THE FOE WE”: EE'JEVLP UF‘E’L‘ WEL [N arr: “r‘ inENCR\ EgEQE: u.- -U bl Dissertaticn Ear Etta Deg EL=L “EP ED. IEEE [CHEM E'E. STATE‘L' "“EE \"SRSE’E P LELE’E‘ELLEEEEEE SCHULTZ 1974 This is to certify that the thesis entitled THE FORMANT BEHAVIOR OF VOWELS IN SELECTED TENOR VOICES presented by Paul Wi11iam Schultz has been accepted towards fulfillment of the requirements for Ph.D. Music Jegree in MKW Major professor Date November 15, 1974 0-7639 E ABSTRACT THE FORMANT BEHAVIOR OF VONELS IN SELECTED TENOR VOICES By Paul william Schultz The Problem The problem of the study was to discover whether there are similar patterns of formant behavior in the spectra of sung vowels judged to be excellent in quality in tenor voices. The study attempted to discover the formants present in a given vowel and any similarity in patterns present, and to compare the results with those of previous research. The study included the vowels [a], [e], [i], [o], and [u] on the pitches e-l65 Hz, b-246 Hz, and g#-4l5 Hz. The voices studied were high school, university, and mature tenor voices selected by qualified choral directors and applied voice teachers at various high schools, universities, and cities in Michigan. The Procedure Thirty singers (l0 high school tenors, 10 university tenors, and 10 mature tenors) sang each vowel on each of the three pitches. Recording techniques were the same for each singer. The recording of each subject took place in a voice studio. The sample used for spec- trographic analysis was selected by 20 adjudicators, all having much vocal teaching experience. Paul William Schultz Samples of voices to be analyzed were prepared in random order on separate tapes for each of the three pitches and for each of the age groups--a total of nine tapes. The adjudication took place on three consecutive days for each adjudicator. Only those vowel samples receiving a simple majority of excellent ratings, with no fair or poor ratings, were used for spec- trographic analysis. The number of samples selected as excellent within a given age level varied according to vowel sound and pitch. Selected samples were prepared into tape loops and converted into spectrographs at the Audiology and Speech Sciences Laboratory at Michigan State University. The spectrographs were analyzed and reproduced graphically, indicating the frequencies, intensities, and locations of the fundamental and first two formants for each subject selected. Conclusions Adjudicator evaluations were found to be highly consistent. Using Hoyt's reliability method, results showed coefficients ranging from .9698 to .9819 for all vowels analyzed. Formants occurred on partial frequencies for all subjects. This was true, even though formants occasionally occurred on different partials for a given vowel or pitch. The first formant was consistent in location for all subjects on all vowels and pitches. Not one variation occurred in the study. The second formant was not as consistent in location, espe- cially on the lower pitches. The location of the second formant differed for two subjects singing the vowel [a] at e-l65 Hz. The Paul William Schultz second formant occurred on four different frequencies for the vowel [e] at e-165 Hz. The location of the second formant differed for two subjects singing the vowel [i], three subjects singing the vowel [o], and one subject singing the vowel [u] at e-l65 Hz. For pitch b-246 Hz, the location of the second formant differed for one subject singing the vowel [a], two subjects singing the vowel [o], and two subjects singing the vowel [u]. There was total consistency in loca- tion of the second formant for the vowels [e] and [i] at b-246 Hz. For pitch g#-415 Hz, the location of the second formant differed for two subjects singing the vowel [a] and one subject singing the vowel [i]. All subjects were consistent in the location of the second formant for the vowels [e], [o], and [u] at g#-415 Hz. Only 17, out of 120 vowels analyzed, differed in formant location. There was little indication that the age of the singer cre- ated basic differences in formant behavior. Formant intensities were relatively similar for all subjects. There was a noticeable variation in the intensities of the fundamentals, especially between high school and mature subjects. A prominent region of energy above the second formant was present in most spectrograms. Although this energy region has been referred to as the third formant in previous research, it was not consistent among subjects and failed to be as prominent as either the first or second formants. THE FORMANT BEHAVIOR OF VOWELS IN SELECTED TENOR VOICES ' By Paul William Schultz A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Music 1974 © Copyright by PAUL WILLIAM SCHULTZ 1974 ACKNOWLEDGMENTS I wish to express my sincere appreciation to the members of my committee, Dr. Gomer L. Jones, Dr. J. Loren Jones, Mr. Richard E. Klausli, and Dr. Robert G. Sidnell of Michigan State University, for their time and unerring guidance in this research. I will be eter- nally grateful to Dr. Dale Bartlett, committee chairman, for his counsel, meticulous approach to research, and warm friendship in bringing this investigation to its conclusion. Special thanks is also expressed to Mr. Don Riggs, Electronic Technician, Department of Audiology and Speech Sciences, Michigan State University, for his assistance in all technical matters. Finally, I offer my special thanks and love to my wife, Donna. Without her continuous encouragement, confidence, and love, this study would have been abandoned on several occasions. ii TABLE OF CONTENTS LIST OF TABLES . LIST OF FIGURES LIST OF APPENDICES Chapter I. INTRODUCTION . Purpose of the Study . Statement of the Problem Methodology . Limitations of the Study . Background of the Problem . Definition of Terms . 11. REVIEW OF RELATED LITERATURE Formant Theories . Comparison of the Relative .and Fixed Formant. Theories . . Directly Related Research . Summary of Related Research III. DESIGN AND PROCEDURES . Selection of Subjects Equipment for Recording Recording of the Singers . . . Preparation of the Adjudicator' 5 Tape . Selection of Adjudicators . . Presentation of Tapes to Adjudicators . Samples Selected for Spectrographic Analysis. Spectography IV. PRESENTATION OF DATA Adjudicator Reliability Spectrographic Analysis . Vowel [a] at Pitch e- -l65 Hz . . . A Comparison of All Subjects Singing the Vowel [a] at e- -l65 Hz . . Page vi vii XV Chapter V. Vowel [e] at Pitch e-l65 Hz A Comparison of All Subjects Vowel [e] at e-165 Hz . Vowel [i] at Pitch e-165 Hz A Comparison of All Subjects Vowel 5i] at e-165 Hz . Vowel [0 at Pitch e-165 Hz A Comparison of All Subjects Vowel [0] at e-165 Hz . Vowel [u] at Pitch e-l65 Hz A Comparison of All Subjects Vowel [u] at e-165 Hz . Vowel [a] at Pitch b-246 Hz A Comparison of All Subjects Vowel [a] at b-246 Hz . Vowel [e] at Pitch b-246 Hz A Comparison of All Subjects Vowel [e] at b-246 Hz . Vowel [i] at Pitch b-246 Hz A Comparison of All Subjects Vowel [i] at b-246 Hz . Vowel [0] at Pitch b-246 Hz A Comparison of All Subjects Vowel [0] at b-246 Hz . Vowel [u] at Pitch b-246 Hz A Comparison of All Subjects Vowel [u] at b-246 Hz . Vowel [a] at Pitch g#-415 Hz A Comparison of All Subjects Vowel [a] at g#-4l5 Hz . Vowel [e] at Pitch g#-415 Hz A Comparison of All Subjects Vowel [e] at g#-4l Hz . Vowel [i] at Pitch g -415 Hz A Comparison of All Subjects Vowel [i] at g#-4lS Hz . Vowel [0] at Pitch g#-415 Hz A Comparison of All Subjects Vowel [0] at g#-4l Hz . Vowel [u] at Pitch g -4l5 Hz A Comparison of All Subjects Vowel [u] at g#-415 Hz Summary . Conclusions iv singing Singing Singing Singing Singing Singing Singing Singing Singing singing Singing Singing Singing Singing the . the . the ° the . the . the O the ° the . the . the ' the . the . the . the ° SUMMARY, CONCLUSIONS, AND RECOMMENDATIONS Page 67 74 79 87 92 lOZ lOZ lll lll 120 l21 l27 13l l4l l4l 148 153 l59 l64 l73 l73 l82 182 l91 l9l 200 200 208 209 209 211 Chapter Page Results Related to Previous Research . . . . . . 2T5 Recommendations for Future Research . . . . . . 2l6 APPENDICES . . . . . . . . . . . . . . . . . . 2T8 BIBLIOGRAPHY . . . . . . . . . . . . . . . . . 272 LIST OF TABLES Table Page 1. Results of the voice experiment to determine recording environment and adjudicator environment . . . . . . 30 2. Subjects selected for Spectrographic analysis on pitch e-lGS Hertz . . . . . . . . . . . . . 50 3. Subjects selected for Spectrographic analysis on pitch b- 246 Hertz . . . . . . . . . . . . . 50 4. Subjects selected for Spectrographic analysis on pitch g#-415 Hertz . . . . . . . . . . . . . 51 5. 'Scores for reliability and standard error of measure- ment for each vowel at each of the three pitches . . . 52 vi Figure 10. ll. l2. l3. l4. LIST OF FIGURES Spectrograph of subject 1 singing b flat-234 Hz in a sound-treated room environment . . Spectrograph of subject l singing b flat-234 Hz in a studio environment . Spectrograph of subject 2 singing b flat-234 Hz in a sound-treated room environment . . Spectrograph of subject 2 singing b flat- 234 Hz in a studio environment. Spectrograph of subject 1 singing g- -396 Hz in a sound- treated room environment . Spectrograph of subject l singing g- -396 Hz in a studio environment Spectrograph of subject 2 singing g- -396 Hz in a sound— treated room environment . Spectrograph of subject 2 singing g- -396 Hz in a studio environment Spectrograph of high school subject 3 singing the vowel [a] at e-165 Hz . Spectrograph of high school subject 4 singing the vowel [a] at e- J65 Hz . Spectrograph of high school subject l0 singing the vowel [a] at e- J65 Hz Spectrograph of university subject 1 singing the vowel [a] at e- -l65 Hz . . . Spectrograph of university subject 4 singing the vowel [a] at e-165 Hz . . . Spectrograph of university subject 9 singing the vowel [a] at e-165 Hz . . vii Page 33 34 35 36 37 38 39 4O 54 55 56 58 59 60 Figure 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. Spectrograph of mature subject 1 singing the vowel [a] at e- -l65 Hz . Spectrograph of mature subject 3 singing the vowel [a] at e-l65 Hz . Spectrograph of mature subject 4 singing the vowel [a] at e- -l65 Hz . Spectrograph of mature subject 7 singing the vowel [a] at e-165 Hz . . Spectrograph of mature subject lO singing the vowel [a] at e- J65 Hz . Spectrograph of high school subject 3 singing the vowel [e] at e- -l65 Hz . Spectrograph of high school subject 4 singing the vowel [e] at e- J65 Hz . Spectrograph of high school subject l0 singing the vowel [e] at e- -l65 Hz . Spectrograph of university subject l singing the vowel [e] at e-l65 Hz . . Spectrograph of university subject 4 singing the vowel [e] at e-165 Hz . . . Spectrograph of mature subject 1 singing the vowel [e] at e- -165 Hz . . Spectrograph of mature subject 3 singing the vowel [e] at e-l65 Hz . . Spectrograph of mature subject 4 singing the vowel [e] at e-l65 Hz . Spectrograph of mature subject l0 singing the vowel [e] at e- J65 Hz . . Spectrograph of high school subject 3 singing the vowel [i] at e- -l65 Hz . Spectrograph of high school subject 4 singing the vowel [i] at e- J65 Hz . Spectrograph of high school subject 10 singing the vowel [i] at e- J65 Hz . viii Page 62 63 64 65 66 68 69 7O 72 73 75 76 77 78 80 81 82 Figure 32. 33. 34. 35. 36. 37. 38.. 39. 4o. 41. 42. 43. 44. 45. 46. 47. 48. Spectrograph of university subject l singing the vowel [i] at e- -l65 Hz . . Spectrograph of university subject 3 singing the vowel [i] at e-l65 Hz . . Spectrograph of university subject 4 singing the vowel [i] at e-l65 Hz . . . Spectrograph of mature subject 1 singing the vowel [i] at e-l6S Hz . Spectrograph of mature subject 3 singing the vowel [i] at e-l65 Hz . Spectrograph of mature subject 4 singing the vowel [i] at e-l65 Hz . . Spectrograph of mature subject 10 singing the vowel [i] at e-l65 Hz . Spectrograph of high school subject 3 singing the vowel [0] at e 165 Hz . Spectrograph of high school subject 4 singing the vowel [0] at e- J65 Hz Spectrograph of high school subject 10 singing the vowel [0] at e- J65 Hz . Spectrograph of university subject 1 singing the vowel [0] at e-l65 Hz . . Spectrograph of university subject 4 singing the vowel [0] at e-l65 Hz . . Spectrograph of university subject 3 singing the vowel [0] at e-165 Hz . . Spectrograph of mature subject 4 singing the vowel [0] at e-l65 Hz . Spectrograph of mature subject 10 singing the vowel [0] at e-l65 Hz . Spectrograph of high school subject 3 singing the vowel [u] at e-l65 Hz . Spectrograph of high school subject 10 singing the vowel [u] at e- J65 Hz . ix Page 84 85 86 88 89 9O 91 93 94 95 97 98 99 100 101 103 104 Figure 49. 50. 51. 52. 53. 54. 55.. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. Spectrograph of university subject 1 singing the vowel [u] at e-l65 Hz . . . Spectrograph of university subject 3 singing the vowel [u] at e-l65 Hz . . Spectrograph of mature subject 3 singing the vowel [u] at e-l65 Hz . Spectrograph of mature subject 4 singing the vowel [u] at e-l65 Hz . Spectrograph of mature subject 10 singing the vowel [u] at e- J65 Hz . Spectrograph of high school subject 5 singing the vowel [a] at b- 246 Hz . . Spectrograph of high school subject To singing the vowel [a] at b- 246 Hz . Spectrograph of university subject 3 singing the vowel [a] at b-246 Hz . . Spectrograph of university subject 9 singing the vowel [a] at b-246 Hz . . Spectrograph of mature subject 3 singing the vowel [a] at b-246 Hz . Spectrograph of mature subject 4 singing the vowel [a] at b-246 Hz . . Spectrograph of mature subject 10 singing the vowel [a] at b-246 Hz . Spectrograph of high school subject 4 singing the vowel [e] at b- 246 Hz . Spectrograph of high school subject 9 singing the vowel [e] at b- 246 Hz . Spectrograph of high school subject 10 singing the vowel [e] at b- 246 Hz . . Spectrograph of university subject 3 singing the vowel [e] at b-246 Hz . . Spectrograph of university subject 9 singing the vowel [e] at b-246 Hz . . X Page 106 107 108 109 110 112 113 115 116 117 118 119 122 123 124 125 126 Figure 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 76. 77. 78. 79. 80. 81. 82. Spectrograph of mature subject 3 singing the vowel [e] at b-246 Hz . Spectrograph of mature subject 9 singing the vowel [e] at b-246 Hz . . Spectrograph of mature subject 10 singing the vowel [e] at b-246 Hz . Spectrograph of high school subject 9 singing the vowel [i] at b- 246 Hz . Spectrograph of high school subject 10 singing the vowel [i] at b- 246 Hz Spectrograph of university subject 3 singing the vowel [i] at b-246 Hz . . . Spectrograph of university subject 9 singing the vowel [i] at b-246 Hz . . . . Spectrograph of mature subject 3 singing the vowel [i] at b-246 Hz . . Spectrograph of mature subject 4 singing the vowel [i] at b- 246 Hz . . Spectrograph of mature subject 9 singing the vowel [i] at b-246 Hz . . Spectrograph of mature subject l0 singing the vowel [i] at b-246 Hz . Spectrograph of high school subject 4 singing the vowel [0] at b- 246 Hz Spectrograph of high school subject 10 singing the vowel [0] at b- 246 Hz . Spectrograph of university subject 3 singing the vowel [0] at b-246 Hz . Spectrograph of university subject 4 singing the vowel [0] at b-246 Hz . . . Spectrograph of university subject 9 singing the vowel [0] at b-246 Hz . . . Spectrograph of mature subject 3 singing the vowel [0] at b-246 Hz . xi Page 128 129 130 132 133 134 135 137 138 139 140 142 143 145 146 147 149 Figure 83. 84. 85. 86. 87. 88. 89.. 90. 91. 92. 93. 94. 95. 96. 97. 98. 99. Spectrograph of mature subject 4 singing the vowel [0] at b- 246 Hz . Spectrograph of mature subject 9 singing the vowel [0] at b- 246 Hz . Spectrograph of mature subject 10 singing the vowel [0] at b- 246 Hz . Spectrograph of high school subject 5 singing the vowel [u] at b- 246 Hz Spectrograph of high school subject 9 singing the vowel [u] at b- 246 Hz . Spectrograph of high school subject l0 singing the vowel [u] at b- 246 Hz . . Spectrograph of university subject 4 singing the vowel [u] at b-246 Hz . . Spectrograph of university subject 9 singing the vowel [u] at b-246 Hz . . Spectrograph of mature subject 3 singing the vowel [u] at b-246 Hz . Spectrograph of mature subject 4 singing the vowel [u] at b-246 Hz . . Spectrograph of mature subject 9 singing the vowel [u] at b-246 Hz . . Spectrograph of mature subject 10 singing the vowel [u] at b-246 Hz . . Spectrograph of high school subject 9 singing the vowel [a] at g#-415 Hz . Spectrograph of high school subject 10 singing the vowel [a] at g#-415 Hz . . Spectrograph of univgrsity subject 3 singing the vowel [a] at g -415 Hz . . . . . Spectrograph of univgrsity subject 9 singing the vowel [a] at g 415 Hz . . Spectrograph of mature subject 4 singing the vowel [a] at g#-415 Hz . . xii Page 150 151 152 154 155 156 157 158 160 161 162 163 165 166 168 169 170 Figure 100. 101. 102. 103. 104. 105. 106.‘ 107. 108. 109. 110. 111. 112. 113. 114. 115. 116. Spectrograph of mature subject 9 singing the vowel [a] at g#-4l5 Hz . . Spectrograph of mature subject 10 singing the vowel [a] at g#-415 Hz . . Spectrograph of high school subject 3 singing the vowel [e] at g#-415 Hz . . Spectrograph of high school subject 10 singing the vowel [e] at g#-415 Hz . . Spectrograph of univgrsity subject 3 singing the vowel [e] at g ~415 Hz . . . . Spectrograph of university subject 9 singing the vowel [e] at g#-4l5 Hz . . . Spectrograph of mature subject 3 singing the vowel [e] at g#-415 Hz . . Spectrograph of mature subject 4 singing the vowel [e] at g#-4lS Hz . . Spectrograph of mature subject 9 singing the vowel [e] at g#-415 Hz . . Spectrograph of high school subject 9 singing the vowel [i] at g#-415 Hz . . . Spectrograph of high school subject 10 singing the vowel [i] at g#-415 Hz . . Spectrograph of university subject 1 singing the vowel [i] at g#-415 Hz . . . Spectrograph of univgrsity subject 3 singing the vowel [i] at g 415 Hz . . . Spectrograph of mature subject 3 singing the vowel [i] at g#-415 Hz . . . Spectrograph of mature subject 4 singing the vowel [i] at g#-415 Hz . . Spectrograph of mature subject 9 singing the vowel [i] at g#-415 Hz . . Spectrograph of high school subject 9 singing the vowel [0] at g#-415 Hz . . xiii Page 171 172 174 175 177 178 179 180 181 183 184 186 187 188 189 190 192 Figure 117. ll8. 119. 120. 121. 122. 123.. 124. 125. 126. 127. 128. Spectrograph of high school subject 10 singing the vowel [0] at g#-415 Hz . Spectrograph of univgrsity subject 3 singing the vowel [0] at g -415 Hz . . . Spectrograph of univgrsity subject 9 singing the vowel [0] at g -415 Hz . . . 