By Joe Armstrong
Boston 2005

An earlier version of this article appeared in
The Flutist Quarterly
Spring 2002

Carl Petkoff 

Carl Petkoff
1925 - 1985



To amplify and enrich certain points in the main text, I have included a number of substantial notes at the end of each section. If you wish to read a section note along with its corresponding subject in the main text, merely click on (N). Click BACK to return to your original place in the text. If you read the main text without skipping to the notes, you'll come to them at the end of each section and can read through them then if you like. Clicking on (S) in the main text leads to source material notes only. Source notes come at the end of the section notes.



Everyone I know who heard Carl Petkoff play thought that he was an extraordinary flutist and an astonishing musician—"a genius," some even said. As a serious flutist myself with nearly fifty years' experience, I have never found any flute playing like his in its total subservience of personality and technical facility to direct musical impulse. I would even go so far as to say that if his poor health hadn't so drastically shortened his career he might easily have achieved renown as one of the finest flutists in the world. I know that's a bold claim to make, but every indication of his brilliance and promise can still be heard in recordings of the last recitals he gave during his late twenties and early thirties (now available on CD). (1N1) Mr. Petkoff (from here on I'll call him "CP") was also remarkable as a teacher in his ability to evoke that directness of musical impulse in his students. I studied with him briefly at summer music camp sessions during high school, full-time from 1962 to 1964 at Illinois Wesleyan University, where he taught, and intermittently after that before his death in 1985; and the more professional flutists I've compared him to, the more I've come to believe that his unique way of producing vibrato and his use of it to enhance the expressive motion of a phrase—as opposed to merely decorating his sound in general—was a crucial and integral part of what made his playing so great.

CP often made suggestions in lessons about using vibrato expressively, but he never tried to describe how to produce it in the subtle way he did. Nor did he seem to want to meddle with his students' vibratos if we already had fairly moderate ones. I think he also believed it wasn't possible to communicate much about vibrato production verbally, maybe because he might have originally discovered his by chance, as other flutists often acquire a more automatic, steady-state vibrato—one day just suddenly "getting it" and considering themselves lucky if it's not too fast and obtrusive (the "nanny goat" type). One of my Wesleyan classmates, however, had this fast kind of vibrato, and CP did become concerned that she find a way to slow it down; but I don't think he ever gave her any specific instructions about how to do this other than by practicing extremely slow, conscious pulsations. She finally managed a change, but, to my ear, what she arrived at was still a steady-state, automatic vibrato, and not like the subtler, more variable one CP used. So, even though he got us to think far more about it than other teachers I later studied with, vibrato still remained a mystery—but also a challenge.

At first I thought that CP could produce vibrato the way he did (as well as make his beautiful, velvet sound (1N2) and articulate so clearly and powerfully) simply because of something unique in the anatomy of his own throat, mouth, tongue, and lips. Then I realized that one of his earlier students, Suan Guess-Hanson, who had studied with him since she was quite young, had learned somehow to produce a vibrato like his (1N3), make the same velvet sound, and articulate his way too. Though I felt that my own vibrato, in spite of being automatic, was fairly acceptable, (1N4) I yearned to understand CP's secret and produce his kind of vibrato because I was so in awe of the way he played. Before long I saw that there was clearly a "Petkoff style," or "school," of playing, and I was determined to become a full-fledged representative of it—no matter how long it took. Eventually, with the help of studying the Alexander Technique very intensively (1N5) and after many years of careful exploration, I believe I discovered how to produce vibrato his way—not from the "diaphragm" (rib cage/abdomen) or from the "throat" (glottis/larynx), where most vibratos are commonly produced, but from higher up, around the back of the tongue and the soft palate.

The "Petkoff vibrato," generally speaking, is probably closer to what some people today might identify as "an intensity vibrato" than it is to a pitch-varying one (1N6), and I'd like to describe in detail here how to develop it, insofar as words can convey the actual operations and experiences involved, in case you might like to try using it too. But first I want to tell you more about CP himself and his teaching approach, because they were both as unique as his flute playing.

The following sound files show how CP's vibrato changed over time, starting with the Delmas Incantation where his vibrato was generally faster than on the Tomasi cadenza, and, finally, extremely flexible and variable in the Prokofiev andante.

Click here to listen to Carl Petkoff playing Delmas' Incantation

Click here to listen to Carl Petkoff playing the cadenza of Tomasi's Concerto en Mi.

Click here to listen to Carl Petkoff playing the Andante of the Prokofiev sonata.


1N1. Carl Petkoff's recital recordings available on CD:

Carl Petkoff: Flute Recital Selections (1950-1960), produced by Joe Armstrong, available in this website. The two-CD set includes works by Tomasi, Prokofiev, Telemann, Ibert, Delmas, Handel, Piston, and Hindemith. These selections were transferred to CD from reel-to-reel copies I made in 1969 of CP's original tapes of the live performances. I had them professionally re-mastered to eliminate as much as possible the surface noise caused by the older recording equipment, but there is an unavoidable loss in fidelity and some altering of speed as a result of the copying, which is especially noticeable in some of the earlier performances. The final movements of the Prokofiev sonata and the Ibert concerto are not complete, but enough remains of the beginning of each to give a clear sense of CP's full rendering of them. (Information and Order Form.)

I don't feel the essence of CP's sound really comes through on his recordings, but you can certainly hear that it's very pure and unforced, with no extraneous breath and no attempt whatsoever to focus it with the so-called "open throat" approach to tone production. I remember CP telling me that the sound of a wood flute was the model he thought we should still follow, even though we have the possibility of a creating a more projectable "edge" with a metal flute; and I always felt that the "velvet" quality I found in his sound was due to his holding the wood flute sound ever in mind. When he lent me his recital tapes to copy in the late 60s it had been over ten years since he had made them, and I doubt if he'd really listened to them much, if ever— particularly because we'd found them stashed away in the back of a closet and he didn't own a reel-to-reel tape recorder to play them on. When I returned the originals to him, we spent a little while listening to them on my machine, and he immediately exclaimed, disdainfully, "I sound like Kincaid!" So even he didn't think the recordings represented his sound faithfully, for, in reality, it was nothing like Kincaid's.

Kincaid, of course, played a platinum Powell, and I often thought that CP's sterling silver Haynes (circa 1940) had a lot to do with the sound he made. In fact, when I eventually came into possession of a Haynes of this vintage myself in 1969, it helped me get much closer to being able to create CP's kind of sound than I could on any of the later-made Hayneses and Powells I had played up to then. But, as I describe later, there was still a lot more to discover about CP's approach to embouchure than I had imagined.


1N2. CP's beautiful, velvet sound:

On reading the earlier, published version of this article, Joyce Wilson, a former student of CP and professor of flute at Indiana State University, wrote me: "I've been haunted by the memory of that sound my entire life. I think about it a lot and sometimes believe that it must have been a youthful dream. I've had many wonderful pedagogical teachers—but no one has ever sounded as he did."


1N3. Suan Guess-Hanson had somehow learned to produce a vibrato like his:

I recently asked Ms. Guess-Hanson if she could remember CP telling her anything specific about how to produce vibrato. She said that she couldn't, and that she thought she somehow just picked it up from listening to him as if by osmosis. But, since she began studying with him in high school, I imagine he played for her a good deal more in her early lessons, simply to demonstrate, than he did for us as college students. She also had more of an opportunity to hear him play in solo recitals during the years while he was still in fairly good health; so it seems likely that he would have had much more of a direct effect on the foundation her vibrato than he did on the rest of ours which were already well-established when we came to study with him at Wesleyan.

The only opportunity I had to hear CP in a solo performance was during high school in a faculty recital at Wesleyan's two-week summer music camp. But I was so stunned by the overall expressive effect of his playing of the piece—I think it was by Couperin—that I couldn't be aware of any details such as his vibrato production, even if I had had the ability then to discern them. Later, while studying with him at Wesleyan and playing in the flute section of the Bloomington-Normal Symphony, I heard CP in various orchestral solos, notably Brahms' first symphony and Piston's The Incredible Flutist. His playing of these was always mesmerizing.


1N4. I felt that my own vibrato, in spite of being automatic, was fairly acceptable:

I can't quite remember how I began to produce a vibrato of my own. I think it was sometime during junior high school, and I really didn't have anyone to imitate (consciously or subconsciously) since none of my early teachers were accomplished flutists themselves. But I think I realized then that if I was going to become a full-fledged flutist I would need to use vibrato, and somehow I just happened onto a way of doing it as I listened to my trumpet-, trombone-, and saxophone-playing friends starting to make theirs. I do remember, however, being quite certain that I didn't want my vibrato to be fast and nervous-sounding, like the vibrato of John Wummer, whom I'd heard often on TV in Leonard Bernstein's Young People's Concerts with the New York Philharmonic.


