not to impune your post, which I thought was excellent. I don't have any suggestions on how to phrase it better than you did, and your use of 'formant' is clear from the context.

  2. You are right Ulf! I would clarify that formants are “frequency bands” that are variable. Depending on their specific bandwidth they influence where the acoustic energy of the vocal tract is primarily centered.

    In that sense, I think of F1 and F2 like magnets that have a certain sphere of influence. Certain notes sound better when influenced by F1 and others by F2. In that sense we can simplistically define F1 as low resonance and F2 as high.

    The singer's formant is important because of its region of influence that coincides with the “sweet spot” of human hearing.

    Particularly for the male passaggio and early high voice and the female middle voice, the second formant is more efficient and does not make the voice lose its inherent color. It does indeed feel foreign when a singer first experiences it. Mastering the acoustic shift used to be a staple part of classical vocal training. Now most tenors I hear do not access F2 in the passaggio and early high range, and most females are spotty on F2 where it could be helpful.

    Keep commenting. I enjoy your input very much.


  3. TS,

    I believe your use of “formant” is perfectly in line with its use in voice science (by Sundberg, Titze, D. Miller and others). “Formants” are simply the resonances of the vocal tract, whose interaction with the sound source are responsible for vowel recognition. But in operatic singing, stronger interaction with the formants are suggested to be crucial for a “resonant” voice. (BTW, f0 has nothing with formants to do, it is the frequency of the fundamental of the note in question. Personally, I prefer the notation H1 (=f0), H2, H3 etc for the “harmonics” of the note that is sung.)

    The question of the acoustic “turning” of the tenor voice is fascinating! With this I mean the acoustic shift you talk about in your blog, that is, when in the spectrum of the voice, there is a shift from a dominating H2 to a dominating H3 or H4 above the passaggio (at say F#4). The consensus seems to be, as you discuss, a shift of resonance strategy from F1 to F2 at the passaggio.

    What I find interesting is that not all tenors, not even some very famous one, display a good control of this register shift! Superstars like Pavarotti, Corelli, Del Monaco and Björling indeed show an excellent ability to perform the “turn”, whereas others do not. I have for instance checked out a few recordings of Bergonzi. He sounds less convincing in this regard (in the recordings I checked) and generally continued with the strong H2 sound above the passaggio. It seems like this technique is not so easy to master!

    I suspect that an ability to phonate with a very harmonically rich source is necessary in order to get F2 to “respond” at all. Higher formants are generally weaker and broader and it is not so much that can be done about that, but the harmonic content of the source certainly can be altered.

    This idea that an ability to phonate “sharply” is beneficiary was substantiated by analyzing the sound of Pavarotti when he demonstrate how to cover (“How to sing Bel Canto 1” on Youtube). First he sings an F4 “uncovered” and then “covered”. What is fascinating is his spectrum for the uncovered F4. It is not a first-formant dominating sound (a la Bergonzi), but a very sharp sound in which H4 completely dominates the spectrum! What he seems to be doing when he covers is to “round” or “filter” this very sharp sound (damping H4) just right so that H3 start dominates the “covered” sound.



  4. Thank you for the excellent observation Martin. It is fascinating that the first clip was actually F2 dominant. In Pavarotti's mind he was singing an “open” [a] when in fact he was singing what might be considered a “covered” [ae] if we take F2 dominance to mean covering.

    What I find is that certain notes need F2 dominance to be balanced, but it is not only F2 dominance but rather F2 dominance with a certain activity in F1 that balances out the spectrum.

    The mindset has been too long about one formant or another dominating the spectrum when in fact it is a balanced spectrum we are looking for in which F1 and F2 have a relationship to each other. A good F4 as in Pavarotti's “covered” example requires a tenuous balance between F1 and F2 dominance. During the length of the note, dominance is exchanged between F1 and F2 on H3 and H4 respectively. There is an ideal formant balance for each note/vowel combination.

    For that reason, the technique is difficult to master. Most tenors I know, indeed baritones as well worry more about their top notes than they do the lower voice that they take for granted. You are right that few tenors master this. Most of the up and coming tenors I hear have not figured this out. Many teachers in the USA do not teach this anymore. It used to be a requirement.

    I agree totally that a strong source tone is necessary for a good quality covered tone. I find that singers do better singing a tone rich in harmonics when they are in the F1 dominated region (tenors F4 and below). It is a good tool to sing F4 open on [a] and then only round the lips in the direction of [o] to accomplish the F2 change. Pay attention not to lose the intensity of the tone as produced for the open version. Below F4 requires greater lip rounding. Above it gradually less. This of course occurs when the open tone is produced with a comfortably low larynx.

    All the best,


Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.