'Spectrograph of mature subject 3 singing the vowel [0] at g#-415 Hz . Spectrograph of mature subject 9 singing the vowel [0] at g#-415 Hz . . Spectrograph of mature subject 10 singing the vowel [0] at g#-4l5 Hz . . Spectrograph of high school subject 9 singing the vowel [u] at g#-415 Hz . Spectrograph of high school subject 10 singing the vowel [u] at g#-415 Hz . Spectrograph of unixersity subject 3 singing the vowel [u] at g 415 Hz . . . Spectrograph (If university subject 9 singing the vowel [u] at g#-415 Hz . . . . Spectrograph of mature subject 4 singing the vowel [u] at g#-415 Hz . . Spectrograph of mature subject 10 singing the vowel [u] at g#—415 Hz . . xiv Page 193 195 196 197 198 199 201 202 204 205 206 207 LIST OF APPENDICES Appendix Page A. Voice Equipment . . . . . . . . . . . . . . 220 B Letter to Directors . . . . . . . . . . . . . 222 C. Adjudicator's Instructions . . . . . . . . . . . 224 D Adjudicator Master Score Charts . . . . . . . . . 226 XV CHAPTER I INTRODUCTION Purpose of the Study It is the purpose of this study to discover formant patterns present in high school, university, and mature.tenor voices that may lead to consistent judgments regarding quality in all voices. Statement of the Problem The problem of this study is as follows: Are there similar patterns of formant behavior in the spectra of sung vowels judged to be excellent in quality in tenor voices? In answering this question, this research will attempt to discover the formants present in a given vowel, any similarity in patterns present, and compare the results with those of previous research. Previous research relating to this study will be presented in Chapter II. Methodology The Sample The voices studied were high school, unversity, and mature tenor voices selected by qualified teachers of applied voice at vari- ous high schools, universities, and cities throughout Michigan. The sample used for Spectrographic analysis was determined from the total sample by twenty adjudicators selected from the Michigan School Vocal Association's list of solo adjudicators. Design and Methods Each tenor in the sample sang the vowels [a], [e], [i], [o], and [u] on three different pitches (e-165 Hz, b-246 Hz, and g#-415 Hz) representing lower, middle, and upper ranges of the tenor voice. Thus, each tenor was rated on fifteen different utterances. Each adjudicator rated the quality of each sound based on a scale of 1--Excellent, 2--Good, 3--Fair, and 4--Poor. This design is a modified version of previously designed procedures used by Jones1 and Wash.2 To implement the above design, the following conditions were necessary: l. Five vowels were included in the study in order to determine the effects of different formant patterns on auditor preference. 2. Three pitches were used to insure covering the entire range of the tenor voice. 3. The singers were chosen on the basis of a strong recom- mendation from a highly respected voice teacher or music educator. 4. The singers were grouped on tape at each maturity level (high school, university, and mature voices) to insure uniform comparisons. 5. The singers were recorded at the same volume level, on the same equipment, and the same distance from the microphone to 1d. Loren Jones, "A Cinefluorographic and Spectrographic Analy- sis of the Effect of Velum Positions on Sung Vowels" (D.M.Ed. disserta- tion, School of Music, Indiana University, 1971). - 2Nathaniel Hubert Nash, "The Formant Behavior of the Vowels [a], [i], and [u] in Baritone Voices in Relation to Different Voice Ranges" (Ph.D. dissertation, Michigan State University, 1971). insure against differences in timbre which might result from uneven signal strengths. 6. The adjudicators chosen for the study were engaged in studying, performing, or teaching voice or choral music at the high school, university, or professional level. In addition, they were designated as solo adjudicators on the Michigan School Vocal Associ- ation's adjudication list. 7. The material was presented to the adjudicators in a totally random manner so that each singer, pitch, and vowel could be evaluated individually. 8. Adjudicator reliability was determined by appropriate statistical computation. A In order to determine the environment in which the singers would be recorded and how the tapes would be presented to the adjudi- cators, the following experiment was undertaken before the study began:. 1. Two mature tenor voices were selected to sing two differ- ent pitches (b-246 Hz and g-396 Hz) on the vowel [a]. 2. Two different recording environments (a sound-treated room and a studio-type room) were selected for recording the singers. 3. The same recording equipment and same testing procedures were implemented in each environment. 4. Ten qualified adjudicators were selected to listen to the recordings and determine which environment produced the most natural sounding result. 5. The presentation of tapes to the adjudicators was designed so that the observations could be tested for reliability, individual differences, and determining the method of presenting tapes to adjudicators in the final research. Analysis of Data Each sung vowel chosen by a simple majority of the adjudicators as excellent in quality was used for Spectrographic analysis. In addi- tion, Hoyt's reliability coefficient was calculated to determine reliability among adjudicators. Limitations of the Study High school, university, and mature tenor voices were used as subjects in this study. The high school singers were limited to those between the ages of 16 and 18 having fully completed their voice change, and were members of either the Michigan School Vocal Association's Honors Choir or the Youth for Understanding Chorale, two of the most highly selective high school choral organizations in the state of Michigan. The university singers were limited to those between the ages of 19 and 24, and the mature tenor voices from 25 to 50. Five vowels were used in this study. Previous studies have used two or three vowels, but have usually recommended that future research include at least five vowels. The three pitches chosen for this research represent the lower, middle, and upper ranges of the tenor voice. These pitches caused some technical difficulty with some singers (i.e., some high school singers had difficulty with either the lower pitch or the higher pitch, and, in some cases with both extremes) depending upon the experience or the vocal and physical maturity of the individual. Three 20-minute tapes were prepared for each of the three classifications of singers. To avoid a possible fatigue factor in the evaluation of the tapes by some adjudicators, the adjudicators were asked to evaluate the tapes on three separate days, with short breaks between each tape. Only those vowel sounds rated excellent by a simple majority . of the adjudicators were used for Spectrographic analysis. Because of the purpose and magnitude of this study, those not selected as excellent were not considered for comparison. Consideration was given only to the sung vowels and not to the singer producing those vowels. Consequently, it was possible that only one or two of the vowels sung by an individual were used in the final Spectrographic analysis, and not necessarily all of the vowels sung by that individual. Background of the Problem The requirements of singing necessarily make it different from speech. Listeners normally understand speech,lmxtthis cannot be said of a listener's ability to recognize the same words when sung. The differences between the act of speech and the act of singing are created because of such variables as range, duration of pitch, tempo, and mood. This is supported by Fletcher after conducting and experi- ment comparing melodic curves of the same sentence when spoken and sung. In the case of the sung sentence the pitch changes are around definite intervals on the musical scale while for the spoken sentence the pitch varies irregularly, depending upon the emphasis given. The pitch of the fricative and stop con- sonants is ignored in the musical score, and since these consonants form no part of the music, they are generally slid over, making it difficult for a listener to understand the meaning of the words. .Some teachers of music object to this statement of the situation, but I think most people will agree that a singer's principal aim is to produce beautiful vowel quality and to manipulate the melodic stream so as to produce emotional effects. To do this, it is necessary in singing to lengthen the vowels and to shorten and give less emphasis to the stop and fricative consonants. It is for this reason ghat it is more difficult to understand song than speech., 4 observes that pronunciation in singing demands that Vennard the sound be enough like normal speech to create an illusion of naturalness. This means that while vowels should be kept as pure and musical as possible, they must be modified were necessary in favor of more familiar or recognizable pronunciation. Pronunciation of vowels and concepts of tone production are often the explanation of a person's ability to Speak clearly and sing unrecognizably. From the research examined in this study and discussed in more detail in Chapter II, claims have been made concerning the character- istic frequencies of the basic vowel formants most evident in, or important to, good singing. The results are not always in agreement as evidence in the following summary which Vennard entitles "The Acoustical Dilemma": If Helmholtz made any mistake, it was in assuming that the voice functions entirely like a musical instrument. He applied to vowels the same principles which explain the tone color of instruments. Those who dispute this assumption are 3Harvey Fletcher, Speech and Hearing in Communication (New York: Van Nostrand Company, Inc., 1953), pp. 54-56. 4William Vennard, Singing, the Mechanism and the Technic (5th ed. rev.; Ann Arbor: Edwards Brothers, Inc., 1964), pp. 166-167. primarily phoneticists. Scripture, for example, based his studies largely upon the spoken word. He analyzed graphs of the recitations of Joseph Jefferson, and found few vowels of the kind that a singer produces. Instead of diphthongs he found triphthongs, continually merging into consonants, no sound prolonged for enough time to establish musicality. Small wonder he found more inharmonic partials than har- monics . . . . Possibly the confusion could be resolved by saying that in singing, where we want the voice to be a musical instru- ment, we shape the resonators to be in tune with the funda- mental, and so our vowel formants are harmonic, whereas in speech, this need not be the case. As a matter of fact Paget analyzed the vowel [o] as sung throughout a chromatic scale and found that "of the fourteen upper resonances heard, ten were actual harmonics of the larynx, i.e., their frequency of vibration was a numerical multiple of that of the larynx note. In two of the remaining cases, the relation, though not har- monic, was musically a simple one, namely an interval of two octaves and a fourth; in the last remaining cases there was apparently no simple relation between the larynx note and the upper resonance hear " (pp. 48-52) . . . . Of course, to make the voice musical, we must also free the valve to provide more and better partials in the glottal tone. Scripture did not recognize this possibility, and phoneticians generally do not concern themselves with it, though of course speech therapists do. There is a critical need for more information resulting from carefully controlled observation and reporting of the several physi- ological and acoustical phenomena in relation to excellence in performance. Because of the broad scope of controversy in this area, this study will be but one step toward building the necessary amount of information needed to begin resolving such controversy. 5Ibid., p. 166, citing Hermann L. F. Helmholtz, Sensations of Tone (4th ed.; New York: Dover Publications, 1954), pp. 39-41; Edward Wheeler Scripture, "Analysis and Interpretation of Vowel Tracks," Journal of the Acoustical Society of America, V (1933), 148- 152; and Sir Richard A. Paget, Human Speech (New York: Harcourt, Brace, and World, 1930), pp. 48-52. Definition of Terms Decibel represents a relative quantity. It is a logarithm of a ratio of two values of power, and equal changes in decibels repre— sent equal ratios. The decibel is most often used in acoustics for expressing the sound-pressure level and the sound level: abbrevi- ated dB. Formant is a region of pitch in which all partials are strengthened. For every vowel there are at least two formants of fixed pitch. The relative strength of a formant is an important fac- tor in determining the quality of a sound. The abbreviation F is fbllowed by a numeral denoting the chronological order of those formants observed within the spectrum of a given sound (i.e., F1, F2, F3). Hertz_is synonymous with cycles per second and abbreviated Hz. International Phonetic Alphabet is abbreviated I.P.A. and designed to represent phonetic sounds and avoid makeshift spellings. The following are the five basic vowel symbols used in the present study: [a] as in calm [e] as in pay [i] as in beet [o] as in tone [u] as in boot Partial refers to any of the simple components of a complex tone, the frequencies of the upper components being exact multiples of the fundamental. The abbreviation P is followed by a numeral denoting the ascending position of the partial, beginning with P1 as the fundamental. Spectrograph is the graphic display of the frequency compo- sition of a sound. Vg_is the abbreviation for volume unit. VU Meter is the meter on a tape recorder indicating signal strength. CHAPTER II REVIEW OF RELATED LITERATURE The research to be reviewed in this chapter will be concerned with both spoken and sung vowels. The linguistic studies are valuable in providing information relevant to the early opinions on formant behavior, many of which apply to sung vowels. Conclusions from the linguistic studies reviewed may not relate directly to the sung vowel, but the overall content of these studies reveals that speech and sing- ing are greatly related. Formant Theories Such classifications as absolute formant theory, relative for- mant theory, and fixed formant theory are the results of linguistic studies. In these studies, attempts have been made to define the number of formants for each vowel and which formants are most important for vowel quality and vowel perception. Since vocal quality, in rela- tion to formant behavior, has been a controversial subject, this portion of Chapter II will present some varying opinions regarding the three formant theories mentioned previously. The Absolute Formant Theory This theory suggests that for a listener to recognize a vowel phoneme, he must depend upon the absolute values of the formant fre- quencies of that vowel. It is also suggested that these values remain 10 11 unchanged regardless of the sex of the performer or the acoustical environment in which the vowel is produced.6 Before the word formant became associated with vocal or vowel quality, the principle of harmonic reinforcement was under investiga- tion. In the nineteenth century Helmholtz studied the vowels [a], [o], and [u] for one position of the oral cavity, and included the vowels [e] and [i] for another position. By placing a series of vibrating tuning forks before the mouth after it was shaped for the phonation of a given vowel, Helmholtz determined the harmonics most strongly rein- forced. Helmholtz discovered one partial reinforced for the vowels [a], [o], and [u], and two for the vowels [e] and [i].7 Helholtz discovered similar formant frequencies for men, women, and children in the above experiment. He reports: The pitch of strongest resonance of the oral cavity depends solely upon the vowel for pronouncing which the mouth has been arranged, and alters considerably for even slight alterations in the vowel quality, such, for example, as occur in the different dialects of the same language. 0n the other hand, the proper tones of the cavity of the mouth are nearly independent of age and sex. I have in general found the same resonances in men, women, and children. The want of space in the oral cavity of women and children can be easily replaced by a great closure of its opening, which will make t e reso- nance as deep as in the larger oral cavities of men. 6William E. Castle, The Effect of Selective Narrow-Band Filtering on the Perception of Certain English Vowels (The Hague: Mouton and Company, 1964), p. 17. 7Helmholtz, op. cit., pp. 104-107. 81bid.. pp. 50-51. 12 Aikin9 performed experiments similar to those of Helmholtz and found that for a given vowel, women resonated a pitch approximately a minor third higher than a man, and children higher than women. This was due to the size of the resonators producing a comparable resonant note. Aikin offers a humble apology in differing with the findings of Helmholtz: I regret to be obliged to mention that the great physi- ologist Helmholtz believed the pitch of this resonant note to be the same "in men, women, and children." I can only attrib- ute this error to the use of artificial aids to hearing, and not trusting to the sensitiveness of the ear alone. I have invariably met with not only the differences between the sexes above described, but also the subtler differences between individuals of the same sex proportionate to the actual measurements of the cavities in their necks and mouths.10 Before Aikin's experiment it was already apparent that propo- nents of this theory were not sound in their logic. Erickson made the following observations concerning vocal quality or timbre: Peculiarities of timbre may then be due quite as much to the presence or absence of partials with certain pitches as to the relative intensities among the partials. This will account for differences in timbre between musical and non- musical voices; between the voices of children and those of adults; between the voices of men and those of women; although all may produce the same vowel sounds at the same pitch. Based on the previous observations, the absolute formant theory is not sound when applied to the singing voice or, for that 9William A. Aikin, The Voice (New York: Longmans, Green and Company, 1951), pp. 48-51. 101616.. pp. 50-51. 1lCarl I. Erickson, "The Basic Factors in the Human Voice," Psychological Monographs, XXXVI (February, 1927), 90—91, 13 matter, to the spoken voice, especially when observing performers of, different sex or maturity levels. The Relative Formant Theory- This theory, also referred to as the harmonic theory, suggests that vocal quality results from a fundamental frequency and the rela- tionship of the upper partials to the fundamental. The frequencies of the partials are all exact multiples of the fundamental. Several studies support this theory. 