1N5. Studying the Alexander Technique very intensively:

I was introduced to the Alexander Technique in 1965 at the National Music Camp by Alex Murray and his wife Joan. (Mr. Murray was then principal flute in the London Symphony Orchestra and eventually became professor of flute at the University of Illinois.) Besides helping me to achieve a more consistent access to my total musicality, the Alexander Technique helped me with control of nervousness, breathing, and flute balance. It also gave me greater accuracy in assessing the best kind of effort for any action (particularly in my neck-head-torso coordination while using my voice to speak or sing), a more reliable ideo-kinetic muscular control, greater power to "leave out" unwanted actions and thoughts, and a fair knowledge of anatomy—all of which were vital in the careful vibrato explorations I began to make. (Alexander study also helped me to learn to swim, get through Army basic training, and overcome a phobia of birds; but those are other stories.) Since completing the three-year training to become a qualified teacher in 1972, I've been teaching the Alexander Technique in Boston, where I've specialized in working with professional musicians. (See my master's thesis in this website: "The Effects of the Alexander Principle in Dealing with Stress in Musical Performance." Tufts University, 1975.)


1N6. Pitch and intensity vibrato:

In The Flute Book (Oxford University Press, New York, 1996. p. 106), while Nancy Toff considers that there are only two viable ways of producing vibrato-"throat" and "diaphragm," she also claims "There are three basic types of vibrato: pitch, intensity (dynamic), and timbre." And John Wion, in his website on flute vibrato ( examines a number of well-known flutists' vibratos at regular speed and then at the same pitch at 300% slower speed. He looks at both the speed of their vibratos and the degree of pitch variation, pointing out that some vibratos "vary from an almost imperceptible pitch variation (the variation is largely in intensity) to a full half step." Some of the widest in pitch variation vibratos seem to me to be "diaphragm" types (Aurèle Nicolet and Ransom Wilson); but others that seem to be "throat" types (Jeanne Baxtresser and Carol Wincenc) vary widely in pitch too. A friend recently examined CP's vibrato at a slowed-down, on-pitch speed, and it proved to be much more of a variation in intensity than in pitch.

To my ear, most of the vibrato samples Wion includes (with, to a degree, the exceptions of Baker's and Kincaid's) don't come anywhere near the Petkoff vibrato in subtlety of production. The majority seem generally to be imposed upon and superficial to the essential expression of a phrase in comparison.

And I think it's important to add that when considering any distinctions between "pitch" vibrato and "intensity" vibrato, it stands to reason that if you conceive of your own vibrato as essentially a variation in pitch, then you'll be much more likely to bring a greater pitch variance into producing it than if you decidedly don't mean for it to be a pitch-varying activity. Then, too, the concept of "intensity vibrato" could also have an exaggerated influence if approached with too much effort. (Some written references to "intensity vibrato" as it's used on other instruments—like bassoon and saxophone—consider "intensity vibrato" to be a vibrato that's "too intense," mainly meaning "too fast.")



Originally from Steubenville, Ohio, CP attended the Cincinnati Conservatory as a student of Alfred Fenboque, first flutist of the Cincinnati Symphony; and he also spent time at Chautauqua and at Tanglewood, where he coached with Georges Laurent. But his study at the Conservatory was interrupted by World War II, during which he served in the Great Lakes Navy Band, Orchestra, and Choir; and when he returned to finish his degree, he developed a form of epilepsy that would ultimately prevent him from pursuing a full-fledged orchestral or solo career. However, after graduating in 1950 and having performed with the Cincinnati Symphony, he received a teaching fellowship to do his master's degree at Illinois Wesleyan School of Music, where he stayed on as a faculty member until he was forced into an early retirement in 1964. He also gave regular recitals there and was principal flutist in the Bloomington-Normal Symphony.

CP was a striking-looking man, around six feet four with black hair, long black eyebrows, and a short mustache. Often his way of doing things could seem eccentric if you didn't know him well, and I think this might have caused some people to be a bit intimidated by him on first meeting; but as a teacher he was very unassuming and kind. Studying with him meant much more than merely having flute lessons. In many ways it was like taking a philosophy course, or even lessons in life itself. Not that he tried to preach or pontificate, but he always wanted us to see that as musicians it's just as important to be conscientious, sensitive, and humble human beings as it is to master our instruments, since our whole attitude toward life and appreciation of the world are reflected in the way we understand a work and perform it. There was simply no place for egotism or pretentiousness in his book, and, by citing examples of those traits in certain performers, he made it clear that if we ever behaved arrogantly or falsely—especially in a rehearsal or performance—he would disown us as students. So, for some of us, he wasn't just our teacher: he was also our mentor and guide.

Often CP would begin a lesson by talking for quite some time, telling about something he'd done or seen since we last met—all with keen discernment of the quality of feeling or character involved in the experience or object. I think some students might have been a bit impatient with this and preferred to get right to the music, but I was always captivated by his observations and comments, which could be about anything from piloting airplanes—another passion his health hadn't allowed him to pursue—to great musicians he had heard or played with, like his friend Frank Miller, first cellist of the NBC and Chicago symphony orchestras. Many times we would completely lose track of time, and a lesson would run way past the hour before we actually started to focus on the music, but, as soon as we did, I always found myself transported into a realm of expressive involvement that was exciting beyond belief.

This heightened expressivity usually lasted for a couple of days, but I never completely understood how CP brought it into being in me. I've had other teachers and conductors who would try to express the energy of a piece in the way he did, but they never seemed to connect so deeply with my musical imagination. The only other time I experienced anything like I did working with CP was when I watched Pablo Casals teaching a young cellist in a master class at the Marlboro Festival in 1968. But as best as I could figure out, CP's ability (and perhaps Casals' too) (2N1) to transport a student so fully into the essence of a piece came somehow from his extremely kinetic way of responding to each figure or phrase as it sounded ideally in his imagination (2N2) while we played for him. This seems all the more likely since he rarely played for us himself during a lesson—although when he did, that of course had a powerful impact too.

With his kinetic responses, CP also combined vocal characterizations that weren't actually sung—even though he would sing a phrase to us from time to time too. For instance, he might start by whispering the softer notes at the beginning of a long, busy crescendo that builds into a big sforzando accent, then gradually increasing the notes' intensity with a kind of growl that would open into a great roar, and finally end by shouting a loud "POW!" on the last, sforzando note. He always matched these vocal sounds exactly with gestures and facial expressions, and sometimes he might even "dance out" a passage if he felt that might draw us more fully into the character of the piece. I could never resist being completely entranced by these portrayals, and often when I played such a passage back to him afterward, I'd be astounded by how much it was transformed from whatever earlier meager offering I'd produced on my own steam.

Suan Guess-Hanson echoes this experience in writing about studying with CP:

I think that in working with Mr. Petkoff, I had a total body experience — it wasn't just the sound the flute made, it was a whole aura to be felt in one's bones. Petkoff's gifts to his students included a love of the "feeling" of music, the physical sensation of being "moved," excited almost to the point of being overcome by emotion—on the edge, perhaps. He often sang phrases and it was the musicality that we student sponges soaked up; we emulated what he said and played for us.

There was nothing more gratifying than acknowledging together with CP the moment I'd made this transition into that fuller realm of expression. Then he'd often say something like, "Now, you're really cooking with that piece, buddy!" And if I didn't manage it very well, he could seem quite crestfallen, yet no less intent on giving me hope that I still might be able to do it—someday. It wasn't that he had any less concern for technical facility though. On the contrary, he expected us to work very hard at it—but never as an activity separated from the fullest expression, even in the most repetitious studies, as was always the case with his own playing. (For instance, it was spellbinding to hear him play just a few bars of an Andersen étude.)

One of the best examples I can remember of CP's natural brilliance for connecting kinetically to a musical impulse and transmitting it to us is not from a private flute lesson but from a concert I was playing with the Army Field Band in Bloomington on our tour of the midwest in 1969. He and his wife Jeanne (a fine clarinetist and teacher herself; in fact, my first flute teacher) had told me they were coming—more out of kindness to me, I'm sure, than out of any special interest in hearing a band concert—so I knew they were in the audience, although I hadn't actually seen where they were sitting. But when the band came to the end of the program and was playing, as always, The Stars and Stripes Forever, we three flutists, of course, took our piccolos, stood up, and came out to the front of the stage to play the famous obbligato. Just as we got into place, I happened to glance out over the audience, and through the open doors at the back of the hall I noticed CP in the foyer, where he must have gone to stretch his long legs and smoke his pipe. Not seeming to be paying attention to us (although I knew he always heard every note), he ambled casually over into the middle of the doorway; and just as we were about to start the obbligato he looked up to the stage, snapped to full military "attention," and gave us piccolos a little flick of a salute right in tempo, sending us on our way "with colors fully flying"!

