Another Incremental Step in Reviving and Revising Coffin's Favorable

315-332_JOS_JanFeb08_Depts B:JOS JanFeb09 12/1/08 12:08 PM Page 329

VOICE RESEARCH AND TECHNOLOGY Ingo Titze, Associate Editor Another Incremental Step In Reviving And Revising Coffin’s Favorable Vowel Chart

Ingo Titze

NTHENOVEMBER/DECEMBER 2007 ISSUE of Journal of Singing, I be- gan to propose a new way of looking at pitch-vowel interaction in singing. The column was entitled “Resurrection from the Coffin” because, al- though the acoustic theory behind interaction has changed somewhat, theI basic goal of understanding why certain vowels are favored at certain pitches has not changed. Coffin was not misled in his quest for a practical way to sort out the pitch-vowel dependence. I don’t mean to be presumptuous in trying to improve on his work in one or two little essays; this is long-term and detailed work in progress. However, the topic is important enough to singing Ingo Titze pedagogy that even a little step is worth reporting as a yearly update. Over a decade ago, my colleague Brad Story enticed two professional singers (a baritone and a soprano) to subject themselves to magnetic resonance imaging (MRI) while singing three vowels (/i/, /å/, and /u/). They also pro- duced the same vowels in a speech mode. Dr. Story processed the imaging data to obtain equivalent cylindrical vocal tract shapes for the vowels. Figure 1 shows the results for the soprano. On the left are six outlines of the airways (trachea and upper vocal tract) stretched out into a straight tube. The singer was well versed in both classical and jazz music. The top two outlines are for the vowel /i/, with the singing shape above the speaking shape. Note that the mouth (far right of airway diagram) is more open for the singing /i/, approximating more of an /I/ shape. We all know that singers tend to do this, especially with rising pitch. We say it creates more space and produces more resonance. But what is the evidence that it helps acoustically? On the right panel of Figure 1 we see some of the evidence emerging. On

top is a keyboard for pitch reference. The keys C3 through C8 are labeled. On the bottom axis are additional numeric labels of the corresponding frequen-

cies for A3 (220 Hz), A4 (440 Hz), A5 (880 Hz), A6 (1760 Hz), and A7 (3520 Hz). The four vertical bars represent frequency ranges of four source harmon-

ics (F0, 2F0, 3F0, and 4F0 from left to right) when the singer produces sounds Journal of Singing, January/February 2009 in her primo passaggio range, D4 to F4. Finally, the horizontally shaded areas Volume 65, No. 3, pp. 329–331 Copyright © 2009 are regions where the vocal tract offers favorable reinforcement to vocal fold National Association of Teachers of Singing vibration. When the shaded areas are above the reference line, we have supra-

January/February 2009 329 315-332_JOS_JanFeb08_Depts B:JOS JanFeb09 12/1/08 12:08 PM Page 330

Ingo Titze

Figure 1. Airways (left), and inertance curves (right) for a Figure 2.

soprano with pitch in the D4 to F4 range.

glottal inertance, which is beneficial. When the shaded In the primo passaggio region shown here, it is all

area is below the reference line, we have subglottal com- about a strong or weak second harmonic, 2F0. Note that

pliance, which is also beneficial. When the shaded areas 2F0 is reinforced by supraglottal inertance for the singing collapse to a straight line, neither the supraglottal nor /i/, but not for the speaking /i/. The same thing happens the subglottal airways are favorable to vocal fold vibration. for the vowel /u/, as can be seen from the bottom two cases in Figure 1. By widening the mouth to migrate from a speech /u/ to more of an /U/, the soprano was

able to keep 2F0 in the favorable inertive region. For the /å/ vowel in the middle of the two panels, the mouth was open enough in both cases (speaking and singing) so that the second harmonic was well within the favorable inertive region. For readers familiar with descriptions of pitch-vowel interaction in terms of formant frequencies, the mes- sage is that “favorable” translates to “harmonics being below a formant” in frequency. The formant frequencies are where the shaded area suddenly collapses from a relatively high value to zero. Note that the first for-

mant frequency (F1) is raised from F4 to E5 when the speaking /i/ is changed to the singing /i/. For the /å/ # vowel, it stays near F 5, and for the /u/ vowel it increases

from G4 to F5. Figure 2 shows the same vowel configurations and corresponding inertances, but now for a pitch range an

octave higher, D5 to F5. This is well above the primo passaggio and therefore the second harmonic has been

“lifted” over the first formant. The fundamental F0 is

now reinforced in the shaded area below F1 and 2F0 is

reinforced to a lesser degree in the F2 region. The third

330 Journal of Singing 315-332_JOS_JanFeb08_Depts B:JOS JanFeb09 12/1/08 12:08 PM Page 331

Voice Research and Technology

and fourth harmonics, 3F0 and 4F0, also get some reinforcement in the higher formant regions, but in general the differences between singing and speaking inertance values are not great for this singer at these pitches; the pri-

mary effect is how the fundamental (F0) is treated by the vocal tract. We hypothesize that the secondo passaggio

occurs when F0 is lifted over F1 with increasing pitch. Thus, it appears that the primo passaggio in females is

all about 2F0 interaction with F1, and the secondo pas-

saggio is all about F0 interaction with F1. In general, pitch-vowel interaction is complex. It in- volves many cases to be looked at. For each of about 10 vowels there are 2–3 formants to consider, and then there are 3–4 harmonics to keep track of for every pitch. We are not sure how all of this complexity can be organized in one simple chart. Coffin gave it an enormous effort; only time will tell if it can be condensed to the level that everyone can use it in the studio. A dynamic comput- erized version is probably the answer for the future, tai- lored to every individual singer.

Ingo R. Titze is Distinguished Professor of Speech Science and Voice at the University of Iowa and Executive Director of the National Center for Voice and Speech at the Denver Center for the Performing Arts. His formal education is in physics and electrical engineering, but he has devoted much of his studies to vocal music and speech. Dr. Titze has published more than 500 articles in scientific and educational journals, coedited two books titled Vocal Fold Physiology, and has authored two books called Principles of Voice Produc tion, and The Myoelastic Aerodynamic Theory of Phonotion. He has lectured throughout the world and has appeared on such educational television series as Innovation, Quantum, and Be yond 2000. He is a recipient of the William and Harriott Gould Award for laryn geal physiology, the Jacob Javits Neuroscience Investigation Award, the Claude Pepper Award, the Quintana Award, and the American Laryngological Association Award. He is a Fellow of the Acoustical Soci ety of America and the American Speech-Language- Hearing Association. Dr. Titze has served on a number of na tional advisory boards and scientific review groups, including the Scientific Advisory Board of the Voice Foundation and the Division of Research Grants of the National Institutes of Health. In addition to his scientific endeavors, Dr. Titze continues to be active as a singer. He is married to Kathy Titze and has four children. Mail should be addressed to Ingo R. Titze, National Center for Voice and Speech, 330 WJSHC, Iowa City, IA 52242. Telephone (319) 335-6600.

January/February 2009