12 concluded that vocal quality is dependent upon two Holmes main factors: (1) the relationship between the fundamental and the harmonic frequencies of a musical sound; and (2) the proper function- ing of the vocal mechanism depending upon anatomical, physiological, and psychological factors. Holmes felt that, in addition to the above factors, correct vowel placement on a pitch suitable for a given voice is essential for good vocal quality. Holmes' second point may seem too obvious to be included in this study but, in reality, the present investigation is dealing with that point in great depth. Without the proper functioning of the vocal mechanism and correct vowel placement, consistent measurement of formant behavior would not be possible. Borchers studied the relationship between intensity and har- monic structure by having four well-known professional male singers 12F. Lincoln P. Holmes, "An Experimental Study of Individual Vocal Quality," Quarterly Journal of Speech, XVI (October, 1930), 351. 14 (two basses and two tenors) sing the vowel [a] at three intensity levels. The results were as follows: 1. There is without exception a greater percentage of energy in the fundamental of tones sung with less intensity but at the same approximate pitch. Inversely, there is less energy in the fundamental of tones sung at greater intensity levels but at the same approximate pitch. 2. There is an increasing amount of energy in the funda- mental from low to high tones which are sung pianissimo. 3. The second partial shows characteristic changes in energy with changes in intensity level which are very similar to the changes in the fundamental. 4. The centroid of the total dispersion of energy changes to markedly higher frequencies in tgnes sung at greater inten— sity levels but at the same pitch. Borchers' conclusions bear out the relative formant theory. However, these conclusions may not be the only important features in vowel recognition. 14 Castle tested the hypothesis that the acoustic information necessary for the recognition of context vowels is different from that for isolated vowels. This hypothesis suggests that formant bandwidths and the fundamental frequency may be as important to vowel recognition as the elements of the relative formant theory. Castle's results indicated: The sufficient acoustic information for the recognition of context vowels differs from that for isolated vowels. It appears that identification of context vowels may be less dependent on steady state formant information and more dependent on temporal variations and comparison vowel infor- mation than is identification of isolated vowels. In addi- tion, there is indication that there may be a perceptual heirarchy [sic] among the formants of a given vowel, that 13Orville J. Borchers, "The Relation Between Intensity and Harmonic Structure in Voice," Psychological Record, III (April, 1939), 64. 14Cast1e, op. cit., pp. 157-160. 15 ratio hypotheses of vowel perception are not completely valid, and that the most important frequencies for the iden- tification of the ygwels of continuous speech may be between 1500 and 2500 cps. The Fixed Formant Theory This theory, also known as the formant theory, is defined by Bartholomew: The formant theory states that the characteristic tone quality of an instrument is due to the relative strength- ening of whatever partial lies within a fixed or relatively fixed region of the Tgsical scale. This region is called a formant of the tone. ‘7 summarizing Concerning these fixed regions, Gray and Wise, several research studies, state that formants are characteristic fre— quency regions which determine the different speech sounds. The summary also reveals approximate formant frequencies for two vowels not used in the present investigation, and that the formants of men are lower than those of women and children. Delattre,18 in an article comparing vowel color and voice quality, made observations concerning fundamental and formant behavior. He stated that formant frequengy and, occasionally, formant intensity are the necessary factors for vowel identification. Regarding voice quality Delattre stated: 1516161., p. 160. 16Wilmer T. Bartholomew, Acoustics of Music (New York: Prentice- Hall, Inc., 1942), p. 17. 17Giles Wilkeson Gray and Claude Merton Wise, The Bases of Speech (3d ed.; New York: Harper and Brothers, 1959), pp. 117-131. 18Pierre Delattre, "Vowel Color and Voice Quality: An Acoustic Articulatory Comparison," National Association of Teachers of Sigging‘ Bulletin (October, 1958), 4-7. 16 The work of correlating voice formants with types and classes of voices has not yet been done successfully. It is generally agreed, however, that voice quality in singing is mainly characterized by the two or three vowel formapas whose frequencies are just above the vowel formants. Delattre identified those frequencies for a male between 2400 and 4000 Hz. He further stated that high frequency overtones are necessary for rich vocal quality in the singing voice. The studies dealing with the fixed formant theory also refer to the number of formants necessary for vowel distinction. Potter, 20 found the third and fourth formants were either very 21 Kopp, and Kopp weak or, in many cases, blended together. Delattre found that the first three formants are necessary for vowel identification of front- unrounded vowels, but the third formant is very close to the second. The first two formants, only, were necessary for identification of other vowels. Winckel stated that each vowel is formed from two formants or more, but those beyond the second formant are insignifi- cant. He continues: The characterization of sound essentially through two formants is of great significance, even though it is not organically adjusted to the hearing process; this is because we do not distinguish single resonance areas of the spectrum, as does the keyboard of the bulging basilar membrane in the inner ear, but rather hear an integrated blending. Howsyer, there is no other characterization of tone color known. 19 20Ralph K. Potter, George A. Kopp. and Harriet Green Kopp, Visible Speech (New York: Dover Publications, Inc., 1966), pp. 54-56, 66, 69. Ibid., p. 5. 2IDelattre, op. cit., p. 5. 22Fritz Winckel , Music, Sound and Sensation (New York: Dover Publications, Inc., 1967), p. 20. 17 Comparison of the Relative and Fixed Formant Theories Bartholomew illustrated the two theories on a chart showing the fundamental and strengthened partials of three separate pitches. He summarized the differences between the theories as follows: Thus, according to the harmonic theory, if the instru- ment plays a different pitch, the vibration form will be similar, since, although the particular harmonic series will be shifted, its members will still retain the same intensity relation to each other. But according to the formant theory, whatever partial lies within or close to the formant range will be strengthened. Ifthe pitch of the fundamental changes, so that some other partial comes into the formant range, this other partial will be strengthened. Thus, although the for- mant range'ksfixed, Ehe vibration form will change as the fundamental changes. 3 Vennard24 referred to the harmonic theory as the Relative Pitch Theory and to the formant theory as the Fixed Pitch or Formant ngpry, He suggests that the formant behavior of various instruments bears out both theories. The relative pitch theory explains why strong formants can be found in certain regions of most instruments. Vennard suggests this explains only 80 or 90 percent of the formant behavior of instruments. Using the flute to illustrate this sugges- tion, Vennard explains: The second partial is always strongest, and this bears out the Relative Pitch Theory; but you will also notice that in the C4 spectrum the first partial is weaker than the third, while as the pitch of the tone rises, the first partial grows in strength and the third diminishes. This can only be explained by the Formant Theory; that is, there could be dis- covered in the flute a formant whose pitch was fixed somewhere between the first and third partials, so that as the first moved up toward it, that partial was augmented, and as ghe third moved away from it, that partial was diminished.2 23Bartholomew, op. cit., p. 17. 24Vennard, 0p. cit., pp. 125-127. 251bid., p. 125. 18 A great amount of research regarding formant behavior has been done in the area of vocal acoustics, but such research is not readily applied to instrumental acoustics. It is evident that formant behavior in instruments depends not only on the range of the formant, but also on the intensity within that range, relative to the funda- mental, to determine timbre. Vennard also stated that formants may be detrimental to the vocal tone and that bronchotracheal formants created vocal problems, i.e., voice sputtering when changing registers. Vennard concluded: The supraglottal formants, however, are essential, and the story of vocal timbre reverses that of instrumental tim- bre. Here eighty or ninety per cent of the proportioning of the harmonics is controlggd by formant, and the rest is rela-‘ tive to the fundamental. Directly Related Research The remainder of this chapter will deal with four studies which relate directly to the present investigation. These studies combine to serve both as motivation for the present research, and to formulate most of the methods and research design used by this inves- tigator. Appelman is one of the leading authorities on the singing voice and formant behavior. He discussed the creation of vowel formants as follows: As the sound passes through the resonating cavities of the throat and mouth, the profile of the spectrum changes, since each cavity resonates to some of the tones in the spec- trum more readily than to others and each adds its own 261bid. the 19 characteristics to such tones. This reinforcement gives the partials greater energy at the point of cavity resonance. These points of greater energy are called formants. In passing through the resonating system of the throat and mouth, the partials in the harmonic sequence do not change from their original location in the tonal spectrum; rather, some are strengthened and reinforced by cavity reso- nance, while others are weakened or damped out. The values of the natural frequencies of the resonating cavities within the vocal tract are determined by their shape; as a result, as the shape of the tract is altered the amplitudes of the partials within the spectrum will be greater at different frequencies. Thus every configuration of the total vocal tract has its own set of characteristic formant frequencies which gives to the laryngeal sound a particular vowel quality. The resonance frequency of any cavity is not necessarily equal to the frequency of any partial of the spectrum. The frequencies of the formants need not be the same as those of the partials, but they may coincide. The formant frequen- cies are determined by the configuration of the total vocal tract as a series of resonators while the partials within the spectrum are determined by the vocal folds. The vocal tract and the vocal folds can change independently of each other. When the cavities of the throat and mouth remain fixed, a laryngeal sound of lower pitch may be passed through the system, and the vowel characteristic will remain the same because the energy within each formant has not varied. Only the fundamental will be lower since it is deteEmined by the frequency of the vibration of the vocal folds. Appelman's discussion of the vowels is also very pertinent to present investigation. He states: One must have a series of acoustically stable sung vowels that will serve as a standard scale or measure by which the singer may compare all other sung vowels when they are pro- duced at any pitch level or any intensity and timbre. Such acoustically stable vowels have been designed by the author and are as follows: 1. The Basic Vowel--A standardized recorded vowel sound to be used as a point of phonemic reference for all singers. 2. The Quality Alternate Vowel-~A modification of the basic vowel sung. ton: 27D. Ralph Appelman, The Science of Vocal Pedagogy (Blooming- Indiana University Press, 1967), pp. 126-127. 20 3. The Pure Vowel-~An identifiable area, surrounding both the basic vowel and its quality alternate, which enables a singer to selsct the phoneme of his choice for any given pitch or intensity. 3 Appelman stated that the pure vowel ”is sung within the stable "29 For tenors, Appelman defined the stable vowel vowel pitch range. pitch range as c-l3l Hz to b-247 Hz. He further stated that when a pitch is sung above this range, vowel modification is necessary to maintain quality and vowel recognition. In other words, the vowel must be modified by changing the basic vowel coloring.l Two of the three pitches used in the present investigation fall within Appelman's stable vowel pitch range (e-l65 Hz and b-246 Hz). The discrepancy in frequency on the pitch b is comnon depending upon the equipment or acoustical reference material used. All subjects in the present investigation were made aware of the fact that the vowels sung on the pitch g#-415 Hz must be modified. 30 dealt with quality judgment of sung The research by Jones vowels relative to different positions of the velum. Cinefluorography was used to observe the proper positioning of the velum during the course of the study. Seven baritone and bass singers were used as subjects, each singing three vowel phonemes, [i], [a], and [u], on two pitches, using three positions of the velum. The two pitches . were e-33O Hz and c-128 Hz, although a discrepancy of 2 Hz occurs throughout the study on the latter pitch depending upon the reference in the text or corresponding charts; the text stating c-128 Hz and the 28 29 Ibid., p. 223. Ibid., p. 228. 30Jones, op. cit. 21 charts c-l30 Hz. The samples were recorded and presented to selected auditors for their judgments of vowel recognition, nasal quality, and preference. Results show that the auditors deviated very little in their opinion concerning vowel recognition and they strongly preferred non- nasal tones. Vowels produced with a completely Open nasal port with the velum sagging down and away from the pharyngeal wall were most preferred by the auditors. Of these, the samples with a preferred rating of 90 percent or better were examined for formant behavior. Conclusions regarding formant behavior were as follows: The evidence presented in the discussion and in the graphs of the acoustical properties shows that, first, the fundamental pitch is consistently strengthened when a pho- neme is sung with some degree of naso-pharyngeal space, and, second, that there is a general tendency for partials in the lower frequencies of the spectrum to increase in strength. If the added strength in the fundamental and in the lower frequencies of the spectrum are attributed to the addition of nasal space into the oral and pharyngeal coupled cavity system, then the increase in strength may be said to be the result of nasal resonance. If this is true, then nasal resonance and nasal quality are the same, for both are asso— ciated with the increase in amplitude of the fundamental and other partials in the lower frequencies of the sound spectrum, as well as those specific alterations pointed out in the above discussion and charts for the several vowel pho- nemes and pitches. Further, since the effect of naso- pharyngeal Opening causes a vowel phoneme to be less pre- ferred by auditors, it can be said that nasal resonance is an undesirable quality characteristic and should be avoided. The sounds which are preferred by auditors are those prg- duced without nasal quality or without nasal resonance. 1 3‘Ibid., pp. 235-236. 22 Sullivan's32 research compared the relationships between sub- jective evaluations of auditors and the acoustical properties of a given phoneme. Specifically, the study dealt with judgment of tone quality, vowel quality, vibrato rate, and the intensity of tone. Eighteen male singers representing a wide range of vocal quality, ranging from excellent to poor, were chosen for the study. The vowels studied were [a], [i], and [u] on two pitches: a#-234 Hz and f-347 Hz. The given frequencies, corresponding to selected pitches, are Sullivan's and may vary with frequencies reported by various acousticians. All samples were recorded and presented to a jury of expert voice teachers. The jury was given two tones to compare for each phoneme and pitch. The mean score of the judges' preferences was used to rank the tone of each singer. Each tone was analyzed spectrographically and the resulting formants were measured in terms of location, intensity, and width. The results were as follows: The intensity of each partial on both sections was mea- sured by taking the average of the four points. Measurement was in millimeters, each millimeter representing a change of approximately 5 db [dB] in amplitude. Each measurement was then described in terms of the ratio of the intensity of the partial to the fundamental, thus pro- viding a measure that could be compared throughout all the tones. Formants were measured for locations, intensity, and width. In many cases the location of the first two formants was so close together they overlapped. Here it was difficult to determine by measurement alone the location of the formant peak. The phonetic transcriptions were consulted. The 32Ernest G. Sullivan, "An Experimental Study of the Relation- ships between Physical Characteristics and Subjective Evaluation of Male Voice Quality in Singing" (Ph.D. dissertation, School of Music, Indiana University, 1956). 1 Still! II [I ‘l. I I [‘5 I11 ,I 23 formants were estimated using the sonagram patterns in con- junction with the known formants of the vowels as transcribed by the phoneticians. The width of a formant was considered to be the total frequency range covered by the formant. It was measured on the section. In the case of overlapping formants, the width was measured between the lowest points on either side of the peak. Where two or even three formants overlapped, the entire band width was also measured.33 Sullivan's conclusions relating to the present study are as follows: 5. The proximity of the strongest formant above formant 2 [F ] to a given optimum location is an index of quality as judged by voice teachers. The optimum location seems to vary accorglng to the vowel and the pitch at which the tone is sung. 2. The intensity of F5 [Fs designated as the strongest formant above F2 ] provides the most significant index of jury tonal preference. The influence of this intensity may be expressed in the following ways: a. As the intensity of Fs (measured absolutely) is increased, the tone is ranked higher in quality. b. The ratio of the intensity of F5 to the inten- sity of F1 has a high positive correlation with jury tonal preference. Fs should be between one and two times as strong as F1. c. There is some evidence of correlation between the ratio of the intensity of F5 to the intensity of the fun- damental and rank order of jury preference. As Fs is stronger in relation to the fundamental, the tone is con- sidered to be better. The evidence is not as conclusive here as in the preceding correlation.35 3. As formant 1 [F1] becomes weaker, the tone is judged better in quality. 