That gesture, both serious and playful, so completely fit the bravura character of the music that I instantly felt a bolt of adrenaline shoot through me. I have to admit that, even though I'd been playing Stars and Stripes night after night for nearly three years, I still found it especially exciting in contrast to most of the other music we usually played. Yet this time it became almost unbearably thrilling just because of that single, lovable "cue" from the lone silhouette of this dear and brilliant man, way at the back of the hall and probably unnoticed by anyone else. I don't know if CP figured that I'd spot him out there, but he certainly brought back the old magic for me that I'd so often felt in lessons. I doubt if he meant anything more by saluting us than merely responding in his usual total way to this special part of the music that he too must have had to perform countless times in his Navy years, but I never played that grand old march again without bringing to it that night's extra charge.

I hope this shows a little of why I feel it was such a great privilege to know Carl Petkoff and to be inspired by his grasp of music and his view of the world. I offer this brief account of him and the following explorations of vibrato as a tribute to his wonderful gifts.


2N1. The only other time I experienced anything like I did working with CP was when I saw and heard Pablo Casals teaching:

See my "Introduction" to 'Just Play Naturally'(Duende Editions, 2002) in this website—my conversations with cellist/Alexander teacher Vivien Mackie on her three-year study with Casals in the early 1950s and her discovery of the resonance of his teaching with the principles of the Alexander Technique.


2N2. CP's extremely kinetic way of responding to each figure or phrase as it sounded ideally in his imagination:

Over time, with the help of the Alexander Technique, I gradually began to bring myself back on my own to the same level of aesthetic involvement in my playing that I'd experienced with CP in my flute lessons. I give recommendations for cultivating the ability in my article "Musical Vision," also in this website.



In my quest for understanding CP's secrets, (3N1) I've attempted to look at flute vibrato in the larger context of how vibrato is produced and used throughout the classical music world—not just by other wind players, but by singers and string players too. To a certain extent, I've also tried to listen closely to the vibratos of popular, jazz, and folk musicians. (3N2) Surveying this broad spectrum has definitely made me more aware of the ways we flutists produce and use our vibratos. I've also had many candid conversations with string players and non-flutist wind players about their reactions to many current flutists' vibratos, which the non-flutists often find so exaggerated that they feel it draws attention away from the expressive essence of whatever piece the flutist plays. For the most part, I find, these flute vibratos fall into the constant, or steady-state variety, and they are usually either of the rib-cage/abdomen "diaphragm") or the glottis/larynx ("throat") type.

Many continuous flute vibratos sound similar to what singers and singing teachers call a "wobble" in the singing voice: a constant vibrato that's usually very wide. Sometimes you see the singer's entire jaw and mouth opening and closing—or even their whole head nodding up and down—with each vibrato pulsation. I heard a prime example of the wobble vibrato in a recent television broadcast of Tristan und Isolde from the Metropolitan Opera. Both lead singers had enormously wide vibratos that distracted you very much from the musical line and its expressive import. In stark contrast though, the long English horn solo at the beginning of Act III was played that evening with the most refined and expressively nuanced use of vibrato you could ever hope to hear. It made the singing in the rest of the act seem all the more superficial and tedious.

A striking example of a "wobble" in flute vibrato can be found in a recording of Otto Luening's "Trio for Three Flutists" (Composers' Recordings #561, 1993). In one movement, the flutists actually synchronize their slow and wide vibrato pulsations on a number of longish notes. When I first heard this, it made me feel as if the walls of my room were alternately shrinking and expanding with the combined flute pulsings. I thought maybe Luening had written it in as a special effect, but when I eventually saw the score, I found no indication for it.

I often wonder why so many more flutists than other wind players have adopted such prominent and exaggerated vibratos, and sometimes I think it comes from the strong influence of orchestral string players, who generally use a constant, whole-forearm/wrist/hand/finger-waving vibrato action mainly to create an overall shimmer in the sound of the section as a whole, as distinct from some string soloists and chamber players who, like Pablo Casals, contain their vibrato more within their hand and finger action and vary it more expressively according to the emotional demands of a phrase. (3N3) In general, singers also seem to use a constant vibrato mainly to animate their sound in a general way. Nancy Toff says, in The Flute Book, that one school of thought considers flute vibrato "a natural part of tone production" (p. 107) as some people claim vibrato to be an inevitable result of natural development in the adult singing voice, believing that singers would have to do some degree of tightening to keep their vibratos from happening if asked to sing without it. This "nature versus nurture" view about vocal vibrato could certainly be used to back up an unquestioning acceptance and encouragement of the steady-state flute vibrato. (3N4) However, the fact that many clarinetists don't ever use vibrato seems to give us grounds for an alternate view. Whatever the exact evolution of and reasons for using a continuous vibrato, I feel that much current flute playing could benefit from more careful examination of the many great instrumentalists who use vibrato expressively and sparingly—and sometimes not at all. (3N5)

The factor of imitation—conscious and subconscious—needs to be taken into account too, as F. M. Alexander pointed out many years ago (S1) when describing how we often go about learning and developing a skill by trying to copy someone who is an "expert" at it. Alexander found that, because of our lack of understanding about what constitutes an integrated use of ourselves as a whole, we often copy only the superficial features of the expert's performance without any knowledge of how to recreate the central integration which lies at the core of what makes that person's playing great. Surely, many flutists—and other musicians too—merely try to mimic the vibrato that's in vogue when they are first learning how to produce their own. If the steady-state, extraneous vibrato is mainly what they hear in their teachers' and famous professionals' playing, then they probably feel compelled to try to duplicate it by whatever means they can, instead of considering if there might be a larger array of possibilities for influencing the air column and then consciously choosing from them what best suits the note or phrase at hand. (I will show how to make an exploration of this broader range of options in Sections V and VI.)

Sometimes I think the current wide-spread use of steady-state vibrato goes hand in hand with the tendency for many flutists—in contrast to most other wind players—to play way too fast, as if impressing listeners with their technical prowess is more important than expressing the essence of the music they play. For example, a conductor I know recently told me about meeting to discuss tempos with a flute soloist who was scheduled to perform one of the standard Baroque flute works with his orchestra. My conductor friend was appalled by the speed at which this flutist insisted on playing each movement of the piece, leaving little or no room for the expressive nuances that most other instrumentalists would automatically allow for in a work from that period by choosing more reasonable tempos. "Fast, faster, fastest!" seems to be the order of the day. (3N6)

Oboists, on the contrary, often produce and use vibrato in a way that much more closely resembles CP's. (3N7) Among singers, Dietrich Fischer-Dieskau comes to mind—especially in his Bach and Schubert performances. Of course, as I mentioned earlier, there are more obviously differing approaches to vibrato among string players, and sometimes I think CP's model for flute vibrato might have come from the more varied vibrato heard in some chamber and solo string playing—particularly like that of Casals, in which the vibrato seems to happen inside the sound (3N8) and can emerge from a straight tone and be carried through slow and moderate to faster and more intense undulations according to the momentum and emotional power of a phrase. CP once told me that his teacher, Alfred Fenboque, had actually been a violinist before he became a flutist; and I've often thought that Mr. Fenboque's teaching must have reinforced CP's attitude of using a more variable vibrato too. And his playing with cellist Frank Miller during their Navy years together may have had an additional influence on CP from the string player's point of view too.

With these general points in mind then, I'd like to review some of the published perspectives on flute vibrato of the last 75 years or so and examine them in light of my understanding of the "Petkoff vibrato." I hope that doing so may help those of you who would like to think freshly about your own playing have a clearer basis from which to make the vibrato explorations in sections V and VI.


3N1. My quest for understanding CP's secrets:

Along with investigating vibrato production, I've also searched deeply into the elements of tone production and articulation, and I think I've discovered how to focus my sound in the same way CP did: it takes a far more vigorous downward direction of the central facial and lip muscles than I had ever considered feasible, while keeping the corners of the mouth and the cheeks going in an upward direction. I've also found out how to use my tongue in single, double, and triple tonguing to achieve the same sharp and definitive attack CP used, particularly on accented notes. It involves the tongue being "exploded" away from contact with the upper gum, rather than the tongue coming from a no-contact position and touching the gum as a mere interruption of the air stream. It's very similar to what Larry Guy describes in his book on the teaching and playing of the great clarinetist Daniel Bonade: "Bonade believed that the tip of the tongue, rather than hitting the reed for articulation, should be lightly placed on the reed, the air built up behind it, and then moved quickly away from the reed a short distance to allow the note to sound. So one starts a note by placing the tongue on the reed before blowing, and one never 'hits' the reed with the tongue in a hard or percussive manner, resulting in an abrupt stop of vibrations. So the tongue, rather than being a muscle that hits or pushes [the reed] is an organ of sensation, which . . . feels the pressure of the air behind itself with great sensitivity." The Daniel Bonade Workbook, Rivernote Press, Stony Point, NY, 2004, p. 35.