4. There was no evidence from this study to justify the assumption that the width of F1 or F5 influences jury evalu- ation of the tone. This, of course, does not preclude the possibility of formant width affecting vocal quality.37 33Ibid.. pp. 46-47. 34 35Ibid.. pp. 164-165. 361919;; P- 155° Ibid., p. 164. 37Ibid., p. 166. 24 5. For tones with the vowel [i] it may be concluded that as the width of F2 becomes narrower, the tone is judged as improved. The [i] vowel was the only one studied which pro- vided a measurable second formant throughout all tones.3 Sullivan's final conclusion is extremely interesting since it relates his findings to a description of good male vocal tone. He concludes: Good male vocal tone may be described as one which can be sung at high intensity, its spectrum showing the following characteristics: a weak first formant; a narrow second for- mant with twice the energy of the first; a group of strong partials between 2,500 cps and 3,000 cps (Fs), the location of this group dependent upon the vowel being sung, and its intenégty equal to or stronger than that of the first for- mant. ‘ The study by Wash40 provides a design and information leading directly to the present investigation. The purpose of Wash's study was to determine the formant behavior of sung vowels selected as excellent by a panel of judges. Five baritone voices, judged to be of professional quality, sang the vowels [a], [i], and [u] on the pitches e-165 Hz, c-256 Hz, and e-33O Hz. The recording of the singers took place in an acoustically treated room. A panel of 22 judges was chosen to evaluate the quality of the tones and identify the vowels. The three most highly preferred tones from each sample were analyzed spectrographically. Spectrograms were analyzed in terms of frequencies, intensities, bandwidths, and locations of the funda- mental and first three formants. Wash's research resulted in the following conclusions: 381bid. 39Ibid., p. 167. 40Wash, op. cit. 25 l. Auditor judgment was relaible [sic] and had a high factpp of agreement in evaluating the quality of sung vow- els. 2. The relationships between the fundamental and F1 varied according to vowel and pitch as follows: a. For the vowel [a] the relative intensity of F1 increased as the pitch rose. Fl had a frequency of 660 Hz. when the fundamental was 165 Hz. or 330 Hz. When the fundamental was 256 Hz., the frequency of F1 was 512 Hz. b. For the vowel [i] F1 was absorbed by the fun- damental as the pitch rose. When the fundamental was 165 Hz., F1 was 330 Hz. When the fundamental was 256 Hz. or 330 Hz., F1 was on the same frequency as the fundamental. c. For the vowel [u] F1 was absorbed by the fun- damental as the pitch rose. When the fundamental was 165 Hz., F1 was 330 Hz. When the fundamental was 256 Hz., it had absorbed much of the energy of F1, but F1 was still evident as a separate energy peak at 512 Hz. When the fundamental was 230 Hz., it had absorbed most of the energy of F1. 3. The relationships between the fundamental and F2 varied according to vowel and pitch as follows: a. For the vowel [a] the intensity of F2 remained constant at the lower pitches whereas theintensity of the fundamental was low. These characteristics were reversed at the highest pitch. When the fundamental was 165 Hz., F2 had a frequency of 990 Hz. When the fundamental was 256 Hz., F2 had a frequency of 1024 Hz. When the fundamental was 330 Hz., F2 had a fre- quency of 990 Hz. b. For the vowel [i] (recognized as [I] by a con- sensus of auditors) the intensity of F2 remained constant at the lower pitch, whereas the intensity of the fundamental was low. On higher pitches the funda- mental increased as it absorbed Fl energy. When the fundamental was 165 Hz., F2 had a frequency of 1650 Hz. When the fundamental was 256 Hz., F2 had a fre- quency of 1792 Hz. When the fundamental was 330 Hz., F2 had a frequency of 1650 Hz. c. For the vowel [u] the intensity of F2 tended to decrease slightly in the middle range and to increase slightly at the highest pitch. There was an increase in the fundamental as the pitch rose and absorbed F1. When the fundamental was 165 Hz., the frequency of F2 “Ibid., p. 64. Ibid 26 was 990 Hz. When the fundamental was 256 Hz., the fre- quency of F2 was 768 Hz. When the fpndamental was 330 Hz., the frequency of F2 was 990 Hz. 3 4. The relationships between F1 and F2 were as follows: a. For the vowel [a] the intensity of F1 increases and that of F2 decreased as the pitch rose. b. For the vowel [i] the intensity of F1 was absorbed by the fundamental and that of F2 remained con- stant as the pitch rose. ~ c. For the vowel [u] the intensity of F1 was gradu- ally absorbed by the fundamsntal and that of F2 remained constant as the pitch rose. 4 5. F2 was brozg when F1 was narrow. The reverse of this was observed also. Wash had several conclusions regarding F3. Since the present investigation will not deal with the third formant those findings are not pertinent. Wash's final conclusion was that formants always occurred on partial frequencies. Summary of Related Research Research relating to the formant theories reveals that the absolute formant theory is not applicable to all voices. However, both the relative formant theory and the fixed formant thegry can easily be applied to vocal acoustics. Most of the related research indicates the first two formants are essential for vowel recognition, and formants beyond the second formant are insignificant and diffi- cult to measure accurately. 43Ibid., p. 65. 44Ibid., pp. 65-66. 4516id.. p. 66. 27 It is interesting to compare the findings of Appelman,46 48 and Wash.49 Jones,47 Sullivan, Wash discovered that formants showed greatest intensity on partial frequencies. Appelman, quoted earlier in this chapter, stated that formant frequencies need not be the same as those of partials, but they may coincide. Furthermore, the fixed formant locations given by Appelman do not agree with those found by Wash. Sullivan found the formants to be in the same general area as those of Wash, but not directly related to the partials. How- ever, Sullivan and Wash were in agreement on the variations in location and strength of F3. The lack of agreement may be due, in part, to recording environments, pitches, and differences in performer ability. The research of Jones agreed with that of Wash, in that formants showed greatest intensity on partial frequencies. Recommendations for future research from the four studies men- tioned above suggest that all vowels be analyzed at all singable pitch levels, and in all voice ranges and classifications. It is also sug- gested that the most desirable formants for singers be determined and a reliable formula be developed for accurately measuring the intensity and width of formants. One of the most interesting recommendations comes from Wash regarding the recording environment. He states: 46Appelman, op. cit. 47Jones, op. cit. Sullivan, op. cit. Wash, op. cit. 48 49 28 Although subjects were selected as excellent by compe- tent voice teachers, the evaluations of tones made by auditors in this study were not as high as expected. A possible explanation for ratings may be the clinical con- ditions under which tones were recorded and presented. It is recommended that future experiments should record tones in a room with acoustic propergies more normal to the average listening environment. 0 501bid., p. 69. CHAPTER III DESIGN AND PROCEDURES Before the research design for the present survey was com- pleted, the recording environment, the adjudicator's environment, and the method of presentation to the adjudicators had to be established. Two mature tenor voices were used for this experiment. Each voice was recorded on the same sound equipment in two different environments: (1) a'sound-treated room in the Department of Audiology and Speech Sciences at Michigan State University and (2) a voice studio. The tenors participating in the experiment were asked to sing two pitches: b flat-234 Hz and g-396 Hz. The results were arranged by flip of the coin as to order by environment for each pitch sung by each per- former, and then placed on tape for the adjudicators. A group of ten adjudicators (five university professors, three graduate assistants in voice, and two high school voice teachers) were asked to respond to the tape in terms of the most natural sound, as determined by each adjudicator, between the two performances of each pitch. The adjudi- cators followed the format of the Voice Experiment found in Appendix A. In addition to choosing the most natural sound, the adjudicators were asked (1) to determine the degree of difference between the two sounds in terms of great, moderate, or very little, and (2) to say whether they preferred to listen to the tape through earphones or from speak- ers. The results are shown in Table l. 29 30 TABLE l.--Results of the voice experiment to determine recording environment and adjudicator environment. Adjudicators . Recording Subject Samples Tone . Percent Average E”V1”°"me"t Selecting Pgrfig?t Differ- Tone q ence l 1* A Treated O 0 0 B Studio 100 1.9 2 2 ‘ A Treated 20 70 1.5 B Studio 10 1.0 l ' 3 A Treated 70 O 2.4 B Studio 30 2.7 2 4 A Studio 70 10 2.1 B Treated 20 2.5 1 5** A Studio 10 50 2.0 B Treated 40 1.3 2 6 A Treated 50 40 1.2 B Studio 10 3.0 1 7 A Studio 70 10 2.4 B Treated 20 3.0 2 8 A Studio 100 0 2.3 B Treated 0 0 *Samples 1 through 4 monitored through earphones. **Samples 5 through 8 monitored through speakers in a studio without earphones. 31 Samples 1 though 4 were monitored by the adjudicators through earphones while samples 5 through 8 were monitored through speakers in a studio. The figures for average difference were computed on the basis of great--3, moderate--2, and very little--1. All adjudicators indicated a preference to monitor the tape using earphones. The sam- ples were presented in the same order when heard through speakers, but the recording environment was still arranged by flip of the coin. For samples heard through earphones, 27.5 percent of the adju- dicators preferred tones produced in a sound-treated room, with 52.5 percent choosing the studio environment and 20 percent finding no difference in the two tones heard. The average difference was 2.3 for those choosing the sound-treated YTKml and 2.1 for those choosing the studio environment. For samples heard through speakers in a studio, 27.5 percent preferred tones produced in a sound-treated room, 47.5 percent pre- ferred tones produced in a studio, and 25 percent found no difference in the two tones heard. The average difference for those choosing the sound-treated room was only 1.6 while for those choosing the studio environment it was 2.4. Although the same tones were presented in samples 5 through 8, but in slightly different order than in samples 1 through 4, the adjudicators' preferences obtained by the speaker method were not consistent with those heard through earphones. Though the adjudica- tors selected the opposite tone from their previous choice approximately 40 percent of the time in samples 5 through 8, the tones produced in the studio were still preferred to those produced in the sound-treated 32 room by a margin of nearly two to one. Degrees of difference were generally higher for the studio environment than for the sound- treated room. In addition to the statistics shown in Table l, a spectro- graphic analysis was run on the samples in the voice experiment for both environments. Figures 1 and 2 show spectrographic comparisons for subject 1 between the sound-treated room and the studio environ- ment for the pitch b flat-234 Hz. Figures 3 and 4 show the same comparison for subject 2. Figures 5 and 6 show spectrographic com- parisons for subject 1 between the sound-treated room and the studio environment for the pitch g-396 Hz. Figures 7 and 8 show the same comparisons for subject 2. Figures 1 through 4 show that the formants occurred on the third and fifth partials for both subjects on the lower pitch, with only the first five and the twelfth partials registering intensity. The formant bandwidths were almost identifical for both subjects in either environment. Figures 5 through 8 show that the formants occurred on the second and third partials for both subjects on the higher pitch. Only the first four and the seventh partials regis- tered intensity. The fourth partial showed a very low intensity peak of approximately 8 dB for both subjects. In general, there is seemingly little difference in the inten- 2 for both subjects in either environment. sity levels for either F1 or F Figures 1 and 3 show a lower intensity level of the fundamental for both subjects when singing in the sound-treated room than that of the studio environment (Figures 2 and 4). However, Figures 5 through 8 Freq. 33 01050!“ 171.67 8919 2885 9199 0989 M N‘DddddNMMNwQQQBb-b 2688:385*g5~232~522 OmeS ognmogoowm 0897 Figure l.--Spectrograph of subject 1 singing b flat-234 Hz in a sound-treated room environment. 34 d a dNuAUImNmOO-h :1 21 71 91 91 21 81 61 oz 1a 22 ca pz 9: 3’ 3 db40 38 36 34 32 30 28 26 24 22 20 18 16 14 12 10 8 6 4 2 O M NCO-h-Od-DNNNN wooed A 4542-0101 0'1 pm 94385565555655;§§s§4§55§a Figure 2.--Spectrograph of subject 1 singing b flat-234 Hz in a studio environment. 35 d d d—b—b—D—l-Id-JN unwammuoocoo-nnwammwoocoo M _o 33 3 db 40 32 28 24 20 18 16 14 12 10 OMOQW V83 89? ZOL 986 Gill 7001 8891 ZL81 9012 0V8Z VLSZ 8083 3908 9L38 0198 VVL8 8L68 Zle 997? 089? 916? 8V19 3889 9199 Freq. Figure 3.--Spectrograph of subject 2 singing b flat-234 Hz in a sound-treated room environment. 0989 36 d d d-b-I-A-h-I d” ”N” anuammwmwo—Muammuamoggwam db 40 2." 3 78C 897 ZOL 986 0L11 7071 8891 ZL81 901Z 0783 7LSZ 8088 Z708 9L38 0198 77L8 8L68 2127 9777 0897 7167 8719 8889 9199 0989 Figure 4.--Spectrograph of subject 2 singing b flat-234 Hz in a studio environment. Freq. 37 8 6 4 2 0 wqaaawnwwgbhmmmmmuuu @4010 (Do-‘Ul‘DUN-‘Ul wfldgwwN-PUICDCQNI-I mugoooouumgmmmb SQNNMMdd8§ 30611000 00110005 mumbomn Figure 5.--Spectrograph of subject 1 singing g-396 Hz in a sound-treated room environment. 38 db 40 7096 0066 I',‘ 8 968 36L 8811 7891 0861 9L8Z ZLLZ 8918 7998 0968 9987 €9L7 8719 7799 0769 9889 Z8L9 8ZlL 7ZSL OZGL 9188 81L8 8016 Figure 6.--Spectrograph of subject 1 singing g-396 Hz in a studio environment. 39 db 40 8 6 4 2 0 NdddNN #0101 uuu Hm §S§§§§§§§§§§§§§§§9§§§§§§§ Figure 7.--Spectrograph of subject 2 singing g-396 Hz in a sound-treated room environment. 40 - gfldd-PNNOJ 8&5mmmmmuuvmmwmm Freq. 68.-56588551? aasgtwaseeexiew co 001M” 801mg 08100050016.)ng Figure 8.--Spectrograph of subject 2 singing g-396 Hz in a studio environment. 41 reveal this is not the case for the higher pitch, as the intensity of the fundamental is nearly identical in either environment. The study of these spectra reveals there is little difference between the two acoustical environments in terms of the formant pat- terns, bandwidths, and over-all formant behavior. Since the studio environment was preferred over the sound-treated room by a margin of nearly two to one, the adjudication results of this voice experiment will serve as a basis and justification for using samples from a studio environment in this investigation. In addition, the adjudica- tors in this investigation will use earphones when evaluating samples since the adjudicators in the voice experiment expressed their pref- erence to do so. All equipment used in recording and adjudicating the voice experiment was deemed to be reliable after a calibration check was made by the technician in the Audiology and Speech Sciences Depart- ment at Michigan State University. Selection of Subjects Tenor voices to be compared in this study were chosen from three levels: (1) high school voices (ages 16 to 18); (2) university voices (ages 19 to 24); and (3) mature tenor voices (ages 25 to 50). Subjects for these three groups were selected upon reconmendation by voice teachers and choral directors chosen from various high schools and universities throughout Michigan. A copy of the initial communi- cation sent to these teachers is found in Appendix B. A total of 35 teachers was contacted and responses were received from 21. A total of 10 subjects at each level was chosen to participate in the study. The selection was based entirely on the order in which affirmative 42 responses were received from the teachers, and the willingness and availability of those recommended to participate. Participating sub- jects were assigned numbers 1 through 10 for identification at each level. Equipment for Recording The recorder used for recording subjects was a Sony Stereo Tapecorder TC—6660. This is a four-track, stereophonic or monophonic machine with a frequency response of 20 to 22,000 Hz. The tape speed was 7-1/2 inches per second with all subjects recorded mon0phonically. The microphone used was a Sony model F-98 Cardioid Dynamic semi- directional, high impedance. Scotch recording tape number 201 was used and each subject was recorded in a studio similar to the one used in the voice experiment. Recording of the Singers After each singer was informed regarding the purpose of the study, he was allowed ample warm-up time and practice runs in order to assure readiness for recording. During this practice time each singer was checked for his ability to produce all pitches with accurate intonation and his ability to produce the five vowel sounds without modification. Volume levels were set for recording each singer during this warm-up period. To further insure uniformity of recording, the micr0phone was placed at mouth level and a line was marked on the floor for the singer's toes so that his mouth was approximately 15 inches from the microphone. The VU meters indi- cating the recording signal strength were calibrated prior to the 43 recording. The signal strength of each voice taped was controlled manually by volume controls so the VU meter remained between a minus two or three at approximately 62 dB. Singers were instructed as to the proper pitch, vowel, and duration of each pitch before each vowel recorded. Attacks and releases were controlled by hand signal with each pitch lasting six seconds. The recorder was stopped after the singing of each vowel. This procedure was repeated until each sub- ject had sung the five vowels used in this study on the three pitches e-165 Hz, b-246 Hz, and g#-415 Hz. The pitches chosen for the singers represent the lower, middle, and upper voice ranges of the tenor voice. All subjects had to be able to produce all pitches on all vowels accurately or they were eliminated from the study and replaced by another singer. Singers were #-415 Hz to insure they were not using especially checked on the pitch 9 falsetto voice. All pitches were chosen upon recommendation of Dr. J. Loren Jones of the voice faculty at Michigan State University. Information as to name, age, and address of each singer was recorded prior to his singing. Prior to singing, each singer identi- fied himself by assigned number on the tape. Preparation of the Adjudicator's Tape The taped samples were prepared in numerical order, beginning with the first singer. A separate tape was made for each of the three pitches and for each of the age groups; a total of nine tapes. Samples representing the five vowel sounds at each of the three pitches were recorded. The order of the samples was selected, by vowel, according to random number tables. Each sample was recorded twice 44 with a pause of three seconds between. Aural instructions on the tape kept the adjudicator informed as to the sample, pitch, and vowel performed. Instructions prior to adjudication were given verbally and then repeated on tape. The instructions can be found in Appendix C. In addition, the investigator was present to answer any questions not covered by instructions before proceeding. The separate tapes for each pitch and for each of the age groups of tenors participating lasted approximately 15 minutes in length per tape. The nine tapes totaled approximately 135 minutes of listening for each adjudicator. Selection of Adjudicators The adjudicators were voice teachers and choral directors, 16 of whom taught at universities and four of whom taught at secondary schools in Michigan. All of them were $010 adjudicators on the Michi- gan School Vocal Association's list of adjudicators. In order to qualify for this list, they must be actively involved in teaching voice and must attend one state sponsored adjudication clinic biennially. A total of 20 adjudicators was used in the present investigation. Presentation of Tapes to Adjudicators Because of a possible fatigue factor, each adjudicator listened to only three tapes at one time. This totaled approximately one hour including time for instructions and changing tapes. In an attempt to insure greater reliability among adjudicators, each adjudicator started with a tape different from that of the previous adjudicator. Tapes 45 were presented sequentially beginning with the first adjudicator. This sequence was repeated with the tenth and eighteenth adjudicators. After the playing of each sample, the adjudicator indicated the quality of the vowel heard based on a rating scale of l to 4: (1) excellent, (2) good, (3) fair, and (4) poor. The adjudicator's quality judgment for each vowel was placed on a score sheet and then transferred to a master chart for statistical analysis (see Appendix D). A Hoyt's Reliability Calculation was used to determine reliability among adjudicators. The equipment used to present the tapes to the adjudicators was a Sony 540 Tape Recorder, having a frequency response of 30 to 20,000 Hz, and Superex model PRO-B earphones, having a frequency response of 18 to 22,000 Hz. All tapes were presented at the same level of intensity to all adjudicators; thus, the playback environment was identical for all adjudicators. Samples Selected for Spectrographic Analysis Since this study deals with judgments of excellent quality, only those vowel samples receiving a simple majority of excellent ratings were used (i.e., 11 or more). In addition, the sample was eliminated from the final analysis if any single adjudicator rated the vowel as fair or poor. The number of samples selected as excel- lent within a given level varied according to vowel sound, age level of the performer, and pitch. For example, a singer may have produced what was judged to be an excellent [a] vowel but did not receive an excellent rating for the four remaining vowels. Some high school 46 tenors had problems producing excellent vowels at either the higher or lower pitches, but were rated excellent in the middle voice range. Preparation of Tapes for Spectrographic Analysis After the tapes had been presented to the adjudicators, results were compiled to determine the samples selected for analysis. The appropriate section of tape from the original recording for each vowel rated as excellent was removed and marked for identification. Each selected sample was then cut, eliminating the attack and release of each vowel, leaving a tape 31 inches in length. After sufficient practice to become proficient in splicing recording tape, loops were made and placed on a pegboard in readiness for transportation to the Department of Audiology and Speech Sciences for analysis. Spectrogrgphy The following design, procedures, and technical information 5] Wash's research continued 52 was first utilized and described by Jones. the same design and was helpful in providing additional information. The loops were converted into spectrographs at the Audiology and Speech Sciences laboratory at Michigan State University. With the assistance of the technician, proper connections and adjustments were made between the Ampex Model AG 440 tape recorder, the Bruel and Kjaer 2107 frequency analyzer, and the Bruel and Kjaer 2305 Level Recorder. When these are combined, the print-out of the spectrum is automatically 51Jones, op. cit., pp. 48-49, 55-57. 52Wash, op. cit., pp. 21-23. 26-27. 47 produced on frequency-calibrated paper. A reading of 10 dB was used on the VU meter to insure uniformity of signal strength. The analyzer has a linear frequency response of 2 to 40,000 Hz. The scanning of the spectrum is done in six stages: 20 to 63 Hz, 63 to 200 Hz, 200 to 630 Hz, 630 to 2000 Hz, 2000 to 6300 Hz, and 6300 to 20,000 Hz. The different ranges may be either included or excluded by a manually operated switch. The accuracy of the band pass characteristics, depending on the attenuation desired, is plus or minus 0.5 dB. Signal shaping is possible through three weighted net- works, or the signal may be analyzed linearly from 20 to 40,000 Hz, or 2 to 40,000 Hz. The level recorder has a frequency range of 2 to 200,000 Hz and is accurate to within 1 dB. It will accept input signals up to 100 volts and will respond to signals as low as 5 mV. The calibrated paper for the print-out is connected automatically to the different frequency ranges of the analyzing systems to which it is coupled. The writing speed of the stylus and the speed of the paper are selectable. The resulting spectrograph is a highly accurate reproduction of the location of frequencies within the tone being analyzed and a highly accurate indication of the relative strengths. Having this informa- tion, a comparison of formant patterns in vowels judged to be excellent in quality may be made. The following settings were used on the Bruel and Kjaer 2107 frequency analyzer and the 2305 Level Recorder: 48 Analyzer Meter range . . 100 dB, s 1 Input potentiometer . 5 Signal input . . . direct Weighting network . linear Frequency range . . 20 to 20,000 Hz Meter switch. . fast RMS Range multiplier . . 0 dB Frequency analysis octave selector . 40 dB Function selector Level Recorder Paper speed Continuous record Voltage selector Potentiometer Input potentiometer Input attenuator Lower limiting frequency. Writing speed . . automatic . 3 mm. per sec. . on . 115 . 50 dB range . 4 . 10 . 20 Hz . 40 mm. per sec. After the tape loop was in proper position, the gain control was adjusted to 20 dB, chart paper was set at 0, and the process started. After the spectrograph was completed, the subject informa— tion was transferred to the spectrogram. Each individual loop was prOcessed in the same manner. After all loops were completed, a detailed analysis of the formant patterns for all vowels, pitches, and age groups were made and developed. In addition, a Hoyt's Reliability Calculation was made at the Computer Center at Michigan State University for each vowel at each of the three pitches to check adjudicator reliability. The results and conclusions regarding these analyses will be presented in the following chapters. CHAPTER IV PRESENTATION OF DATA Statistics from the master score charts (Appendix D) show the evaluations of all subjects by the adjudicators. Only those vowel samples receiving a simple majority of excellent ratings were used for spectrographic analysis. A datum was eliminated from the spectro- graphic analysis if any adjudicator rated the vowel as fair or poor. Subjetts at each maturity level (high school, unversity, and mature voices) were given numerical identification from 1 to 10. Tables 2, 3, and 4 identify those subjects (by numerical identification) selected for spectrographic analysis for each pitch and vowel. The number of samples selected as excellent within a given level varied according to vowel, maturity level of the performer, and pitch. A sample of 120 tones was chosen for spectrographic analysis. There were 45 samples chosen at e-165 Hz, 41 at b-246 Hz, and 34 at # g -415 Hz. A sample of 51 mature tones was chosen as compared to 33 university and 36 high school tones. Adjudicator Reliability To determine adjudicator reliability, a Hoyt's Reliability Coefficient was calculated by the Control Data 6500 Computer at the Michigan State University Computer Center. A separate program, using all data, was run for each vowel at each of the three pitches, 49 50 TABLE 2.--Subjects selected for spectrographic analysis on pitch e-165 Hertz. High School University Mature VOW33S Subjects Subjects Subjects [a] 3 1 l 4 4 3 10 9 4 7 10 [e] 3 1 1 4 4 3 10 4 10 [i] 3 1 l 4 3 3 10 4 4 10 [o] 3 l 3 4 4 4 10 10 [u] 3 1 3 10 3 4 10 TABLE 3.--Subjects selected for spectrographic analysis on pitch b-246 Hertz. High School University-””3 Mature —3 3 VOWE3S Subjects Subjects Subjects [6] 5 3 3 1o 9 4 10 [e] 4 3 3 9 9 9 1o 10 [i] 9 3 3 1o 9 4 9 , 10 [0] ' 4 3 3 1o 4 4 9 ~ 9 10 [U] 5 4 3 9 9 4 1o 9 10 51 TABLE 4.--Subjects selected for spectrographic analysis on pitch g#-415 Hertz. High School University Mature Vowels Subjects Subjects Subjects [a] 9 3 4 10 9 9 10 [e] 3 3 3 10 9 4 9 [i] 9 1 3 10 3 4 9 [o] 9 3 3 10 9 9 10 [u] 9 3 4 10 9 10 producing a total of fifteen programs. The resulting print-out pro- duced a reliability score (r) and the standard error of measurement (SE) for each of the fifteen sets of data. The results are found in Table 5. Table 5 shows extremely high reliability coefficients for all vowels at all pitches, ranging from .9698 for the vowel [u] at g3-415 3-415 Hz. The standard errors of Hz to .9819 for the vowel [a] at 9 measurement are also very consistent, ranging from 2.2916 for the vowel [u] at b-246 Hz, to 2.7804 for the vowel [u] at g3-415 Hz. These data show adjudicators for this study to have been highly con- sistent in their judgment of vowel sounds and voice quality. The adjudicators used in this research apparently had similar concepts of tone quality for tenor voices at all maturity levels. 52 TABLE 5.--Scores for reliability and standard error of measurement for each vowel at each of the three pitches. Pitch and Vowel r SE e-165 Hz [a] .9797 2.3494 [e] .9761 2.6243 [i] .9783 2.4515 [0] .9791 2.2935 [u] .9744 2.4948 b-246 Hz [a] '.9745 2.4095 [b] .9767 2.3561 [i] .9788 2.3052 [0] .9723 2.5147 [u] .9785 2.2916 93-415 Hz [a] .9819 2.3520 [e] .9791 2.4537 [i] .9778 2.4933 [0] .9784 2.4980 [u] .9698 2.7804 Spectographic Analysis The remainder of this chapter deals with the evaluation of the 120 tones selected for spectrographic analysis. Each tone will be evaluated in terms of (1) intensity of the fundamental, (2) location 2 and intensity of F3, (3) location and intensity of F , and (4) location and intensity of prominent energy regions above F3, often referred to 53 as F3. A comparison will also be made within each maturity level and between the three groups used in this investigation. Figures 9 through 128 are graphs presenting the information found within each spectrograph. Each figure accurately illustrates the locations and intensities of various energy peaks. However, these figures do not show information such as vibrato rate, formant bandwidth, or other scale reproductions of various acoustical data shown on the original spectrograph. The figures do provide the data required for the purpose of this investigation. Each figure is identified as to performer, maturity level, pitch, and vowel. Vowel [a] at Pitch e-165 Hz Figures 9 through 11 are spectrographs showing the formant behavior of high school subjects singing the vowel [a] at e-165 Hz, and include the following information: 1. The intensities of the fundamental range from 28 to 32 dB. The variation in fundamental intensity for these subjects, and those to follow, are the result of differences in each voice and not the methods used in recording each singer. Much care was taken, as men- tioned in previous chapters, to insure that each subject was recorded at the same intensity on the VU meter of the tape recorder. 1 2. F occurs on the fourth partial (660 Hz) for all three subjects. Decibel readings are 38, 36, and 40, respectively, for high school subjects 3, 4, and 10. 2 3. F occurs on the sixth partial (990 Hz) for high school subjects 4 and 10. Decibel readings are 34 and 32, respectively. db 40 9911 0881 Figure 9.--S singing the vowel [a 9971 1 54 d d d-odd—o—s-o “Nubmmgm‘DO-‘K’wbmm‘lm d-fi-DNMMNMB moocoacongoo mdmb-fifl ON 0010010010019 ectrograph of high at e-l65 Hz. N a 38 Z 8 HM (DO (AG) (400.) 5 ‘00 01V -‘ ca GNOGN 0'1 01001001 school subject 3 Freq. 0'0me 55 d d duh-Odd ‘NubmmNm‘Do-inbmm p 991 088 967 099 938 066 9911 0381 9871 0991 9181 0861 9713 0183 9L73 0793 Figure lO.--Spectrograph of singing the vowel [a] at e-165 Hz. ddunns ammo- 8BS“88 888888 high school 96L8 0968 9317 subject 4 Freq. Omem 99 1 088 singing 56 '3’ 3 Part. A m fl-fi-I-‘d-DMNNNN 0300 (A) b m 9°885888888§§8888a888 Figure ll.--Spectrograph of high school subject 10 the vowel [a] at e-165 Hz. 57 F2 occurs on the seventh partial (1155 Hz) for high school subject 3, with a decibel reading of 36. 3 is present in all three 4. A prominent energy region above F spectrograms. This energy region ranges from the eleventh partial (1815 Hz) to the nineteenth partial (2970 Hz) at 14 to 20 dB for high school subject 3. The range for high school subject 4 is from the thirteenth partial (2145 Hz) to the twentieth partial (3300 Hz) at 10 to 21 dB. For high school subject 10 the energy region ranges from the tenth partial (1650 Hz) to the sixteenth partial (2640 Hz) at 10 to 18 dB. While previous studies may have designated this energy region F3, the present investigation deals only with the first two formants. Conclusions and recommendations regarding this energy region will be discussed in Chapter V. Figures 12 through 14 are spectrographs showing the formant behavior of university subjects singing the vowel [a] at e-165 Hz, and include the following information: 1. The intensities of the fundamental range from 29 to 32 dB. 2. F1 occurs on the fourth partial (660 Hz) for all three subjects. Decibel readings are 40, 38, and 34.5, respectively, for university subjects 1, 4, and 9. 3. F2 occurs on the sixth partial (990 Hz) for all three sub- jects. Decibel readings are 34 for all three subjects. 2 is present in all 4. A prominent energy region above F three spectrograms. This energy region ranges from the sixteenth partial (2640 Hz) to the twenty-fourth partial (3690 Hz) at 8 to 16 dB for university subject 1. The range for university subject 4 is Freq. '8 8 8 3 3 8 B 8 8 815 58 ddd-fid-fi—b ‘NUfimmNm‘Do-inkwm 991 088 967 099 938 066 9911 0381 9871 0991 9181 0861 9713 0183 9L73 0793 Figure 12.--Spectrograph of singing the vowel [a] at e-165 Hz. '0 d 33 8 7 93 .0 0 q F’ dd-fin NNCDO 9083 0L63 9818 0088 9978 0898 96L8 0968 9317 university subject 1 db 59 d d debMU’NQQO-fi 31 81 71 91 91 L1 81 61 03 13 33 83 7 9 8 8 8 8 823 991 088 967 099 938 066 9911 0381 987 0991 9181 0861 9713 0183 9L73 0793 9083 0L63 9818 0088 9978 0898 96L8 0968 9317 Figure 13.--Spectrograph of university subject 4 singing the vowel [a] at e-165 Hz. Freq. 0105mm 60 d d u-I-I—l—bd inwbmmum‘Do-‘N‘mem 991 088 967 099 938 066 9911 0381 9871 0991 9181 0861 9713 0183 9L73 0793 Figure 14.--Spectrograph of singing the vowel [a] at e-l65 Hz. 13 33 3 3 3 dd...” NGCOO 9083 0L63 9818 0088 9978 0898 96L8 0968 9317 university subject 9 61 from the fifteenth partial (2475 Hz) to the twenty-fifth partial (4125 Hz) at 10 to 19 dB. The energy region ranges from the tenth partial (1650 Hz) to the eighteenth partial (2970 Hz) at 6 to 10 dB for university subject 9. Figures 15 through 19 are spectrographs showing the fOrmant behavior of mature subjects singing the vowel [a] at e-165 Hz and include the following information: 1. The intensities of the fundamental range from 28 to 33 dB for all subjects except mature subject 10, who registers a funda- mental intensity of 22 dB. This lower intensity cannot be accounted for as the adjudicators did not evaluate subject 10 significantly better or worse than the other four subjects. 3 occurs on the fourth partial (660 Hz) for all five 2. F subjects. Decibel readings are 38, 39, 36, 38, and 36, respectively, for mature subjects 1, 3, 4, 7, and 10. 3. F2 occurs on the sixth partial (990 Hz) for all subjects except mature subject 3. Decibel readings on the sixth partial are 32, 32, 36, and 32, respectively, for mature subjects 1, 4, 7, and 10. The decibel reading for mature subject 3, occurring on the seventh partial (1155 Hz),is 40. 2 is present in all 4. A prominent energy region above F Spectrograms except that of mature subject 7. This energy region ranges from the twelfth partial (1980 Hz) to the twenty-second partial (3630 Hz) at 14 to 18 dB for mature subject 1. The range for mature subject 3 is similar to that of mature subject 1, but at 8 to 17 dB. The range for mature subject 4 is a very intense peak from the db 3 3 B Freq. 62 40 38 36 34 32 30 28 26 24 16 14 12 10 8 6 4 2 0 8888888888§8§§§§§§S§8§§§ 0'10 00100100100191” 01001 Figure 15.--Spectrograph of mature subject 1 singing the vowel [a] at e-165 Hz. 9317 63 dwbmoo d-fi-e-i-I-DNNMN 8888888888888888 momomomomo Figure 16. Spectrograph of singing the vowel [a] at e-165 Hz. 8 5 a w 3 8 B 5 mature subject 3 db Freq. 64 ‘ ... d-b-l—b—I-b—l-DN anubmmuoocoo-snuammuoocoo 13 33 83 73 40 32 28 26 24 20 18 16 14 12 10 ONAO’N 991 088 967 099 938 066 9911 0381 9871 0991 9181 0861 9713 0183 9L73 0793 9083 0L63 9818 0088 9978 0898 96L8 0968 Figure l7.