3N2. The vibratos of popular, jazz, and folk musicians:

One folk singer's vibrato I particularly noticed was that of Joan Baez. Her vibrato was so rapid in her early career that I could never enjoy listening to her then. But I was surprised to find that later on, in the 80s or early 90s, she found a way to slow it down. Around that time I happened to read an interview with her where she talked about her vibrato, and she said that when she began singing she didn't actually have any vibrato at all. She felt so envious of other singers who did that she resolved to stand in front of a mirror, jiggling her Adam's apple (thyroid cartilage) up and down with her fingers and thumb while she sang, until she could make the pulsations with her throat muscles alone. Then she told about how, much later, she began working with a vocal coach who eventually helped her to tame the fast vibrato down quite a bit.


3N3. The many string soloists and chamber players who, like Pablo Casals, often vary their vibrato more expressively according to the emotional demands of a phrase:

Richard Dyer, senior music critic for the Boston Globe, writes of the Beaux Arts Trio's recent 50th anniversary concert at Tanglewood: "The string playing by Hope and Meneses was superb, dead-on in intonation, blended and contrasted in ensemble. Vibrato came and went at various speeds and degrees of intensity. It was a spice that perked up elements of the music, not a gravy that enveloped and masked it." (July 19, 2005). On the other hand, a week later, Dyer wrote of Pinchas Zukerman's performance there of the Beethoven violin concerto: "At 57, Zukerman still operates at a high instrumental level, although you must have a tolerance for throbbingly intense and unvaried vibrato to enjoy his playing as much as he wants you to." (July 26, 2005.)


3N4. An unquestioning acceptance and perpetuation of the steady-state vibrato:

I think some untrained listeners are often more aware of excessive vibrato than we musicians realize. Though they may not be able to speak about it in an informed way, non-musicians nevertheless frequently realize that there's something about an instrumental or vocal vibrato that makes a performance less congenial than it might be. Once, in the 1980s, when I began to discover how to create the Petkoff vibrato, I was very surprised to find out how much more a non-musician was aware of vibrato than I had expected. I was staying for a few days in Philadelphia with my dear friend, the Alexander teacher Kitty Wielopolska, who was then in her 80s and had never studied voice or played any instrument. One day I decided to do some practicing while she was busy in another part of the house. Afterward, she told me how much more she enjoyed my playing than that of other flutists she'd heard, and she added, "And I think you've really got the 'warble' just right!" I was astounded to hear her say this, especially because I'd never mentioned to her that vibrato was something I'd been working on so diligently—assuming that she would neither know what I was talking about nor have enough listening experience to recognize such seemingly subtle differences.


3N5. Great instrumentalists who use vibrato expressively and sparingly—and sometimes not at all:

A friend recently sent me a quote from "The Flute List" commenting on the vibrato of the flutist Fred Selden, who plays in the sound track of the film March of the Penguins: "I couldn't help but notice his flawless use of vibrato . . . often very little or none, and when he did use it, he used it purely for color, for expressive touches, constantly and artfully varied. For a professional lesson in effective use of vibrato (not to mention great flute-playing generally) I can't think of anything I've heard lately I'd recommend more." (Charles Andrews, Los Angeles, July 17, 2005.) On seeing the film myself, I certainly found the flute playing to be quite lovely, and I also felt that, especially because of the reverberation that was added to the entire music soundtrack, a constant or too wide vibrato would have severely disrupted the underlying serenity of the music. It was striking, too, how the flutist seemed to mirror so well the only other solo wind, bassoonist Rose Corrigan, in her exquisitely subtle vibrato coloring, often emanating from notes begun with no vibrato at all.

A few weeks later, this same flutist friend sent me a significant comment of his own about listening to a recording of a well-known flutist who "was vibrating so much that, in fast passages, you'd hear two or three little stuttering notes where there was really only one. The interruptedness of the playing (for want of a better word) was what bothered me more than anything else." August 2, 2005


3N6. Fast, faster, fastest! Seems to be the order of the day:

The Spring 2005 issue of The Flutist Quarterly notes that the recent Chicago Flute Club Flute Fair featured a "Chicago's Fastest Fingers" concert. ("Across the Miles," p. 13.) I've addressed this issue of flutists' gratuitous speed in an article I wrote on my two-year study with the late, former first oboist of the Boston Symphony, Fernand Gillet—particularly noting my work with him on reining in my tempos in the Bach sonatas and the Mozart D major concerto, which, of course, is the same as the C major oboe concerto. I alos compare tempos of various well-known flutists' and oboists' recordings of this concerto. ("Oboe Master Fernand Gillet's Legacy to Flutists: His Methods for Developing Superior Technique and Expressive Control," The Flutist Quarterly, Winter 2004. Expanded version in this website.)


3N7. Oboists sometimes use vibrato in a way that resembles CP's:

The most recent example of this that I've heard is in the playing of oboist Alan Vogel in his recording Bach's Circle, Delos International, 1998.


3N8. Like that of Casals, in which the vibrato seems to happen inside the tone, as if alternating more in intensity of volume than in pitch:

Vivien Mackie points out that "as Casals was quick to say, quick to imply, there's really no such thing as a long, plain note. A note is always either going or coming; so there was no place for a long straight note—and yet that's what we're traditionally encouraged to practice." ('Just Play Naturally', p. 25.) A good example of Casals' varied use of vibrato in this way is found in his recording of the slow movement of Bach's first gamba sonata, now available on Sony CD SMK66572. Mackie teaches cello vibrato by likening it to gently polishing a pearl with only the tip of your finger where it contacts the string.


3S1. Constructive Conscious Control of the Individual, Integral Press, Kent, England, 1923, "Imitation," Part III, Chapter II, pp. 158-160. Available through Mouritz Ltd., London: uk



As I've said, CP didn't try to explain the exact way he produced vibrato but very frequently suggested varying the speed or intensity of it for expressive reasons. For instance, in The Afternoon of a Faun he recommended starting the opening solo C sharp without any vibrato at all, and then allowing a pulsation to come in gently, possibly increasing its speed a little as the note moves toward the B natural of the following sixteenths. You can do this even more effectively in the later repetitions of this passage where the beginning note is extended much longer, and where, if you want, you can also bring the vibrato back into to a straight tone for an instant before moving off the long note, so as to let it seem to linger slightly before moving to the sixteenths. In CP's own playing I often had the impression that certain parts of a figure or phrase were hovering magically in a kind of suspended animation, even though the strictest beat was continuing underneath them; and, in retrospect, I think this was partly due to a tapering off of his vibrato just before he moved on to a more energized note or series of notes—and maybe also due to the greater flexibility in legato he had because his vibrato wasn't distorting or jolting the air column as much as I believe the "diaphragm" and "throat" vibratos do. It was heart-stopping when this hovering happened. (A good example of this is found in his playing of the andante of the Prokofiev sonata, which can be heard by clicking on the link at the end of Section I.)

CP often brought my attention to vibrato at the end of pieces that finish on a long note with a diminuendo, as so many do. There he might suggest letting the vibrato gradually slow down and then taper off into a straight tone to enhance the feeling of coming to rest. But he never made such recommendations into hard and fast rules. He just offered them as possibilities for bringing the feeling and motion of a particular passage closer to its fullest expression than a steady-state vibrato allows.

Even if CP never believed in making anything formulaic, it was very clear that he considered a constant, too fast, or too wide vibrato neither tasteful nor appropriate. For example, as we worked on The Afternoon of a Faun, he also described how many flutists immediately begin the solo with a steady-state vibrato and continue to pulsate that same way throughout the piece on every note longer than a sixteenth—seeming more like a thoughtless habit than a sensitive musical choice. When the solo is played with the most blatant of these continuous vibratos, he said it reminded him more of a "bad night of a neurotic bird" than the languid summer afternoon Debussy's faun was surely meant to be having!

Over the years since I studied with him, I've spent a good deal of time listening to CP's recital recordings and comparing his vibrato quite carefully with those of other flutists on recordings of some of the same works (particularly the Prokofiev sonata). And it's pretty clear that none of the many other flutists I've listened to produce vibrato the way he does. Nor do most of them vary theirs much expressively as he suggested. (4N1) On the whole, they use a constant vibrato everywhere they have a sustained note, with some difference from flutist to flutist in number of pulsations per beat. (4N2) The only players I feel come anywhere near producing vibrato as CP does are: Julius Baker, in his exquisite 1955 Decca recording with Lillian Fuchs and Laura Newell of the Debussy Sonate for flute, viola and harp (2nd movement available on Special Retrospective Collection, The Manhattan Flute Center); Georges Laurent, in his recording of Hanson's Serenade for flute, harp and strings with the Boston Symphony (RCA Red Seal LM2900) (4N3); and James Pappoutsakis, in his recording of the Telemann Suite in A minor with the Zimbler Sinfonietta (Decca DL8522), and in other solos he recorded with the Boston Pops, such as Piston's Incredible Flutist. But I think these three flutists resemble CP in their playing chiefly because they focus their sound more with their lips than with their throats, which also lends itself to a less severe vibrato that can be somewhat more responsive to expressive demands—but not much, since theirs are still vibratos produced primarily in the glottis/larynx ("throat/voice box") area and are therefore harder to vary because the muscular structures in that mid-throat area tend to cut so sharply into the air column.

It's important to point out here that the glottis is the opening between the two vocal cords in the larynx. You can hear and feel this opening tighten and close off completely right when you start speaking (or whispering) the vowel sounds at the beginnings of these words: "eel," "aim," "eye," "oh," "ooze," "under," "ever," "all." Voice teachers call what happens when the glottis completely closes at the start of each of these vowels a "glottal shock," or a "glottal attack." (Notice the difference in sound and feel when you don't have a glottal shock as you speak these same vowels right after a consonant: "heel," "late," "night," "low," "choose," "lumber," "never," "tall.") In many flutists' "throat" vibratos, you can often hear the effect of glottal shock, or glottal friction, happening around the air column, which sometimes also gives a raspy quality to their tone because they are, in fact, creating a whisper sound in the larynx at the same time they are producing their flute tone at the embouchure plate. This whisper sound in their throat can also lead to extra, wispy noise at the lips as their breath rushes through their embouchure because these sharp glottal vibrato contractions produce such strong and jolting bursts of air that restraining them requires extra effort at the lips just to maintain a focused tone. But, as I show later on, this extra noise at the lips is just as likely to occur with "diaphragm" vibrato pulsations (4N4), because they usually also create fairly strong thrusts of breath, compared to what happens with the subtler and gentler "Petkoff vibrato."

Though I haven't made an exhaustive survey of writings on vibrato pedagogy, the many flute recordings I have listened to and the discussions I've had with a number of prominent flutists who've studied the Alexander Technique with me over the last 30 years have given me the strong impression that not much has changed in the knowledge of flute vibrato production and the approaches to teaching it since William Kincaid, Marcel Moyse, Robert Willoughby, and Clarence Kelly wrote and talked about it between the late 1930s and early 1960s.

In John Krell's Kincaidiana (notes of his study with William Kincaid from 1939 to 1941), Kincaid pretty much backs up CP on vibrato's expressive function:

Vibrato is a wonderfully expressive tool when used with taste and discretion. Superficial to the tone itself, the frosting on the cake, it nevertheless adds a lyrical quality and an element of freedom to the flow of sound. Applied intelligently and incorporated in the body of sound, it becomes a very distinctive and individual part of the player's expressive resources.

In practice, however, the vibrato is more noted for its abuse. All too frequently the mechanical, vibraphone type of quaver is superimposed indiscriminately (and continuously) to the extent that the tone becomes all frosting and no cake.

To complicate the situation, there is also a kind of mystique about vibrato, some claiming that, with the production of good, supported sound, it occurs like spontaneous combustion, or that an angel kisses you on the forehead and suddenly there is vibrato! It probably is intuitive (we are surrounded by examples), and most likely evolves naturally through imitation of a teacher or a favorite performer. Yet, since it is so personal, many instructors get very evasive and mysterious about its production, hesitating to commit themselves to a specific method of production. The consensus, however, is that it should be a shallow, controlled and even undulation of sound, avoiding the nervous, automatic and uneven type of shaking. It is most probably produced by a combination of the delicate vibration of the throat and the elastic reinforcement of the diaphragm, acting together and sympathetically. [My italics.]

The conscientious musician, like a good string player, will analyze, practice and develop a repertoire of vibrato speeds, contours, amplitudes, intensities and pitch variations, each style subject to the implications of the music being performed.

Ideally there should be consensus of vibrato style in each section of an orchestra or ensemble; it takes only one instrument with a machine-gun or heart throb vibrato to destroy the blend of an entire section and disconcert the tuning. (4S1)

This is very well stated; nevertheless, I find Kincaid's vibrato to be of the more limited, "throat/voice box" (glottis/larynx) variety—though certainly one of the least obtrusive of those and varying much more in intensity than pitch. (4N5) It's too bad that Kincaid didn't give any suggestions about how to create or change vibrato other than calling it "the delicate vibration of the throat," since I understand that he altered his own in the 30s or 40s from a rapid, constant one (presumably influenced by his study with Barrère, whose vibrato was very fast) to one with slower and more variable undulations so that it would be more compatible with the rest of the woodwind section in the Philadelphia Orchestra. (4S2)

In 1960, Moyse more or less reinforced Kincaid's perspective:

A vibrato appropriate to the general sense of the phrase, intensified on certain notes or passages; suppressed or discreet on others; according to the fluctuations of the said phrase; is as necessary to the interpretation as that employed judiciously by a good actor to convey a poem, narrate an anecdote or to relate a dramatic story.

When I talk of vibrato, it is obvious that I do not include the sort of oowah-oowah (fake vibrato) more or less precipitated on a note and of its systematic application to each note in the hope of rendering the entire phrase more expressive. There are certain notes, even passages, that do not need it. This poor kind of vibrato is called "cache-misère," which means that it hides the insufficiency of the player. (4S3)

Moyse also adds that in his earlier years, because of his "fiery temperament and enthusiasm," his vibrato was sometimes excessive and colleagues "warned him about it." And when asked "How do you produce vibrato? From the throat, lungs, diaphragm?" he could only respond "When Faust declares his love to Marguérite, when Pelléas says to Mélisande, 'I love you,' when a mother says to her child, 'Mon enfant chéri,' do they measure the number of vibrations according to the ardor of their sentiments?" All this seems compatible with what CP recommended, yet when listening to Moyse's recordings, I also find that his vibrato, like Kincaid's, falls into the glottis/larynx ("throat/voice box") category and is quite constant (his son, Louis', even more so).

In The Instrumentalist, in 1957, Robert Willoughby unequivocally advocated "diaphragm" vibrato;

Throat vibrato has as its by-product an impeding of the air column, as a result of tightening the throat. Consequently there is a loss of projection and, at its worst, a sort of nanny goat effect. In this method, too, there is an insufficient range of intensity.

Diaphragm vibrato, when properly executed, proves the ideal solution. It is extremely flexible, can be easily controlled, in no way interferes with an open, free throat, and aids sound projection. (Reprinted in Flute Talk, January, 1992, p. 24.)

While Willoughby echoes Kincaid and Moyse in saying, "vibrato is an expressive device linked with one's feeling for the music being played, and as tension mounts, the vibrato will naturally increase in speed. Conversely, in a quiet, passionate section, a slower vibrato is more appropriate," he doesn't give any explanation of how you actually go about producing a "diaphragm" vibrato, except by saying that "producing a vibrato is simply a matter of blowing alternately louder and softer while maintaining . . . [the] embouchure constant . . . and an unchanging position of the head." Otherwise, he only recommends practicing a gradual increase and decrease in number of pulsations per beat, as do Kelly and Toff (see below). Although the closing of the glottis is no doubt what Willoughby objects to as a "tightening of the throat" in "throat" vibrato, his own "diaphragm" vibrato (heard in his recordings) (4S4) seems to cause him to make just as much, if not more, extraneous, "whispery" breath noise in his glottis and at his lips as players using a "throat" vibrato often do.

In 1961, Clarence Kelly attempted to describe precisely how to produce "throat" vibrato. He says that to get beyond the "inadequate diaphragm vibrato, too slow for practical use, effective perhaps at times," (4S5) you should pronounce five times the syllable "Ah," strong but short, on one breath at mm 60 with a quarter rest in between each "Ah." Then you do the same thing again, but leave out the rests between the "Ah's," maintaining what he calls "the cutting edge" at the start of the syllable. Then you pronounce the connected "Ah's" in a whispered sound, still keeping the "cutting edge." Eventually you transfer this action over to the flute, increasing the number of pulsations per beat to three and then to four, and eventually allowing as many to occur as happen spontaneously. But, of course, this "cutting edge" quality that comes from making the "Ah" sound actually involves the "glottal shock" tightening action in the "throat/voice box" area that I described earlier, and it's not at all conducive to achieving the subtler expressive range that CP and others clearly hold out to us as ideal.

More recently, in 1996, Nancy Toff described in The Flute Book how to develop both "diaphragm" and "throat" vibrato:

For diaphragm vibrato, the first step is diaphragm pushes. Attack each note of a quarter- or eighth-note exercise with forceful contractions of the abdominal muscles. Only later add the tongue. In other words, "tongue with the breath." Next, play a long tone, slowly punctuating or pulsing it with diaphragm pushes. Using a metronome to monitor your progress, gradually speed up the tempo until the sound appears continuous. At this point the frequency of the pulses will be greater, but their intensity will diminish. (p. 109)

For "throat" vibrato, Toff pretty much reflects Kelly's 1961 procedure when she says:

. . . the first exercise does not use the flute. Say who or cough lightly at four counts per second. Then do the same with the instrument, playing a single note and gradually increasing the frequency of the whos or coughs. (p. 109)

But I think it's significant for our overview here to point out that Toff seems to consider pronouncing the word "who" to be interchangeable with the action of making a "light cough." To me, these are two quite different muscular operations—the light cough clearly involving an extra rib cage/abdomen/diaphragm action that saying "who" does not—and therefore produce distinctly different effects on the air column. This contrast will be more obvious if you try the exercises in Section V, Figure 1 which use a broader range of word-syllable formations as vibrato sources that are all quite different from each other.