--Spectrograph of mature subject 4 singing the vowel [a] at e-165 Hz. 9317 Freq. onumoo 65 d d debmmNQQDO-fi 31 81 71 91 91 991 088 967 099 938 066 9911 0381 9871 0991 9181 0861 9713 0183 9L73 0793 Figure 18.--Spectrograph of singing the vowel [a] at e-165 Hz. NNUQU coco-owbggg ouuomumg 01001001601 mature subject 7 9317 66 ... g m d—b-A-A-l—LMNNM m§m§~§dwgmmmawbg 01 0'1 0'1 GIN m—ooxaau 0'10 0010010010 Figure l9.--Spectrograph of singing the vowel [a] at e-165 Hz. m ‘00me -A CVQSQQ‘DGN mom 01001001 mature subject 10 67 sixteenth partial (2640 Hz) to the twenty-fourth partial (3960 Hz), the highest intensity peaking on the twenty-first partial (3465 Hz) at 27 dB. The range for mature subject 10, also with an intense peak, is from the sixteenth partial (2640 Hz) to the twenty-third partial (3795 Hz), the highest intensity peaking on the nineteenth partial (3135 Hz) at 29 dB. A Comparison of All Subjects Singing_ the Vowel,[a] at e-165 Hz 1. With the exception of mature subject 10, the intensities of the fundamental range from 28 to 33 dB for all subjects. . 2. F1 occurs on the fourth partial (660 Hz) for all subjects with decibel readings ranging from 34.5 to 40. 3. F2 occurs on the seventh partial (1155 Hz) for high school subject 3 and mature subject 3 with decibel readings of 36 and 40, respectively. F2 occurs on the sixth partial (990 Hz) for all other subjects with decibel readings ranging from 32 to 36. V 4. A prominent energy region above F2 is present in all spec- trograms except that of mature subject 7. The length and intensity of this energy region varies among subjects, occurring between the thirteenth (1980 Hz) and twenty-first (3465 Hz) partials at 10 to 20 dB. Vowel [e] at Pitch e-165 Hz Figures 20 through 22 are spectrographs showing the formant behaviortyfhigh school subjects singing the vowel [e] at e-165 Hz, and include the following information: 68 d d debmmNm‘DO-P 31 81 71 91 91 L1 81 61 03 14 onumoo 991 088 967 099 938 066 9911 0381 87 0991 9181 0861 9713 0183 9L73 0793 9083 0L63 9818 0088 9978 0898 Freq. Figure 20.--Spectrograph of high school singing the vowel [e] at e-165 Hz. 96L8 0968 9317 subject 3 69 Part. db 40 Freq. duh m a... ..5 O) ‘0 ”8‘“ ddMNMMMMQM (db) 5 0'80,ng amended» coco—c.1353 ~18 8898888888888888888 Figure 21.-—Spectrograph of high school subject 4 singing the vowel [e] at e-165 Hz. 70 d d dNQ§01mNm®Od 31 81 71 91 91 L1 81 61 03 db 4O 991 088 967 099 938 066 9911 0381 871 0991 9181 0861 9713 0183 9L73 0793 9083 0L63 9818 0088 9978 0898 96L8 0968 9317 Freq. Figure 22.--Spectrograph of high school subject 10 singing the vowel [e] at e-165 Hz. 71 l. The intensities of the fundamental range from 28 to 34 dB. 2. F1 occurs on the third partial (495 Hz) for all three subjects. Decibel readings are 38, 40, and 38, respectively, for high school subjects 3, 4, and 10. 3. F2 occurs on the tenth partial (1650 Hz) for high school subject 10, with a decibel reading of 34. F2 occurs on the eleventh partial (1815 Hz) for high school subjects 3 and 4 with decibel read- ings at 34 and 28, respectively. 2 is present in all 4. A prominent energy region above F three spectrograms. This energy region begins at approximately the fourteenth partial (2310 Hz) at 18 to 24 dB with a gradual decrease in intensity to 0 dB at the twenty-fifth partial (4125 Hz). Figures 23 and 24 are spectrographs showing the formant behavior of university subjects singing the vowel [e] at e-165 Hz, and include the following information: 1. The intensities of the fundamental are 31 and 33 dB. 2. F1 occurs on the third partial (495 Hz) with decibel readings at 36 for both subjects. 3. F2 occurs on the tenth partial (1650 Hz) for both subjects, with decibel readings at 34 for university subject 1 and 33 for uni- versity subject 4. 4. A prominent energy region above F2 is present in both spectrograms. This energy region ranges from the twelfth partial (1980 Hz) to the twenty-fifth partial (4125 Hz) with energy peaks occurring at 24.5 dB and 30 dB at approximately the fifteenth partial (2475 Hz) and the twenty-first partial (3465 Hz), respectively, with Freq. 72 31 81 71 91 91 L1 81 61 03 dd dNMbU‘ONQ‘DO-fi a 3 3 8 8 8 8 S 8 8 813 duh-Ad 0104501 OMbOm 991 088 967 099 938 066 9911 0381 871 0991 9181 0861 9713 0183 9L73 0793 9083 0L63 9818 0088 9978 0898 96L8 0968 9317 Figure 23.--Spectrograph of university subject 1 singing the vowel [e] at e-165 Hz. 73 db 40 38 36 34 32 30 28 26 24 22 20 18 16 14 12 10 8 6 4 2 O as 8§§§§§88§§§§§§§§§§8§§§§§5 0'1 omomomomom 01001 01 Figure 24.—-Spectrograph of university subject 4 singing the vowel [e] at e-165 Hz. 74 a gradual decrease in intensity of between 10 and 15 dB through the twenty-fifth partial (4125 Hz). Figures 25 through 28 are spectrographs showing the formant behavior of mature subjects singing the vowel [e] at e-165 Hz, and include the following information: 1. The intensities of the fundamental range from 30 to 32 dB for all subjects except mature subject l0. Subject l0 shows a funda- mental intensity of 22 dB, identical to his fundamental intensity for the vowel [a] at the same pitch. 2. F1 occurs on the third partial (495 Hz) for all four subjects. Decibel readings are 36, 38, 38, and 40, respectively, fbr mature subjects 1, 3, 4, and 10. 3. F2 occurs on the eighth partial (l320 Hz) for mature sub- 2 occurs on the ninth partial ject 4, with a decibel reading of’ 32, F (1485 Hz) for mature subjects 3 and l0, with decibel readings of 32 and 24, respectively. F2 occurs on the eleventh partial (l8l5 Hz) for mature subject 1, with a decibel reading of 34. 4. A prominent energy region above F2 is present in all four spectrograms. This energy region ranges from approximately the fourteenth partial (23l0 Hz) to the twenty-second partial (3630 Hz) at 26 to 29 dB with a sharp drop to 0 dB at the twenty-fifth partial (4125 Hz). A Comparison of All Subjects Singing the Vowel_[e]_at e-l65 Hz l. With the exception of mature subject lo, the fundamental intensities range from 28 to 34 dB for all subjects. Freq. 75 ONAG’Q duh-03m dd-I-D-D-DNNNNN (AN (A) b assse§asaeeezesse§a§§§a§a momomomomomom 01001 0'1 Figure 25.--Spectrograph of mature subject 1 singing the vowel [e] at e-l65 Hz. 76 d «5 m d-‘d-fi-‘HNMNNN NC») “(A & e§e§a§asaeeexeagg§a§§ea§a WOUOMOU‘OU‘ 00" 01001 01 Figure 26.--Spectrograph of mature subject 3 singing the vowel [e] at e-165 Hz. 77 8 8 g 8 g 8 8 fl 8 833 18 16 14 12 owamm Freq. dwbmmg-‘d-‘dd-OMNNNN (JO-3001.000 5 $819 a flu DING-“(~15 Bdwbmflgd “8 °esgeasaaa§§aeseeaea Figure 27.--Spectrograph of mature subject 4 singing the vowel [e] at e-165 Hz. Part. 78 dd de'bmmumOOd 9 8 8 31 81 wt 91 91 L1 31 61 3 8 8 24 20 18 16 14 10 010me 991 088 967 099 938 066 9911 0381 891 0991 9181 0861 9713 0183 SLVZ OVQZ 9083 OLSZ 9818 0088 SSVE 0898 SGLC 0968 SZlb Freq. Figure 28.--Spectrograph of mature subject l0 singing the vowel [e] at e-l65 Hz. 79 2. F1 occurs on the third partial (495 Hz) for all subjects with decibel readings ranging from 36 to 40. 3. F2 occurs on the eighth partial (1320 Hz) for mature sub- ject 4, on the ninth partial (l485 Hz) for mature subjects 3 and l0, on the tenth partial (1650 Hz) for high school subject 10 and both university subjects, and on the eleventh partial (1815 Hz) for high school subjects 3 and 4, and mature subject l. With the exception of high school subject 4 and mature subject l0, decibel readings for F2 range from 32 to 40. 4. A prominent energy region above F2 is present in all spec- trograms. The pattern is very similar for all subjects, ranging from approximately the fourteenth partial (2310 Hz) to the twenty-second partial (3630 Hz) at 26 to 29 dB. Most Spectrograms show a sharp decrease in energy to 0 dB at the twenty-fifth partial (4l25 Hz) with no prominent energy peaks in the energy region. Vowel [i] at Pitch e-l65 Hz Figures 29 through 3l are spectrographs showing the formant behavior of high school subjects singing the vowel [i] at e—165 Hz, and include the following information: l. The intensities of the fundamental range from 30 to 38 dB. 1 2. F occurs on the second partial (330 Hz) for all three subjects. Decibel readings are 38, 38, and 40, respectively, for high school subjects 3, 4, and l0. 3. F2 occurs on the eleventh partial (l8l5 Hz) for high 2 school subject 10, with a decibel reading of 32. F occurs on the db 40 Freq. 80 1 Z 8 P 9 9 L 8 6 0 1 31 81 171 S1 91 L1 81 61 OZ 13 ll 83 V3 9 3 2" #00501“ ddddd-‘NMNMM wwwwu & eesee§asaeeezesse§a§333§a 01001001001001001001 01001001 Figure 29.--Spectrograph of high school subject 3 singing the vowel [i] at e-l65 Hz. Freq. 8] ddddddddd '0 MN duubmmqmocanuaamummB—SuJ-g Part. 8 B 8 3 B 8 S 8 8 815 6 4 2 0 Figure 30.--Spectrograph of high school singing the vowel [i] at e-l65 Hz. 5628 0968 5219 subject 4 Freq. 82 ‘095 m dd-‘d-A-DNMNNM “MUN“ b e88§e§asaeeaxesss§a§s§aga 01° 00100100100100“ 0‘ U1 0'1 Figure 31.--Spectrograph of high school subject 10 singing the vowel [i] at 3-165 Hz. 83 thirteenth partial (2145 Hz) for high school subjects 3 and 4, with decibel readings of 32 recorded for both subjects. 2 is present in all 4. A prominent energy region above F three spectrograms. It appears as an energy peak at approximately the seventeenth partial (2805 Hz) at 26 to 30 dB with a gradual decrease to 20, 3.5, and 9 dB at the twenty-fifth partial (4l25 Hz) for high school subjects 3, 4, and lo, respectively. Figures 32 through 34 are spectrographs showing the formant behavior of university subjects singing the vowel [i] at e-l65 Hz, and include the following information: 1. The intensities of the fundamental range from 34 to 36 dB. 2. F1 occurs on the second partial (330 Hz) for all three subjects. Decibel readings are 36, 40, and 38, respectively, for university subjects l, 3, and 4. 3. F2 occurs on the eleventh partial (l8l5 Hz) for all three subjects. Decibel readings were 30 and 28 for university subjects 1 and 4, respectively, with university subject 3 registering a much 2 intensity of 20 dB. lower F 4. A prominent energy region above F2 is present in all three spectrograms. This energy region ranges from the thirteenth partial (2145 Hz) to an energy peak at the nineteenth partial (3l35 Hz) at 23 and 26 dB for university subjects 1 and 3, respectively. The energy peak occurs at approximately the twenty-second partial (3630 Hz) at 19.5 dB for university subject 4. All three subjects show sharp decreases in intensity through the twenty-fifth partial (4l25 Hz) after the energy peak occurs. 84 20 18 16 14 12 10 8 6 4 2 0 .. addd-fidNMNNM wwwww b Hm $§§§§§aseaasasss§§a§§§§§s Figure 32.--Spectrograph of university subject l singing the vowel [i] at e—165 Hz. 85 d an. ‘wabUIOJNQIDO-D Z1 81 71 91 91 L1 81 61 OZ db 40 a 8 8 8 8 8 8 8 8 8 8 I.” 8 991 088 967 099 9Z8 066 9911 0Z€1 8V1 0991 9181 0861 SVlZ 01€Z SLVZ ovez 908Z OL6Z 9818 0088 9978 0898 96L€ 0968 931v Figure 33.--Spectrograph of university subject 3 singing the vowel [i] at e-l65 Hz. db 86 d .5 d-D-l—Dd—b-l-IMN ”N” uuwammammo—nwammummo-ngubm Part. 8 8 813 L" 8 991 088 967 099 9Z8 066 9911 Z81 871 0991 9181 0861 971Z 018Z 9L7Z 079Z SO8Z OL6Z 9818 0088 9978 0898 96L8 0968 9Z17 Figure 34.--Spectrograph of university subject 4 singing the vowel [i] at e-lGS Hz. 87 Figures 35 through 38 are spectrographs showing the formant behavior of mature subjects singing the vowel [i] at e-l65 Hz, and include the following information: l. The intensities of the fundamental range from 28 to 34 dB. 2. F1 occurs on the second partial (330 Hz) for all four subjects. Decibel readings are 36, 4D, 40, and 38, respectively, for mature subjectsl, 3, 4, and lo. 3. F2 occurs on the eleventh partial (l815 Hz) for all four subjects. Decibel readings are 30, 26, 28, and 28, respectively, for mature subjects l, 3, 4, and 10. 2 is present in all four 4. A prominent energy region above F spectrograms. ‘It appears as two energy peaks covering partials from the fourteenth (23l0 Hz) to twenty-first (3465 Hz) at between 23 and 29 dB. All subjects show a sharp decrease in intensity to 0 dB at the twenty-fifth partial (4125 Hz) after the second peak in this energy region occurs. A Comparison of All Subjects Singing the Vowel [i],at e-l65 Hz l. The intensities of the fundamental range from 28 to 38 dB. 2. F1 occurs on the second partial (330 Hz) for all subjects with decibel readings ranging from 36 to 4D. 2 3. F occurs on the thirteenth partial (2145 Hz) for high school subjects 3 and 4, with decibel readings of 32 for each subject. F2 occurs on the eleventh partial (lBl5 Hz) for all other subjects with decibel readings ranging from 26 to 32. Although university subject 3 shows a very strong fundamental and F1 intensity, his 88 d-ldd-fl-I-b-IN anuammummo-‘nwamo’umwo 14 ova-mm 991 088 967 099 9Z8 066 9911 0Z81 871 0991 9181 0861 971Z 018Z 9L7Z 0798 908Z 0L6Z 9818 0088 9978 0898 96L8 0968 Freq. Figure 35.--Spectrograph of mature subject l singing the vowel [i] at e-l65 Hz. 9Z17 89 d d dfiwammqmood Z1 81 71 91 91 L1 81 61 OZ db 40 8 8 8 28 26 24 22 20 18 16 12 10 010me 991 088 967 099 9Z8 066 9911 Z81 871 0991 9181 0861 971Z 018Z 9L7Z 0792 908Z OL6Z 9818 0088 9978 0898 96L8 0968 9Z17 Freq. Figure 36. Spectrograph of mature subject 3 singing the vowel [i] at e-l65 Hz. 90 d .5 d—D—i-fidd—b-fi” aniwammucocoo-suwammuoowo MNNNN auubm Part. 8 8 8 8 8 8 8 8 8 8 813 1‘ 8 991 088 967 099 9Z8 066 9911 0Z81 87 0991 9181 0861 971C 018Z 9L7Z O79Z 9083 0L6Z 9818 0088 9978 0898 96L8 0968 9Z17 Figure 37.--Spectrograph of mature subject 4 singing the vowel [i] at e-l65 Hz. 91 d d anwbmmuoocoo-n 1Z 7Z 9Z Part. Z1 81 71 91 91 L1 81 61 OZ ZZ 8Z db 4O 32 28 26 24 22 2O '18 16 14 12 1O ONDO’Q 991 088 967 099 9Z8 066 9911 0Z81 9871 0991 9181 0861 971Z 018Z 9L7Z 0792 908Z OL6Z 9818 0088 9978 0898 96L8 0968 9Z17 Freq. _ Figure 38.-—Spectrograph of mature subject 10 Singing the vowel [i] at e-165 Hz. 92 2 decibel reading for F is only 20 dB, considerably lower than all other subjects. 2 is present in all 4. A prominent energy region above F spectrograms. It appears as one or two energy peaks covering partials from the fourteenth (23l0 Hz) to the twenty-second (3630 Hz) at between 23 and 30 dB. All subjects show a decrease in intensity through the twenty-fifth partial (4125 Hz) after the last energy peak OCCUY‘S . Vowel [o]_at Pitch e-l65 Hz Figures 39 through 4l are spectrographs showing the formant behavior of high school subjects singing the vowel [0] at e-l65 Hz, and include the following information: l. The intensities of the fundamental range from 28 to 33 dB. 2. F‘ occurs on the third partial (495 Hz) for all three subjects. Decibel readings were 38, 38, and 40, respectively, for high school subjects 3, 4, and lo. 3. F2 occurs on the fourth partial (660 Hz) for high school 2 occurs on the fifth partial subject 3, with a decibel reading of 34. F (825 Hz) for high school subjects 4 and lo, with decibel readings of 32 for each subject. 4. A region of intensity was present in all three spectrograms. This region is relatively weak in intensity, ranging from the eleventh partial (l8l5 Hz) to approximately the twenty-second partial (3630 Hz) at 16 to 4 dB. 93 9L7Z 079Z 908Z OL6Z 98 18 0088 9978 0898 96L8 0968 9Z 17 ... b m d—b-l-O-h-AMM e§a§e§asaeeeze 010 0010016 Figure 39.--Spectrograph of high school subject 3 singing the vowel [0] at e-l65 Hz. 94 db 8 8 8 8 813 28 26 24 2O 18 16 14 12 onhmco Freq. dub m ...-h ......oNN asmng—nugmmmdu m m°eemeaeas 9L7Z 079Z 908Z 0L6Z 9818 0088 9978 0898 96L8 0968 9Z17 Figure 40.--Spectrograph of high school subject 4 singing the vowel [0] at e-l65 Hz. 95 d d duofbmmuootoo-t Z1 81 71 91 91 L1 81 61 0 1 Z db 4O 8 8 8 8 8 28 26 24 22 20 18 16 14 12 010me 991 088 967 099 9Z8 066 9911 0Z81 9871 0991 9181 0861 971Z 018Z 9L7Z 0792 909Z 0L6Z 9818 0088 9978 0898 Freq. Figure 4l.--Spectrograph of high school singing the vowel [0] at e-l65 Hz. 9Z17 0968 96L8 subject l0 96 Figures 42 and 43 are spectrographs showing the formant behavior of university subjects singing the vowel [0] at e-l65 Hz, and include the following information: 1. The intensities of the fundamentals for the two subjects are 30 and 32 dB. 2. F1 occurs on the third partial (495 Hz) for both subjects, with both decibel readings at 40. 3. F2 occurs on the fourth partial (660 Hz) for university subject l, with a decibel reading of 38. F2 occurs on the fifth par- tial (825 Hz) for university subject 4, with a decibel reading of 34. 4. A region of intensity is present in both spectrograms. This region is relatively weak in intensity, ranging from the seventh partial (ll55 Hz) to approximately the twenty-second partial (3630 Hz) at l6 to 4 dB. Figures 44 through 46 are spectrographs showing the formant behavior of mature subjects singing the vowel [0] at e-l65 Hz, and include the following information: l. The intensities of the fundamental range from 24 to 30 dB. 2. F1 occurs on the third partial (495 Hz) for all three subjects. Decibel readings are 38, 38, and 40, respectively, for mature subjects 3, 4, and lo. 3. F2 occurs on the fourth partial (660 Hz) for all three subjects. Decibel readings are 32 for all three subjects. 4. A prominent region of intensity is present in all three spectrograms. This region ranges from the fourteenth partial (2310 Hz) to the twenty-second partial at 4 to 2l dB for mature subjects 3 and 4. 97 Freq. 991 088 967 099 9Z8 066 9911 0Z81 9871 0991 9181 0861 971Z 018Z 9L7Z 079Z 908Z 0L6Z 9818 0088 9978 0898 96L8 0968 9Z17 Figure 42.--Spectrograph of university subject l singing the vowel [0] at e-165 Hz. 8 8 8 Freq. 98 d b m ddddd-‘NNNNN 0000 OJ e§a§e§aeeeeezese§§a§§e (’10 0010010010 001 010 Figure 43.--Spectrograph of university singing the vowel [0] at e-l65 Hz. 96L8 0968 SZ 17 subject 4 Freq. onamm 99 d a dd-fi-D—D-O-i-IB duwammuoowo-nuwammqoom 1Z ZZ Z 991 088 967 099 9Z8 066 9911 0Z81 9871 0991 9181 0861 971Z 018Z 9L7Z 079Z 908Z 0L6Z 9818 0088 9978 0898 96L8 0968 9Z17 Figure 44.--Spectrograph of mature subject 3 singing the vowel [0] at e-l65 Hz. Freq. Omem 100 d d duubmmqmmo-n Z1 81 71 91 91 L1 81 61 OZ 991 088 967 099 9Z8 066 9911 0Z81 871 0991 9181 0861 971Z 018Z 9L7Z 079Z 908Z 0L6Z 9818 0088 9978 0898 96L8 0968 9Z17 Figure 45.--Spectrograph of mature subject 4 singing the vowel [0] at e-l65 Hz. lOl d-n-o—o—n-a du‘r‘mm‘lmmo-‘wam db 4O 38 36 34 32 3O 28 26 24 22 20 18 16 14 12 1O 8 6 4 2 0 He eeeeeeggegggggg Figure 46.--Spectrograph dad-org mucocoo O79Z 908Z 0L6Z 9818 0088 9978 0898 96L8 0968 9Z17 of mature subject l0 singing the vowel [0] at e-l65 Hz. 102 The region of intensity ranges from the tenth partial (l650 Hz) to the twenty-second partial (3465 Hz) at 8 to 28 dB for mature subject l0. A Comparison of All Subjects Singing the VowelgIo]:at e-l65 Hz l. The intensities of the fundamental range from 24 to 33 dB. 2. F1 occurs on the third partial (495 Hz) for all subjects. Decibel readings range from 38 to 40. 