Along with my observation that "throat" vibrato cuts so sharply and deeply into your air column, I find that "diaphragm" (rib cage/abdomen) vibrato usually interferes too much with your basic source of air flow. The main operation of breath support should be for tone production and control of dynamics, which, apart from the various degrees of accents and some forms of staccato, require a fairly smooth and even flow of air. But if you are also making extra little vibrato contractions and releases in the same musculature that you're attempting to maintain a steady breath flow with (which is essentially what you have to do when you make the "pushes" with your abdominal and torso muscles required in Toff's and Willoughby's approach to "diaphragm" vibrato) you really impose an additional burden on the basic breath support action coming from that region. It's as if you're creating a kind of tug of war in these areas; and, obviously, it would be much better to produce vibrato in the way that is the least burden upon or disruption of the basic function of your air flow.

Both "throat" and "diaphragm" vibratos, then, can easily create such excessive bursts of breath that they demand extra tension in your lips to focus your sound and maintain pitch—both of which have been considered paramount, dating back at least to Taffanel. (4S6)

A major problem in examining vibrato production stems from the fact that the two terms "throat" and "diaphragm" really don't accurately identify, from an exact anatomical standpoint, the locations and actions of what's taking place in us when we are producing either one of these types as they are commonly described and taught. First, the word "throat" can include quite an extensive area that involves many muscles and movable cartilages all the way from the base of the back of the tongue and the pharynx (the hooded wall behind the base of the tongue), the hyoid bone, the cricoid cartilage, the epiglottis/glottis/larynx ("voice box/vocal cords") and the thyroid cartilage ("Adam's apple") that surrounds them, and on into the trachea ("wind pipe") still farther down. What mainly happens with "throat" vibrato on flute is an opening and closing right around the larynx/glottis (vocal cords/voice box). But when you take a closer look, you see that this action involves only a small part of the larger region that the word "throat" can actually encompass, and it's only one of several possible constellations of contractions and expansions all along the upper vocal/respiratory tract that can be brought under conscious control to affect the air column in a number of quite different-sounding ways.

Then, the term "diaphragm" vibrato is also pretty much of a misnomer because our diaphragm is mainly an involuntary muscle, and we actually only have an indirect control over it by way of either lengthening or shortening our abdominal and other torso muscles that move the rib cage and the viscera in, out, up, and down around it. When our torso lengthens and widens, our diaphragm descends to create more space for the lungs to expand and let in air; and when our torso shortens and narrows our diaphragm ascends to assist the air flow out of our lungs, but we don't move our diaphragm by direct volition separately or independently from these general torso dynamics.

Of course, an equalized, overall torso control is what the Alexander Technique cultivates as a means for achieving the best action of our breathing in whatever we do. Alexander saw that the diaphragm and lungs function best if we "get out of their way" and let breathing happen as it's designed to, instead of vigorously trying to control the intake and outflow of air by specific, overt muscular action. He found that our in-breath will tend to happen more by itself and be much freer and fuller if we manage to keep as much elastic lengthening and widening as possible going on in our whole torso (especially across our back) while we propel air outwards to speak, to sing, or to play a phrase on a wind instrument—the very opposite of how many actors, singers, and wind players are taught to do so-called "diaphragmatic breathing" by "taking a (big) breath" and then focusing on contracting inwards (usually by squeezing with their abdominal muscles) to support and propel their sound. (4N6)

No doubt much more could be said about the precise anatomical aspects of vibrato production, (4N7) but I hope I've shown good reason here for being open to the possibility of using types of vibrato other than the common "throat" and "diaphragm" varieties. The next two sections describe how to do just that by drawing on the steps I took in my own search to re-create CP's vibrato—the "Petkoff vibrato." I hope, by now, that I've sparked your curiosity to read further and to experiment with these varieties yourself.


4N1. None of the many other flutists I listened to produce vibrato the way CP did. Nor do most of them vary theirs expressively as he suggested:

On CP's recital recordings, which range mainly from the early to late 1950s, he varies his vibrato even more in the later performances than in the earlier ones, which seems to show that he did work consciously on developing his expressive control of it. If you listen closely, I think you can easily hear that, in general, he isn't using a "throat" or "diaphragm" vibrato, even though at times he might come close to using a "throat" one. Yet, because his basic vibrato production is happening higher up, it never cuts sharply into the air column at the glottis as most "throat" vibratos do.

More recently I've heard recordings of Robert Willoughby (National Flute Association Historic Recordings Series, Volume 3, 2004) and Peter-Lukas Graf ("Flute Classics," Claves Records, 1994), both of whom do vary their "diaphragm" vibratos somewhat; however, theirs seem to be even further removed from the intrinsic kind of vibrato that CP produced than, say, the "throat" vibratos of Kincaid and Baker.


4N2. With some difference from flutist to flutist in number of pulsations per beat:

Robert Philip examines this aspect of speed of pulsation from player to player in his survey of flute vibratos in Early Recordings and Musical Styles: Changing Tastes in Musical Performance, 1900-1950, Chapter 5, Cambridge University Press, 1992.


4N3. Georges Laurent, in his recording of Hanson's Serenadefor flute, harp, and strings with the Boston Symphony:

It's interesting that Laurent, though his vibrato is quite regular throughout the piece, uses a straight tone on the final, held note. This seems to illustrate the "all or none" characteristic of most flutists who use the "throat/voice box" (glottal/laryngeal) vibrato. So it's also puzzling to read that Laurent denied using vibrato (Thompson, p. 155. See Note 2 below.).


4N4. Extra breath noise in the larynx and at the lips with "throat" and "diaphragm" vibratos:

I think many flutists feel that this breath noise doesn't matter much because they believe it's not audible at a distance; but I certainly find it very noticeable in many players' sounds, both in live performances and in recordings of solo and orchestral works. Sometimes all you can hear of this type of flute sound, both in woodwind section playing and full orchestra passages is the extraneous breath. It lets you know that there's flute playing going on in the ensemble somewhere, but it's not really the "body" of the flute's tone that you're hearing. Of course, some players' lip shape and dental formation may stand in the way of focusing their sound without extra breath noise; but usually there is still a good chance that they can focus better if they leave off other intrusive aspects like those that are likely to emanate from throat and diaphragm vibrato production.


4N5. Kincaid's vibrato less obtrusive, and varying more in intensity than pitch:

Even though I stated earlier that CP's sound was nothing like Kincaid's, there were probably more similarities between the playing of the two men, including their vibratos, than there were differences—especially when you compare them to most others of their and later generations (with the exception of Laurent, Baker, and Pappoutsakis, as I mentioned earlier.)


4N6. The Alexander Technique deals very much with developing the best practical understanding of the functioning of our breathing:

Alexander wrote several very elucidating chapters on "respiratory re-education" and "respiratory mechanisms" in his first two books, Man's Supreme Inheritance (1910) and Constructive Conscious Control of the Individual (1923). These chapters amplified earlier articles he'd written on the subject: "Introduction to a New Method of Respiratory Vocal Re-education" (1906) and "Respiratory Re-education" (1907). Alexander's books are available through Mouritz Ltd., London:

Also see my article, "Working on Breathing and Vocal Production," Published by STAT Books, London, 1993. Available through AmSAT Books at


4N7. Vibrato and anatomy:

An example of close anatomical analysis of vibrato is the work of oboist Andrew Brown of Central Missouri State University, who describes the typical kind of "throat/voice box" (glottis/larynx) vibrato and shows its operation in great detail in his article "A Cinefluorographic Pilot Study of the Throat while Vibrato Tones are Played on Flute and Oboe," which appears at I would be very interested in examining the Petkoff vibrato with a cinefluorograph. I feel certain that it would show different results from the "throat" vibratos Brown observed, particularly because of the different source of vibrato production I illustrate here.


4S1. John Krell, Kincaidiana, Trio Associates, Culver City, California, 1973, pp. 14-15


4S2. Shannon Thompson, A History and Analysis of the Philadelphia School of Clarinet Playing. University of Texas, Austin, 1998, p. 122. (Laurent's claim to use no vibrato is attributed to James Collis, "From Where I Sit," Symphony, 2/9, September 1949, p. 4.)


4S3. "The Art of Vibrato," Flute Forum, W.T. Armstrong Co., Autumn, 1960, p.7.


4S4. Robert Willoughby, Historic Recording Series, Volume Three, The National Flute Association, NFA-3. 2004.


4S5. "Teach Flute Vibrato Early," Flute Forum, Spring 1961, p. 14.


4S6. Taffanel: Genius of the Flute, Edward Blakeman, Oxford University Press, New York, 2005, p. 210.