2 3. F occurs on the fifth partial (825 Hz) for high school subjects 4 and l0, and university subject 4, with decibel readings 2 occurs on the fourth partial (660 Hz) for ranging from 32 to 34. F all other subjects with decibel readings ranging from 32 to 38. 4. A region of intensity is present in all spectrograms. This region is relatively weak in intensity for high school and uni- versity subjects. However, a prominent region of intensity is present in the spectrograms of the mature voices, ranging from the twelfth partial (l980 Hz) to the twenty-second partial (3630 Hz) at 8 to 28 dB. Vowel [u] at Pitch e-l65 Hz Figures 47 and 48 are spectrographs showing the formant behavior of high school subjects singing the vowel [u] at e-l65 Hz, and include the following information: l. The intensities of the fundamentals are 32 and 34 dB. 2. F1 occurs on the second partial (330 Hz) for both subjects. Decibel readings were 38 for high school subject 3, and 40 for high school subject l0. db 5 8 8 8 Freq. 103 d d dNQbO‘IOfiNQ‘DO-P z: 81 71 91 91 L1 81 61 oz tz at 2: ya 9: 8’ 3 4O 32 28 26 16 14 12 10 onumoo 991 088 967 099 9Z8 066 9911 Z81 871 0991 9181 0861 971Z 018Z 9L7Z O79Z 908Z 0L6Z 9818 0088 9978 0898 96L8 0968 9Z17 Figure 47.--Spectrograph of high school subject 3 singing the vowel [u] at e-l65 Hz. Freq. 104 Omem -' 5mm ddddd-‘NMNMN NM 0000 b gge ”g-“ mee-e. ge-§emse- “8“ 8888888888 38 38383 Figure 48.--Spectrograph of high school subject l0 singing the vowel [u] at e-l65 Hz. l05 3. F2 occurs on the third partial (495 Hz) for high school subject 3 at 35 dB. F2 appears to be combined with F1 for high school subject 10 at 40 dB. 4. A weak region of intensity is present in the spectrogram of high school subject 3. This region ranges from the eighth partial (l320 Hz) to the twenty-fifth partial (4125 Hz) at approximately 4 to 14 dB. No region of intensity above F2 is present in the spectrogram of high school subject 10. Figures 49 and 50 are spectrographs showing the formant behavior of university subjects singing the vowel [u] at e-l65 Hz, and include the following information: l. The intensities of the fundamental are 36 dB for both subjects. 2. F1 occurs on the second partial (330 Hz) with decibel readings at 40 for both subjects. 2 appears to be combined with F1 for both subjects at 3. F 40 dB. 4. A weak region of intensity above F2 is present in both spectrograms. It ranges from the twelfth partial (1980 Hz) to the nineteenth partial (3l35 Hz) at approximately 2 to 9 dB. Figures Sl through 53 are spectrographs showing the formant behavior of mature subjects singing the vowel [u] at e-l65 Hz, and include the following information: l. The intensities of the fundamental range from 28 to 30 dB. 2. F1 occurs on the second partial (330 Hz) for all three subjects. Decibel readings are 40, 40, and 38, respectively, for mature subjects 3, 4, and lo. 106 db 4o 38 36 34 32 3O 28 26 24 22 20 18 18 14 12 1O 8 6 4 2 O ... add-bd-AMNNMN wwwww 5 mm a§§§§§aeaaaasgas§§a§§§§§s Figure 49.--Spectrograph of university subject 1 singing the vowel [u] at e-165 Hz. Freq. 107 9 8 8 8 3 8 § 8 8 8 16 14 10 8 6 4 2 0 a§§§§§asaa§§§§§§ 0‘ 01°01 010010 Figure 50.--Spectrograph of singing the vowel [u] at e-l65 Hz. 333w§§383 outage: wgn wow 01 m 01 university subject 3 Q. U' o n a m m 3 3 3 3 8 3 3 fl 3 8 8 8 fl 8 813 Freq. 108 d .n d-b-Idd‘d-I SONGNQ‘DO-‘NUOU‘G‘IQ‘Do-‘Nub an F'F §9l 088 96? 099 538 066 Sgtl OZEl 8ft 099l Sl81 086l SVlZ OlCZ SLVZ OVQZ 9083 OLGZ 981$ DOES 997$ 0898 SGLC 0968 SZIV Figure 5l.--Spectrograph of mature subject 3 singing the vowel [u] at e—165 Hz. a 8 3 g 3 B 8 8 9 8 813 Freq. 109 dddd—nddMMSNN 991 088 967 099 938 066 SSll OZCI 871 OSQl Sl8l 086l SVlZ OISZ 9L7Z 0093 9088 OLSZ EELS 0088 997€ 0398 96L8 0968 Figure 52.--Spectrograph of mature subject singing the vowel [u] at e-165 Hz. SZlP b llO OICZ 9618 0968 5319 SLVZ OVQZ 9083 OLGZ §€l€ 0082 9978 0898 Ha a§§§§§§§ggggg Figure 53.-~Spectrograph of mature subject l0 singing the vowel [u] at e-l65 Hz. lll 2 l 3. F appears to be combined with F at 40 dB for mature 2 subjects 3 and 4. F occurs on the third partial (495 Hz) with a decibel reading of 38 for mature subject l0. 2 is present in 4. A prominent region of intensity above F all three spectrograms. It ranges from the thirteenth (2145 Hz) to the twenty-first partial (3465 Hz) for all three subjects. Decibel readings range from 4 to approximately 24. A Comparison of All Subjects Singing the Vowel [u] at e-l65 Hz l. The intensities of the fundamental range from 28 to 36 dB. 1 2. F occurs on the second partial (330 Hz) for all subjects. Decibel readings range from 38 to 40. 2 occurs on the third partial (495 Hz) for high school 2 3. F subject 3 and mature subject l0 at 35 and 38 dB, respectively. F 1 for all other subjects at 40 dB. appears to be combined with F 4. A weak region of intensity above F2 is present in high school and university subjects; however, no consistent pattern is evident among subjects. The mature subjects show a more prominent region of intensity. Vowel [a] at Pitch b-246 Hz Figures 54 and 55 are spectrographs showing the formant behavior of high school subjects singing the vowel [a] at b-246 Hz, and include the following information: 1. The intensities of the fundamentals are 28 and 32 dB. Freq. ll2 cream 338983s§§§see§§ssez§safi§s Figure 54.--Spectrograph of high school subject 5 singing the vowel [a] at b-246 Hz. Freq. o u a a m 8 S 3 8 E 3 S g 8 8 8 8 2 8 813 113 d ‘...a dd-n—a-b-n—adlo dnqammqmwOanwbmmqmmo MbNCD-idd-DNNNBQ wgbfibbmfiflm g asesasswaaamageaaweea2e§g omwgbomnm omwooaomnm Figure 55.--Spectrograph of high school subject l0 singing the vowel [a] at b-246 Hz. 114 2. F1 occurs on the third partial (738 Hz) for both subjects. Decibel readings are 36 and 40, respectively, for high school subjects 5 and TD. 3. F2 occurs on the fourth partial (984 Hz) at 34 dB for both subjects. 2 is present in both 4. A prominent energy region above F spectrograms. This energy region ranges from the sixth partial (l476 Hz) for both subjects, with decibel readings ranging from approximately 4 to 22. Figures 56 and 57 are spectrographs showing the formant behavior of university subjects singing the vowel [a] at b-246 Hz, and include the following information: 1. The intensities of the fundamentals are 32 and 33 dB. 1 occurs on the third partial (738 Hz) for both subjects. 2. F Decibel readings are 33 and 32, respectively, for university subjects 3 and 9. 3. F2 occurs on the fourth partial (984 Hz) at 34 dB for both subjects. 2, with a strong peak at 4. A prominent energy region above F the twelfth partial (2952 Hz), is present in both spectrograms. This energy region ranges from the eighth partial (l968 Hz) to the seven- teenth partial (4l82 Hz) for both subjects. 11K: intensity peak at the twelfth partial ranges from l5 to 28 dB. Figures 58 through 60 are spectrographs showing the formant , behavior of mature subjects singing the vowel [a] at b-246 Hz, and include the following information: llS dub-ld-Id-l dNmemNm‘DO-‘N‘Jémw 8 8 8 8 8 8 B 8 8 833 18 1‘ 3 973 867 88L 786 0€Zl 9L7l ZZLl 896l 7lZZ 0973 90LZ 8963 8618 7778 0698 9868 Figure 56.--Spectrograph of singing the vowel [a] at b-246 Hz. -..—DNNNNNN game-Iowan: Part. 38l7 8377 7L97 OZ67 99l9 Zl7§ 8999 7069 0§l9 university subject 3 8 8 8 8 8 8 8 8 8 8 813 (O on b 9L7l ZZLL OCZL Figure 57 singing the vowel l16 896l 7tZZ 0978 SOLE ZSGZ 86l8 0698 9868 Z8l7 .--Spectrograph of [a] at b-246 Hz. 8877 7L97 0367 Zl7§ 7069 09l9 99l9 university.subject 9 Freq. 117 d d at at rt 9: 9: L1 81 61 oz 13 ca 8: rt 9: 3’ 3 d dNQbUIONmOO 8 8 8 8 8 8 8 8 8 8:5 18 16 14 12 ONbOQ 973 367 88L 786 0831 9371 33L1 8961 7133 0973 9033 3963 8618 7778 0698 9868 3817 8377 7L97 0367 9919 3179 8999 7069 0919 Figure 58.--Spectrograph of mature subject 3 singing the vowel [a] at b-246 Hz. dad-o OMbm Freq. 8.8 8 8 8 8 8 8 8 8 833 118 31 81 v: 91 9: L1 81 61 oz 13 at 8: rt 9 3’ 3 u... dNmea’NmmOd 01050“ 0367 9919 3179 8999 7069 0919 973 367 88L 786 0831 9L71 33Ll 8961 7133 0973 90L3 3963 8618 7778 0698 9868 3817 8377 7L97 Figure 59.--Spectrograph of mature subject 4 singing the vowel [a] at b-246 Hz. 119 d‘d d antitammqoooo add-A Mwbmm 8 8 8 8 8 8 8 8 8 8 8 813 2‘ .3 99 zev 88L 786 08:: 9L71 zzzl 8961 7133 osvz 90L3 zssz 8618 vvve 0698 9868 Figure 60.—-Spectrograph of singing the vowel [a] at b-246 Hz. 3 8 8 8 8 8 8 8 8 3 8 3 8 3 8 m g 9 (browns—g 01 MmbO N o mature subject 10 120 l. The intensities of the fundamental range from 24 to 30 dB. 2. F1 occurs on the third partial (738 Hz) for all three subjects. Decibel readings are 38, 38, and 40, respectively, for mature subjects 3, 4, and l0. 3. F2 occurs on the fourth partial (984 Hz) for mature subjects 4 and 10 with decibel readings at 34 and 36, respectively. F2 occurs on the fifth partial (1230 Hz) for mature subject 3 with a decibel reading of 34. 4. A prominent energy region above F2, with a strong peak at the twelfth partial (2952 Hz), is present in all three spectrograms. This energy region ranges from the eighth partial (1968 Hz) to the seventeenth partial (4182 Hz) for all three subjects. The intensity peak at the twelfth partial ranges from 26 to 30 dB. A Comparison of All Subjects Singing the Vowel [a]*at b-246 Hz l. The intensities of the fundamental range from 24 to 33 dB. 2. F1 occurs on the third partial (738 Hz) for all subjects. Decibel readings range from 32 to 40. 2 3. F occurs on the fifth partial (l230 Hz) for mature sub- 2 occurs on the fourth partial ject 3 with a decibel reading of 34. F (984 Hz) for all other subjects with decibel readings ranging from 34 to 36. 4. A prominent energy region above F2, with a strong peak at the twelfth partial (2952 Hz), is present in nearly all spectro- grams of university and mature subjects. This energy region ranges from the eighth partial (1968 Hz) to the seventeenth partial (4l82 Hz). 121 The intensity peak at the twelfth partial ranges from 15 to 30 dB. This energy region for both high school subjects ranges from the sixth partial (1476 Hz) to the seventeenth partial (4182 Hz), with decibel readings ranging from approximately 4 to 22. Vowel [e] at Pitch b-246 Hz Figures 61 through 63 are spectrographs showing the formant behavior of high school subjects singing the vowel [e] at b-246 Hz, and include the following information: 1. The intensities of the fundamental range from 30 to 34 dB. 2. F1 occurs on the second partial (492 Hz) for all three subjects. Decibel readings are 40, 4D, and 36, respectively, for high school subjects 4, 9, and 10. 3. F2 occurs on the sixth partial (1476 Hz) for all three subjects. Decibel readings are 32, 32, and 36, respectively. ~ 4. A prominent energy region above F2 is present in all three 2 to approximately spectrograms. This energy region extends from F the fifteenth partial (3690 Hz) at 16 to 29 dB. Figures 64 and 65 are spectrographs showing the formant behavior of university subjects singing the vowel [e] at b-246 Hz, and display the following information: 1. The intensities of the fundamentals are 33 and 34 dB. 2. F1 occurs on the second partial (492 Hz) for both sub- jects. Decibel readings are 40 and 38, respectively, fbr university subjects 3 and 9. 3. F2 occurs on the sixth partial (1476 Hz) at 34 dB for both subjects. 122 diamante:umooznazalaliaagfisaam Part. 40 :m as 34 32 30 23 26 24 22 20 13 18 14 12 10 8 6 4 2 o “a“mazzasxzae massacre 2 3”8§83£838883§§8883Bafiggs Figure 61.--Spectrograph of high school subject 4 singing the vowel [e] at b-246 Hz. 123 db 4O 8 3 8 S B 8 8 24 20 18 16 14 12 OM50!” 0919 NONCD-‘d-D-INMN u b «550‘ #0 Mbfl‘DNbNnggg-bgm‘D-lg “Nggeasazesss oeeezesa $3 Freq. Figure 62.--Spectrograph of high school subject 9 singing the vowel [e] at b-246 Hz. 3 a 8 3 § 3 B 8 8 9 8 823 ddd Nb ONAQQO Freq. Mb #49 (”N B 786 OCZt 9LVl Figure 63.--Spectrograph of high school singing the vowel 124 ‘dNNN (a) 5 550101 0105 B102.) N86§§883m¢0d2§0d Neeésse eacezaec 28 subject 10 [e] at b-246 Hz. 125 3 Part. ouamoo 036? 99 l 9 3 WS 8999 9069 05 £9 Nb‘ltD-I-l-fi-fiNMM co 5 b asaeusasegg§a§§§a§s OGMmb Mm oasrooob Freq. Figure 64.-—Spectrograph of university subject 3 singing the vowel [e] at b-246 Hz. a 3 S 9 8 B 8 S g 8 813 126 El 9t 9t 9t d d dMUbO‘QNmmO-DM 99 369 88L 986 0€3l 9L9l 33Ll 896l 9l33 0993 90L3 3963 86l£ 9998 0698 9868 Figure 65.--Spectrograph of singing the vowel [e] at b-246 Hz. Ll Bl 6l 03 l3 33 83 93 93 Part. 5555010101010: dbm‘D-Ibg -' oouwng-o 0‘ tho N!” O university subject 9 127 4. A prominent energy region above F2 is present in both spectrograms. This energy region extends from the eighth partial (1968 Hz) to the twenty-first partial (5166 Hz) at approximately 16 to 28 dB. Figures 66 through 68 are spectrographs showing the formant behavior of mature subjects singing the vowel [e] at b-246 Hz, and display the following information: 1. The intensities of the fundamental range from 24 to 28 dB. 2. F1 occurs on the second partial (492 Hz) for all three subjects. Decibel readings are 36, 32, and 38, respectively, for mature subjects 3, 9, and 10., . 3. F2 occurs on the sixth partial (1476 Hz) for all three subjects, with decibel readings at 36, 38, and 34, respectively. 4. A prominent energy region is present in all three spectro- grams. This energy region extends from the eighth partial (1968 Hz) to the seventeenth partial (4182 Hz), with decibel readings ranging from 20 to 32. Mature subjects 9 and 10 show strong intensity peaks on the ninth partial (2214 Hz) at 32 dB. A Comparison of All Subjects Singing the Vowel [e] at b-246 Hz 1. The intensities of the fundamental range from 24 to 34 dB. 2. F1 occurs on the second partial (492 Hz) for all subjects. Decibel readings range from 32 to 40. 3. F2 occurs on the sixth partial (1476 Hz) for all subjects with decibel readings ranging from 32 to 38. db 128 d d dd-l—ldd-l-INN ”NM nuammqooto0a'n3uom Part. d HMUOO'IONIQCDO 8888 1" 8 993 369 88L 986 O83l 9L9l 33Ll 896l 9l33 0993 90L3 3963 86l8 9998 0698 9868 3819 8399 9L99 0369 99l9 3l99 8999 9069 09l9 Figure 66.-~Spectrograph of mature subject 3 singing the vowel [e] at b-246 Hz. 129 a*°a;:aa~sae§~aets Freq. anggwuwgagomm agmnq <3 cute .5 care 63.5 c: no a:.u 0369 9919 3199 8999 9069 0919 Figure 67.--Spectrograph of mature subject 9 singing the vowel [e] at b-246 Hz. Freq. 130 i d ddddd wamm ... unubmmumwo 8 8 8 8 8 8 8 8 8 833 dad-5.. mONbaO ONOO 993 369 88L 986 0831 9L91 33L1 8961 9133 0993 90L3 3963 8618 9998 0698 9868 Figure 68.—-Spectrograph of singing the vowel [e] at b-246 Hz. 3 8 8 8 8 8 8 8 8 8399 9L99 0369 9919 3199 8999 9069 0919 3819 mature subject 10 131 4. A prominent region of energy is present in all spectro- grams. Although this energy region is not consistent for all subjects, it extends from approximately the eighth partial (1968 Hz) to the seventeenth partial (4182 Hz), with decibel readings ranging from 16 to 32. Mature subjects 9 and 10 show strong intensity peaks on the ninth partial (2214 Hz) at 32 dB. VowelLiJgat Pitch b-246 Hz Figures 69 and 70 are spectrographs showing the formant behavior of high school subjects singing the vowel [i] at b-246 Hz, and display the following information: 1. The intensities of the fundamentals are 38 and 40 dB. 2. F‘ is absorbed by the fundamental for both subjects at 38 and 40 dB. 3. F2 appears as a prominent region of energy for both sub- jects, ranging from an intense peak on the seventh partial (1722 Hz) to the twelfth partial (2952 Hz). Decibel readings on the seventh partial are 38 for both subjects. 2 is present in both 4. A prominent energy region above F spectrograms. This energy region extends from the twelfth partial (2952 Hz) to the seventeenth partial (4182 Hz) at approximately 20 to 32 dB. Both subjects show intense peaks on the fifteenth partial (3690 Hz) at 28 to 30 dB. Figures 71 and 72 are spectrographs showing the formant behavior of university subjects singing the vowel [i] at b-246 Hz, and display the following information: 8888881388188 132 8 Part. d PUOC‘ONQ‘DO 369 88L 986 0831 9L91 33L1 8961 9133 0993 90L3 3963 8618 3819 8399 9L99 0369 9919 3199 8999 9069 0919 £933 Figure 69.--Spectrograph of high school subject 9 singing the vowel [i] at b-246 Hz. 133 dddddddddd N ”MN dnwbmmunmeauubmmummguSQbm Part. db 40 38 36 34 32 30 28 26 24 22 20 18 16 14 12 10 8 6 4 2 0 N N‘D-h-e-I-DNNN u w b 5501 01 be 3§89§§§§§§§§§§§§§§§§§§§§§ Figure 70.--Spectrograph of high school subject 10 singing the vowel [i] at b-246 Hz. Freq. 134 ON§OQ 0369 9919 3199 8999 9069 0919 NbNCD-D-D-‘dNNN (a) (a) b b e ”some sass a-ss 3~8gassaz§ssss§ssgs Figure 7l.--Spectrograph of university subject 3 singing the vowel [i] at b-246 Hz. 135 1‘ 3 993 369 88L 986 0831 9L91 33L1 8961 9133 0993 90L3 3963 8618 9998 0698 9868 3819 8399 9L99 0369 9919 3199 8999 9069 0919 Figure 72. Spectrograph of university subject 9 singing the vowel [i] at b-246 Hz. 136 1. The intensity of the fundamental is 40 dB for both subjects. 2. F1 is absorbed by the fundamental for both subjects at 40 dB. 3. F2 appears as a prominent region of energy for both sub- jects, ranging from an intense peak on the seventh partial (1722 Hz) to the ninth partial (2214 Hz). Decibel readings on the seventh partial are 28 and 24, respectively, for university subjects 3 and 9. 4. A prominent energy region above F2 is present in both spectrograms. This energy region extends from the ninth partial (2214 Hz) to the seventeenth partial (4182 Hz), with a strong inten- sity peak of 29 dB for university subject 3, occurring on the fourteenth partial (3444 Hz). The energy region extends from the ninth partial (2214 Hz) to the sixteenth partial (3936 Hz) for university subject 9, ranging from 10 to 20 dB. Figures 73 through 76 are spectrographs showing the formant behavior of mature subjects singing the vowel [i] at b-246 Hz, and display the following information: 1. The intensities of the fundamental range from 34 to 38 dB. 2. F1 is absorbed by the fundamental for all four subjects. Decibel readings are 34, 38, 32, and 36, respectively, for mature sub- jects 3, 4, 9, and 10. 3. F2 occurs on the seventh partial (1722 Hz) for all four subjects. Decibel readings are 30, 28, 34, and 32, respectively. 4. Two additional energy peaks occur for all four subjects on the ninth partial (2214 Hz) and the thirteenth partial (3198 Hz). 137 dd-I-L-I-b-I-I-b-INN NNN duuammstopcodnubmmuoocoo-ofiwam Part. dbao 38 36 34 32 30 28 26 24 22 20 18 16 14 12 10 8 6 4 2 0 M \lm-O-b-DANNM to & 5&0! O) Fm- siaiessa:ee%s§§§e§:§e§§$e Figure 73.