A WORD OF CAUTION: If you are reluctant to tamper with your vibrato because you don't want to risk causing irreversible changes, you should probably not attempt any of the following experiments. However, if you do decide to try them out, I hope that my descriptions will shed useful light on this illusive subject for you—especially if you or your students are concerned with changing rom a continuous or too fast vibrato to a more expressively variable one.


I recommend beginning your approach to developing an expressive vibrato by first examining a wide range of different ways you can make pulsations in sound by consciously controlling, alternately, your lips, jaw, tongue, glottis/larynx ("throat/voice box"), or your rib cage/abdomen ("diaphragm"). You should be able to do this exploration fairly easily if you think it out carefully enough and if you always let each type emerge from a straight tone. Of course, you may need to develop more control over producing a straight tone first if you have a deeply-ingrained continual vibrato.

SUGGESTION: Warming up on long tones without vibrato is helpful in developing this control. And an exercise I adapted from one I learned from Fernand Gillet is also particularly useful to do without any vibrato at all: you play individual half-steps, by tongueing a single note of nine beats (at mm. 60) at a piano dynamic; crescendo to fortissimo over the first three beats; then decrescendo all the way down to pianissimo over the next six beats; then (on the same, original breath) re-tongue the same pitch with a sforzando attack for another beat, and end by tongueing the next half-step up at piano, for one final beat. That's eleven beats all on one breath. And remember to take plenty of time for allowing your breath to return—even allow several breaths to come and go before starting the next sequence of half steps, beginning with the last note you played. (Pay careful attention to intonation too, of course. Sometimes I also try to extend the first long note to twelve beats, still doing the crescendo to fortissimo on the first three beats and taking even more time to descrescendo over the last nine beats to a delicate pianissimo.)

joe (257K)

Once you're confident at leaving out vibrato entirely, start with the first five of the following exercises without your flute by whispering each sound while also looking at its corresponding drawing and identifying the particular areas in yourself marked by the arrows pointing to its main source. It's absolutely essential that you not let any vocal sound come into these exercises though. Always remember that there's a big difference between pronouncing some of these sounds in a whispered tone and pronouncing them in a vocal tone, particularly TOO-HOO-HOO-HOOH [No. 5] In a whispered "hoo" the rib cage/abdomen/diaphragm can be used on this sound pretty much exclusively, whereas in a vocal "hoo," the glottis/larynx ("throat/voice box") of course must come into action too—that is: tighten—simply because the vocal cords have to vibrate to produce the sound.

Take as much time as you need to become confident in doing each of these types accurately in a whispered tone before you actually try them on your instrument, where I think you'll quickly hear and feel the differences from one to another. The lips, jaw, and tongue vibratos [Nos. 1, 2, and 3] may seem bizarre when you try them out because most of us would never dream of using them in actual playing; but I think these first three types are even more important to work out than the next two more common glottis/larynx and ribcage/abdomen vibratos [Nos. 4 and 5] in developing the awareness and control you need for gaining access to the more gently responsive and versatile area where I found CP created his vibrato—the back of the tongue and soft palate region of type 6.

Once you've identified each of the first five above types of undulation very consciously and can do them easily at will, both in a whispered sound and on your flute, then you should be ready to approach the sixth type, which holds the potential for a much greater range of expression than many flutists commonly have access to. But the most important thing in trying No. 6 and in progressing from here on is to be able to "leave out" or refrain from activating any of the areas you've used to produce the first five modes.

SUGGESTION: If you want to work on refiniing your awareness and control even further, you might try exploring the whole gamut of possible ways of affecting your air column. Use the same approach as in 1 through 6 above on all the whispered sounds you can make from the lowest to the most forward: "hoo," "oo," "goo," "ghoo," [French uvular] "roo," "koo," "yoo," "noo," "doo," "too," "loo," [American] "roo," "thoo," "moo," "boo," "poo," 'soo," "voo." Some of these, of course, fall into the category of articulation.

When you're really sure that you can "not do" the first five vibrato types in Figure 1 and are able to maintain a completely steady, straight tone, the next step is to locate a region considerably farther up from the voice box [4] that lies more towards the back of your tongue and your soft palate (back of the roof of your mouth [6]). This sixth type of undulation is tricky to do and needs to be approached with infinite subtlety and care, but you can activate the area pretty well away from the flute if you whisper the sound "ghoo" very gently. The "ghoo" sound is formed very near the region where the sound "koo" is made, but "koo" (or even "goo") produces a complete break (an actual articulation) in the air stream when the back of the tongue and soft palate meet to enunciate the "k", whereas the softer, whispered "ghoo"can allow the air stream to keep on flowing steadily through in an unforced tone, with only a light caress around its perimeter. (You might go ahead and try saying "koo" and "goo" several times—both in a whisper and in full voice—just to be extremely clear about what you don't want to do in that area at the back of the tongue and soft palate.) Unfortunately, I can't find any words in English that use this soft "ghoo" sound, (5N1) but it seems to me to be something like the one that babies often make before they can say a really solid "goo." For those of you who speak French, you'll find the "ghoo" sound is made not too far from where you pronounce the uvular "r"(5N2); but, again, the uvular "r" creates too much of a break in the air stream to be useful in working on subtler vibrato. You might also try refining the "ghoo" sound by slowly whispering "hoo-ghoo." Beginning with a single, gentle, very slow whispered "hoo" can sometimes help you arrive at the whispered "ghoo" sound even more gently than if you try to start right out with "ghoo." This is because the "hoo" sound is created a little farther back from the "ghoo" sound and with a gentler muscular action. But if you make short, repeated (vocal or whispered) "hoos," as Nancy Toff suggests in her "throat" vibrato approach, you, of course begin to use definite rib cage/abdominal pushes to produce the "hoos" more continuously—and then, as I pointed out in Section IV, this actually becomes "diaphragm" vibrato as in example 5 above.

Making the whispered "ghoo" sound still doesn't bring you all the way to the Petkoff vibrato though, but you're very close to it. In order to get closer still, I recommend working with the "ghoo" sound methodically and progressively by starting with a straight tone on, say, a low G and trying exercise A below, which has only one pulsation on a longish note of one breath. When you can do one pulsation accurately and confidently, then move on to two pulsations on one note (B); then three (C); four (D); six (E); and eight (F). By the time you get to eight, you're almost there. Next try exercise G, all on one breath, starting on a straight tone, then adding two pulsations, three, four (or eight), and going back to three, two, and finishing on a straight tone again. This will prepare you best for ultimately surrendering your entire vibrato—really, your entire sound—over to the expressive character of a note or phrase (Section VI).

Ghoo pulses

SUGGESTION: If you're not aware of where your soft palate is, you can find it easily by touching the tip of your tongue to the back of your upper front teeth. Then slowly move that contact of the tip of your tongue up onto your gum just above your teeth, then on up to your hard palate (which is what you probably call the roof of your mouth). As you continue slowly moving your tongue backward along the center of your hard palate you'll eventually sense when the firm surface turns into a much softer tissue, just before you get to your uvula—the small, tubular shape that you can see hanging down over the top of your throat when you look into your open mouth with a mirror. This softer tissue is your "soft palate," and it has an ability to raise and lower above the back of your tongue—particularly in conjunction with how you make certain vowel sounds resonate. (Various regional accents tend to raise or lower the soft palate, according either to how "open" or how "flat" they sound. The open sound "om," for example, used by some meditational practices, usually raises the soft palate quite high—as do the "nasal" vowel sounds "on" and "en" in French. Country-Western singing, on the other hand, tends to flatten, or depress, the soft palate quite a bit—especially on the consonant "r.)"


5N1. I can't find any words in English that use this soft "ghoo" sound:

In modern Greek the sound of the letter "gamma" comes very close to the softer "ghoo" sound I'm recommending here. The Collins Pocket Greek-English Dictionary says "To produce this sound ["gamma"] make a continuous g sound without withdrawing the tongue." Greek words like "ghouna" ["fur"] use this soft "ghoo" sound.


5N2. The French influence on American vibrato:

In The Flute Book, Nancy Toff says "Vibrato as we know it today—a more or less continuous pulsation or shimmer in the tone—originated in the late nineteenth century in Paris. Paul Taffanel and oboist Fernand Gillet were two of the instigators." But I think she may be mistaken in saying that it was Fernand Gillet (1882-1980) and not his uncle Georges Gillet (1854-1920), who frequently performed with Taffanel (1844-1908) and was professor of oboe at the Paris Conservatory at the same time, as made clear in Edward Blakeman's biography of Taffanel (2005). Georges Gillet was actually the teacher of Fernand Gillet (Georges' nephew), as well as Marcel Tabuteau. However, it does seem that Toff must be correct when she says that [Fernand] Gillet, Tabuteau, Barrère, and Laurent later brought vibrato to the U.S. when they came to play in American orchestras: Barrère in 1905 to the New York Symphony, Tabuteau in 1915 to the Philadelphia Orchestra, and Laurent in 1918 and Fernand Gillet in 1925 to the Boston Symphony.

Whatever the exact origins and lineage, I can't help but wonder if these Frenchmen were aided in their vibrato production by the greater control they would have had over their uvular, soft palate, and pharyngeal regions from speaking the French "r" and the nasal vowel sounds in words spelled with "on," "en," "an," and "in." And, in light of this, I also wonder about the way native English-speaking U.S. flutists might have tried to emulate these Frenchmen's vibratos, since, for the most part, while studying French for many years myself, I have rarely found an American able to pronounce well the uvular "r" and the nasal vowels.

Although I studied with Fernand Gillet for two years, as I mentioned before, we never once discussed vibrato, and I regret never having asked him his views on the subject. This was in the early 1970s before I learned how to produce the Petkoff vibrato, but Gillet never objected to the type I was using and how I was using it at that point. He was in his early nineties then and had recently stopped playing after breaking his arm in a fall; so he never demonstrated anything to me on the oboe and I never got to hear his own vibrato. But Tabuteau's use of vibrato on the CD Marcel Tabuteau's Lessons (Boston Records, BR1017CD, Duxbury, MA, 1996) is subtle and varied, and he often uses none at all on certain notes or parts of notes. However, there certainly is a big difference between the vibratos of Barrère and Laurent in the recordings of them I have heard. Barrère, as widely noted, had a very rapid "chevrotement"—nanny goat—vibrato (as did his student, John Wummer, for most of his career); but Laurent's vibrato was definitely much slower and subtler. Of course, general neck and throat tightness can surely have an effect on the quality of vibrato one uses; and often this general tightness is a result of tense speech habits and voice production, which, in turn, can be linked up with personality traits. There certainly can be both tight-throated and free-throated French speakers, just as there can be tight-throated and free-throated speakers of any other language.



The final step in starting to produce an expression-based vibrato in actual playing takes a great leap of faith, because you need to stop deliberately making the "ghoo" pulsations—stop counting them, and merely let them come and go by themselves in response to the expressive character of the music. Before trying to do this in a piece though, you might work on playing, all on one breath, a single, calm, still note, letting it gently begin to stir with this subtler vibrato, and bringing it more and more alive until it reaches a fiery excitement; then let it gradually subside back to that original, tranquil calm and finally taper it off into nothing.

Now you should be ready to bring this more versatile vibrato into a musical phrase, but your whole mode of operation from here on must be linked directly to your feeling-image of what the life of a note or phrase demands rather than to any direct muscular control ("ideo-kinetic," as opposed to "physio-kinetic," control). (6N1) So you're merely leaving this general region (the "ghoo" area) at the back of your tongue and your soft palate (back of the roof of your mouth) available to govern the pulsations of your air stream however much the expression of a note demands it—if at all —watching carefully not to grab up this area or stretch it into too tight an opening. This kind of vibrato, if done well, sounds like it's coming more from inside your sound rather than being imposed from the outside and threatening to turn your tone into what Kincaid described as "all frosting and no cake."

When you're finally producing vibrato this way it will still remain a bit of a mystery because you're no longer manipulating the undulations by direct conscious control as you were in the "ghoo" exercises, yet neither are you resorting to the stronger bursts of air that usually come with the typical glottis/larynx ("throat") and ribcage/abdomen ("diaphragm") vibratos—which, by the way, you should still have access to if you ever feel you need to use them. I think that you can hear in CP's recordings that he came close to drawing on the common "throat" vibrato at times, but I believe that because he controlled his vibrato in general from the soft plate and back of the tongue area this higher focus rendered any throat vibrato he might have used much less severe and less extraneous to his sound and the expressive character of what he was playing.

If you find that these exercises don't work so well for you, there may be other aspects of your playing standing in your way. For instance, if you use certain kinds of extra tension for tone production (as in the so-called "open throat" tone (6N2)—where your throat may, in reality, be less open than when it's merely left alone as part of a free neck because you are trying to stretch it or hold it in a way that seems to feel more open, as in yawning, for instance), this tightening may also be hardening the back of your tongue and soft palate area and barring them from the subtle responsiveness to vibrato that the expressive momentum of a note or phrase can evoke. If you do make extra throat or other tension to control your sound, I recommend trying all of the above exercises (A through G) with a softer and less focused tone, just to make sure you're leaving all the surrounding throat, tongue, and lip musculature as free as possible. Then you can work on bringing focus back into your sound later—maybe using lips more than soft palate and back of the tongue, because your lips can be used more sensitively to focus when they aren't having to strive so hard to combat the stronger thrusts of air that come from a glottis/larynx ("throat") or rib cage/abdomen ("diaphragm") vibrato. By working in this gentler way, you may even discover realms of tone quality and flexibility you never had access to before.

If you've succeeded in getting this far, congratulations! Now you should be ready to try the Petkoff vibrato in an actual phrase of music. I suggest starting with pieces that are slow and not too full of moving notes, like some of the famous arias and lyrical passages in Marcel Moyse's Tone Development Through Interpretation. Maybe even take just one phrase and go over it a number of times—pausing to reflect quietly after each—so that you can give yourself the best chance to uncover any deeper movement and expression that you might miss if you forge straight through the piece from beginning to end. Simply "be with" the phrase. Absorb it. Contemplate it, until you've really plumbed its depths of feeling (6N3). Don't try to make it go anywhere. Let it take you and your vibrato—when, and if, vibrato is truly needed.(6N4)

Good luck! I hope you find as much satisfaction in using the Petkoff vibrato as I do.


(Please allow for my senior citizen status and lack of recent regular practice!)

Click here to listen to me playing the Andante of Mozart's D major flute quartet with the Petkoff vibrato.

Click here to listen to me playing the Andante of Mozart's D major flute quartet with automatic, throat vibrato.

Click here to listen to me playing the Andante of Mozart's D major flute quartet with automatic, diaphragm vibrato.

Click here to listen to me playing Debussy's Syrinx with the Petkoff vibrato.

Click here to listen to me playing Syrinx with automatic, throat vibrato.

Click here to listen to me playing Syrinx with automatic, diaphragm vibrato.

Click here to listen to me playing Afternoon of a Faun with the Petkoff vibrato.

Click here to listen to me playing Afternoon of a Faun with the automatic, throat vibrato.

Click here to listen to me playing Afternoon of a Faun with automatic, diaphragm vibrato.

Click here to listen to me playing the "piu andante" solo from Brahms' First Symphony with the Petkoff vibrato.

Click here to listen to me playing the "piu andante" solo from Brahms' First Symphony with automatic, throat vibrato.

Click here to listen to me playing the "piu andante" solo from Brahms' First Symphony with automatic, diaphragm vibrato.


Special thanks for their help in reviewing and editing this article must go to Jeff Mitchell, Suan Guess-Hanson, Marianne Gedigian, and Cathy Payne.



6N1. Ideo-kinetic, as opposed to Physio-kinetic:

See New Pathways to Piano Technique: A Study of the Relations Between Mind and Body with Special Reference to Piano Playing, by Luigi Bonpensiere, Philosophical Library, New York, 1953. Bonpensiere devised exercises that anyone can do to experience the difference between an action happening with an image as its source and an action happening because of direct manipulation of our muscles. Even though piano was the instrument he applied his discoveries to, they can work on any instrument.


6N2. "Open" throat, or free neck?

Achieving and maintaining fullest neck freedom (and therefore "throat" freedom) is fundamental to the overall integration of the postural mechanisms that is the goal of the Alexander Technique. Frank Pierce Jones demonstrated the effect of the Technique on vocal production in a study he conducted at Tufts Institute for Psychological Research in 1972 entitled "Voice Production as a Function of Head Balance in Singers," which is reprinted in his book on the Alexander Technique called Freedom to Change, Mouritz, London, 1997.

It seems to me that what many flutists are doing to create and maintain what they call an "open throat" is actually a very strong tensing of their soft palate, along with some pressing down of their larynx/glottis region (Alexander called it "depressing the larynx"), as we often do in yawning very strongly. But I think this so-called "opening" essentially causes an actual narrowing of the air column—the opposite of what they believe is happening—even though it may serve to get a more edgy, projected sound.


6N3. Contemplate it, until you've really plumbed its depths of feeling:

See my article "Musical Vision" in this web site, for suggestions about getting more in touch with the character of a piece.


6N4. . . . when, and if, vibrato is truly needed:

I recently read through Koechlin's Sonatine Modale, Op. 155, for flute and clarinet with a professional clarinetist who, like most, used no vibrato. I felt no need whatsoever to use any either, particularly because of the character of the music, which doesn't really have great dramatic intensity in it, but is calme, doucement expressif, lumineux, assez large, expansif, etc., according to Koechlin's own markings. I asked the clarinetist afterward if she felt anything missing because I didn't use any vibrato, and she said, "Not at all."