--Spectrograph of mature subject 3 singing the vowel [i] at b-246 Hz. db 9 8 813 9 91.91 138 :‘I—tNNSBg wggtbbgg 2 s§§§sasg§sss§§sa§§s Figure 74.--Spectrograph of mature subject 4 singing the vowel [i] at b-246 Hz. 139 .... dddddddd N anemone:unmeanuammummo-ssaam Part. db40 38 36 34 32 30 28 26 24 22 20 18 16 14 12 10 8 6 4 2 0 q ...-.... F... awheeeafisfimiieifis Figure 75.--Spectrograph of mature subject 9 singing the vowel [i] at b-246 Hz. 140 dddddddddd ””3”” dNQbGQNG‘DOdeémmNQOO-fi b u 81 Part. db 40 38 36 34 32 30 28 26 24 22 20 18 16 14 12 10 8 6 4 2 0 qu-b—h-I-DMNN «5 he afiesessate§§§§§§e§§§§§§§§ Figure 76.—~5pectrograph of mature subject 10 singing the vowel [i] at b-246 Hz. 141 Decibel readings range from 28 to 33 on the ninth partial, and from 26 to 32 on the thirteenth partial. A Comparison of All Subjects Singing the Vowel [i] at b-246 Hz l. The intensities of the fundamental range from 34 to 40 dB. 2. F1 is absorbed by the fundamental for all subjects. 3. F2 occurs on the seventh partial (1722 Hz) for all sub- jects, with decibel readings ranging from 24 to 38. However, F2 varied among subjects from a broad region of intensity for high school subjects, to a narrow intensity peak for mature subjects. F2 shows greater intensity for high school subjects (38 dB) than that of the university or mature subjects. 4. A prominent energy region above F2 is present in all spectrograms. This energy region ranges from the ninth partial (2214 Hz) to the seventeenth partial (4182 Hz) at 20 to 33 dB. Intensity peaks are present for all four mature subjects on the ninth partial (2214 Hz) at 28 to 33 dB, and on the thirteenth partial (3198 Hz) at 26 to 32 dB. Both high school subjects show intensity peaks on the fifteenth partial (3690 Hz) at 28 to 30 dB, and university subject 3 shows an intensity peak on the fourteenth partial (3444 Hz) at 29 dB. Vowel[o] at Pitch b-246 Hz Figures 77 and 78 are spectrographs showing the formant behavior of high school subjects singing the vowel [0] at b-246 Hz, and display the following information: 1. The intensities of the fundamentals are 30 and 32 dB. 142 dd—I-b-I-D—I-l-b-IN urtubmazuoocoo-anubmoummo db40 8 8 8 3199 E .3 99 369 88L 986 0831 9L91 33L1 8961 9133 0993 90L3 3963 8618 9998 0698 9868 3819 8399 9L99 0369 9919 Figure 77.--Spectrograph of high school singing the vowel [0] at b-246 Hz. 13 33 8999 9069 0919 subject 4 Freq. 143 ddddddddd an N dsfuammuoocoo-anuammumcoo-a who: Part. 8 8 8 8 8 8 8 8 8 815 xeggazsasssee gesssaee 2 COM b Mm OONmbO N C Figure 78.--Spectrograph of high school subject 10 singing the vowel [0] at b-246 Hz. 144 1 2. F occurs on the second partial (492 Hz) at 40 dB for both subjects. 2 occurs on the third partial (738 Hz) at 34 dB for both 3. F subjects. 4. A prominent energy region above F2 is present for high school subject 10, ranging from the seventh partial (1722 Hz) to the eighteenth partial (4428 Hz). Decibel readings range from 12 to 19. High school subject 4 shows an intensity peak on the ninth par- tial (2214 Hz) at 22 dB. Figures 79 through 81 are spectrographs showing the formant behavior of university subjects singing the vowel [0] at b-246 Hz, and display the following information: 1. The intensities of the fundamental range from 32 to 34 dB. 2. F] occurs on the second partial (492 Hz) at 40 dB for all three subjects. 3. F2 occurs on the third partial (738 Hz) for university subjects 3 and 9 at 36 and 34 dB, respectively. F2 occurs on the fourth partial (984 Hz) at 38 dB for university subject 4. 4. A prominent energy region above F2 is present in all three spectrograms. University subject 3 shows an energy region ranging from the ninth partial (2214 Hz) to the sixteenth partial (3936 Hz) at 16 to 32 dB. University subject 4 shows an energy region ranging from the tenth partial (2460 Hz) to the twenty-first partial (5166 Hz) at 8 to 21 dB. University subject 9 shows an energy region similar to that of university subject 3, but at 10 to 19 dB. Freq. 8 8 8 8 8 8 8 8 8 813 145 - n ION dhliubmaum033333333303u385m Part. dad-... NON‘OQ 010501 8399 9L99 0369 9919 3199 8999 9069 0919 3819 §a§§;;:‘8"38e m" sasfiz§sss “9 Figure 79.--Spectrograph of university subject 3 singing the vowel [0] at b-246 Hz. 146 d dTUOMONQDO db 40 38 36 34 32 so 28 26 24 22 20 18 16 14 12 10 8 6 4 2 0 fl ...-1an am 0’ be géeiesesse§§§§§§§§§ss§§§§ Figure 80.--Spectrograph of university subject 4 singing the vowel [0] at b-246 Hz. 147 ll zz ez rt sz 4‘ 0 3 ddd-fldd-fi—Idd” dnrwammucomo-nwwbmaaumcoo MbNJCD-I-D-h-bNNNEQ wgaaaamgmmm 388E833“33“m6§8 asesa-§§* oangbognm ognmbomn 8 Figure 81.--Spectrograph of university subject 9 singing the vowel [0] at b-246 Hz. 148 Figures 82 through 85 are spectrographs showing the formant behavior of mature subjects singing the vowel [0] at b-246 Hz, and display the following information: 1. The intensities of the fundamental range from 26 to 28 dB. 2. F1 occurs on the second partial (492 Hz) for all four subjects. Decibel readings are 38, 38, 34, and 40, respectively, for mature subjects 3. 4, 9, and lO. 3. F2 occurs on the third partial (738 Hz) for mature subjects 3, 4, and lo. Decibel readings are 34, 32, and 36, respectively. F2 occurs on the fourth partial (984 Hz) at 38 dB for mature subject 9. 2 is present in all 4. A prominent energy region above F four spectrograms. This energy region occurs from the ninth partial (2214 Hz) to the fifteenth partial (3690 Hz) with decibel readings ranging from l4 to 34. A Comparison of All Subjects Singing the Vowel Io] at b-246 Hz 1. The intensities of the fundamental range from 26 to 34 dB. 1 occurs on the second partial (492 Hz) for all subjects 2. F with decibel readings ranging from 34 to 40. 3. F2 occurs on the third partial (738 Hz) for all subjects 2 occurs on the except university subject 4 and mature subject 9. F fourth partial (984 Hz) for these two subjects at 38 dB. Subjects for whom F2 occurs on the third partial show decibel readings from 32 to 36. 4. A prominent energy region above F2 occurs for all sub- jects. This energy region is inconsistent, ranging from the 149 ONDO Freq. sagasssgsssasgaassasesggs 8 m n a 8 8 n m o 8 n m a 8 8 n c Figure 82.--Spectrograph of mature subject 3 singing the vowel [0] at b-246 Hz. 150 db 8 S 2 8 833 28 26 24 2O 18 18 14 12 ONbO’Q ea§§szsaaxsae gazes GIN gflMg-FQSMCD UQNN ON 50 NO omnooa 0367 9919 2195 8999 9068 O§l9 Freq. Figure 83.--Spectrograph of mature subject 4 singing the vowel [0] at b-246 Hz. Q. 0' o n a a m 8 3 3 a a 3 8 § 8 8 8 8 g 8 815 Freq. 151 ID Nbuthbu —- d arc-.5 3~Bgsasazesss.assaasasm s Figure 84.--Spectrograph of mature subject 9 singing the vowel [0] at b-246 Hz. 152 «Fatima:ummoznagaaaaagggabg Part. db40 38 36 34 32 30 28 26 24 22 20 18 16 14 12 10 8 6 4 2 0 Mbuo-n-b-o-nnuvo c.) w a and: mum Hm 3888§§§§§§§§§§§§§§s§§§§§§ Figure 85.--Spectrograph of mature subject l0 singing the vowel [0] at b-246 Hz. 153 seventh partial (1722 Hz) to the twenty-first partial (5166 Hz) at 8 to 34 dB. Vowel [u] at Pitch b-246 Hz Figures 86 through 88 are spectrographs showing the formant behavior of high school subjects singing the vowel [u] at b-246 Hz, and display the following information: l. The intensities of the fundamental range from 32 to 35 dB. 2. F1 occurs on the second partial (492 Hz) for all three subjects. Decibel readings are 4D, 38, and 38, respectively, for high school subjects 5, 9, and lO. 3. F2 occurs on the third partial (738 Hz) for high school subjects 9 and lo. Decibel readings are 33 and 34, respectively. F2 occurs on the fourth partial (984 Hz) at 34 dB for high school subject 5. 4. An energy region above F2 occurs in all three spectrograms. This energy region is very weak and insignificant for high school subjects 9 and lo. The energy region ranges from the tenth partial (2460 Hz) to the eighteenth partial (4428 Hz) at 8 to 23 dB for high school subject 5. Figures 89 and 90 are spectrographs showing the formant behavior of university subjects singing the vowel [u] at b-246 Hz, and display the following information: l. The intensities of the fundamentals are 36 and 38 dB. 2. F1 occurs on the second partial (492 Hz) for both sub- jects. Decibel readings are 34 and 38, respectively, for university subjects 4 and 9. ' 18 Freq. 8 B E 3 8 8 8 8 8 823 154 91 L1 81 61 02 12 22 82 72 92 Part. .a ddddd ‘1?)wbmmNO‘DO-‘NU501 16 14 12 01050 972 267 88L 786 0821 9L71 2221 8961 7122 0972 90L2 2962 8618 7778 0696 9668 2817 8277 7L97 0267 9919 2179 8999 7069 0919 Figure 86.--Spectrograph of high school subject 5 singing the vowel [u] at b-246 Hz. 8 8 8 8 B 8 8 8 8 823 dud-ed “0&1‘0” 01030! Freq. 155 N dddddddddd N N N amwammwaoCanuammqmmg-sauam Part. '0 b m 8 N 8‘ ”:3 E8 9L71 2221 Figure 87 singing the vowel .--Spectrograph of high school subject 9 [u] at b-246 Hz. Q. U’ Freq. 156 d d dub-Ad... dNmemflmmO-‘NU-bmm & O o n a m m 3 3 Z 8 a 8 8 8 8 8 8 8 8 8 8 Nh‘l‘D-‘d-‘HNNN U SS8§~38883“88§§8 Figure 88.--Spectrograph of singing the vowel [u] at b-246 Hz. 12 at c: vz sz '0 0 3 dds-IN ummo Ataamgmgm .. aim- at .. d S 8 E 8 8 u 8 8 high school subject l0 db 40 8 8 8 8 8 8 8 24 20 18 16 14 10 ONJSODQ Freq. 157 Nbfl‘O—ndd-‘NNN O) (a) b #301 010) 8S88§38§3§§88§888§88‘$888 Figure 89.--Spectrograph of university subject 4 singing the vowel [u] at b-246 Hz. 158 21 81 v: s: 9: L1 81 61 oz 12 at ca rt 9: 3’ 3 ...—5 de-bmmflme-D db 4o 38 36 34 32 30 28 26 24 22 20 ' 18 16 14 12 10 8 6 4 2 O “0-....- MN 5 b0! 01 no géaigast§§§§§§§§§§§§§§§Eé Figure 90.-—Spectrograph of university subject 9 singing the vowel [u] at b-246 Hz. 159 3. F2 occurs on the third partial (738 Hz) for both sub- jects. Decibel readings are 26 and 29, respectively. 4. A very weak energy region occurs in both spectrograms. This energy region is very insignificant, showing a decibel reading of l2 at the highest peak. Figures 9l through 94 are spectrographs showing the formant behavior of mature subjects singing the vowel [u] at b-246 Hz, and display the following information: l. The intensities of the fundamental range from 24 to 35 dB. 2. F1 occurs on the second partial (492 Hz) for all four subjects. Decibel readings are 36, 36, 36, and 40, respectively, for mature subjects 3, 4, 9, and l0. 3. F2 occurs on the third partial (738 Hz) for mature sub- jects 3, 4, and lo. Decibel readings are 30, 32, and 30, respectively. F2 occurs on the fourth partial (984 Hz) at 40 dB for mature subject 9. 4. A prominent energy region above F2 is present in all four spectrograms. Each subject shows a high intensity peak covering three different partials: the eleventh partial (2706 Hz), the twelfth partial (2952 Hz), and the thirteenth partial (3l98 Hz). Decibel readings for the high intensity peaks range from 22 to 28. A Comparison of All Subjects Singing_ the Vowel [ulgat b-246 Hz l. The intensities of the fundamental range from 24 to 38 dB. 2. F1 occurs on the second partial (492 Hz) for all subjects with decibel readings ranging from 34 to 40. db Freq. 160 N d 22 82 72 d-I-L—b-D-I-b-l-b-IIN anuammumoo—Mwbmaummo 40 0105010 972 267 882 786 0821 9L71 22L1 8961 7122 0972 90L2 2962 8618 7778 0698 9868 2817 8277 7L97 0267 9919 2179 8999 7069 Figure 9l.-—Spectrograph of mature subject 3 singing the vowel [u] at b-246 Hz. 0919 161 14 Qua-moo Mad“) b10801 USN b Freq. 7122 0972 90L2 2962 8618 7778 0698 9868 2817 8277 7L97 0821 9L7l 22L1 8961 0267 9919 2179 8999 7069 Figure 92.-~Spectrograph of mature subject 4 singing the vowel [u] at b-246 Hz. 0919 Part. 1152 8277 7L97 0267 2817 9919 2179 8999 7069 0919 E88 1" i5 so: too 88L was 0821 92v: zzc: 8961 7122 oer: 902a zsoz 8618 Figure 93.4-Spectrograph of mature subject 9 singing the vowel [u] at b-246 Hz. Freq. 163 8 6 4 2 0 asge~sswat~ma§8a Figure 94.--Spectrograph of singing the vowel [u] at b-246 Hz. h 5&01010'10101 atmwaomwa mnungamom noobo nacho mature subject l0 164 2 3. F occurs on the third partial (738 Hz) for all subjects except high school subject 5 and mature subject 9. F2 occurs on the fourth partial (984 Hz) for these two subjects at 34 and 40 dB, respectively. Decibel readings for the third partial range from 26 to 34. 2 in all spectrograms. 4. An energy region is present above F This energy region is very weak and insignificant for both university subjects and two of the high school subjects. The energy region for high school subject 5 ranges from the tenth partial (2460 Hz) to the eighteenth partial (4428 Hz) with a decibel reading of 23 at the highest intensity. Each mature subject shows a prominent energy region with an intensity peak covering three different partials: the eleventh partial (2706 Hz), the twelfth partial (2952 Hz), and the thirteenth partial (3l98 Hz). Decibel readings for the high intensity peaks ranged from 22 to 28. Vowel [a] at Pitch g#-4l5 Hz Figures 95 and 96 are spectrographs showing the formant 1 behavior of high school subjects singing the vowel [a] at g#-415 Hz, and display the following information: l. The intensities of the fundamentals are 28 and 34 dB. 1 occurs on the second partial (830 Hz) at 38 dB for 2. F both subjects. 3. F2 occurs on the third partial (1245 Hz) at 34 dB for both subjects. 4. A prominent energy region is present in both spectro- grams. This energy region ranges from the sixth partial (2490 Hz) 165 d a ...-5de ‘wammfim‘oo-‘wama 8 8 8 8 8 8 8 8 8 8 843 9721 0991 9L02 0672 9062 0288 98L8 0917 9997 0867 9689 0189 9229 0799 Figure 95.--Spectrograph of singing the vowel [a] at g#-4l5 Hz. no «are :3 £3 33 23 =3 81 29.5 (n Part quuoooomw d Ongfl-fig O GIN d N 01°01 0'1 U" a“ high school subject 9 “N O. O" b O 8 8 8 8 8 8 8 8 8 8 8 singing 166 d wammummoanuammuoom d ddNN wwbbbmmm uqvmooco N cagwu-omcogoowgoag \l—- b qwowwmmg 610 01“ d8 01 momomom 01001001001 0" Figure 96.--Spectrograph of high school the vowel [a] at g#-4l5 Hz. n: :3 =3 01 Part 9796 0966 91.801 subject l0 167 to the twelfth partial (4980 Hz), showing a high intensity peak on the eighth partial (3320 Hz) at 3l dB for both subjects. Figures 97 and 98 are spectrographs showing the formant behavior of university subjects singing the vowel [a] at g#-4l5 Hz, and display the following information: 1. The intensities of the fundamentals are 26 and 34 dB. 2. F1 occurs on the second partial (830 Hz) for both sub- jects. Decibel readings are 40 and 38, respectively, for university subjects 3 and 9. 3. F2 occurs on the third partial (l245 Hz) at 36 dB for university subject 9. F2 appears to be combined with F1 for univer- sity subject 3. 4. A prominent energy region above F2 is present in both spectrograms. This energy region ranges from approximately the fifth partial (2075 Hz) to the twelfth partial (4980 Hz) at 14 to 24 dB for university subject 9, and l2 to 28 dB for university subject 3. Figures 99 through l0l are spectrographs showing the formant behavior of mature subjects singing the vowel [a] at g#-415 Hz, and display the following information: l. The intensities of the fundamental range from 26 to 36 dB. 2. F1 occurs on the second partial (830 Hz) for all three subjects. Decibel readings are 38, 36, and 40, respectively, for mature subjects 4, 9, and l0. 2 occurs on the third partial (1245 Hz) for mature sub- jects 4 and 9. Decibel readings are 32 and 36, respectively. F2 appears 3. F to be combined with F1 for mature subject l0. 8 8 8 8 8 8 8 8 8 8 8E3 ...—add ONOU’ Omem 088 a Freq. a! singing 168 ddMNBO-Jwbbbmmm 888‘o888888888 Figure 97.--Spectrograph of the vowel [a] at g#-415 Hz. uquoooocotogg-a aag~2~a a university subject 3 db 8 8 823 singing 169 d a d—b-b—b-b de&mm\lm‘DO-‘Nw5mm 5888888388§888 bmqwouwmmg dNb 01001001001001 (110010 Figure 98.--Spectrograph of the vowel [a] at g#-4l5 Hz. N O 12 ...—o-b Nmm qummw 8d azgwid§ 8 university subject 9 8 8 8 8 8 8 8 8 8 8 813 singing 170 d d debmmNQQO-l 21 81 71 91 91 #61010 035$ 0'1 01 8888888§8§888§ 01 01001001 01 01001 Figure 99.--Spectr graph of the vowel [a] at g -415 Hz. uuuoooococogga aagwr~2 8 010018018018; mature subject 4 8 8 8 8 8 8 6 4 2 o b on Freq. 51 8 singing 171 9721 0991 9202 0672 9062 0288 9828 0917 9997 0867 9689 0189 9229 0799 9902 0272 9882 0088 Figure l00.--Spectrograph of mature the vowel [a] at g#-4l5 Hz. 9128 0816 9796 0966 92801 subject 9 Freq. 172 8 6 4 2 0 bad-OM10 wwaaammmmuuummom -b asggacg§a§§§§§§§§§§§a§§§§ Figure l0l.--Spectrograph of mature subject l0 singing the vowel [a] at g#-4l5 Hz. 173 4. A prominent region of energy above F2 is present in all three spectrograms. This energy region ranges from the fifth partial (2075 Hz) to the eleventh partial (4565 Hz), with high intensity ranging from 30 to 36 dB. A Comparison of All Subjgcts Singing the Vowel_[a]_at g -415 Hz l. The intensities of the fundamental range from 26 to 36 dB. 2. F1 occurs on the second partial (830 Hz) for all subjects with decibel readings ranging from 36 to 40. 3. F2 occurs on the third partial (1245 Hz) for all subjects 2 appears to be 2 except university subject 3 and mature subject 10. F combined with F1 for these subjects. Decibel readings for F on the third partial range from 32 to 36. 4. A prominent energy region above F2 is present in all spectrograms. This energy region is centered around the eighth partial (3320 Hz) with intensity peaks ranging from 24 to 36 dB. Vowel [e] at Pitch g#-415 Hz Figures 102 and 103 are spectrographs showing the formant behavior of high school subjects singing the vowel [e] at g#-415 Hz, and display the following information: 1. The intensities of the fundamentals are 32 and 36 dB. 2. F‘ is absorbed by the fundamental for both subjects. 3. F2 occurs on the fourth partial (1660 Hz) for both sub- jects. Decibel readings are 38 and 40, respectively, for high school subjects 3 and 10. 8 8 8 8 8 8 8 8 8 813 singing d .b dMQbMU’NQ‘DO-P d-‘MN 800055 8988 sadg Figure 102.--Spect the vowel [e] at g 1 21 0867 5 74 9689 0189 9229 0799 9902 0272 9882 ograph of high ~415 Hz. ”@1019 d §3-g§o .. u mam a school subject 3 175 1 2 8 7 9 9 2 8 6 0 1 21 81 71 91 91 21 81 61 02 12 22 82 72 9 a? 2" db 8 8 823 8 8 8 24 22 20 18 16 14 12 10 010505” a: 8 singing Freq. ddMM 0055-5010101 NNVNQCDCD -I N 03 N-fiUICD NM 05 (JV-5 O A NQONUMOJ #10 0'1“ 9.. 0.) U1 01001001001 01001001001001 0" a Figure 103.--Spectrograph of high school subject 10 the vowel [e] at g#-415 Hz. l76 4. A prominent energy region above F2 is present in the spec- trogram of high school subject l0. This energy region ranges from the fifth partial (2075 Hz) to well beyond 10,000 Hz at l0 to 23 dB. High school subject 3 shows a very weak and insignificant energy region, never exceeding 10 dB. Figures l04 and 105 are spectrographs showing the formant behavior of university subjects singing the vowel [e] at g#-415 Hz, and display the following information: l. The intensities of the fundamentals are 32 and 38 dB. 2. F1 is absorbed by the fundamental for both subjects. 3. F2 occurs on the fourth partial (l660 Hz) for both sub- jects. Decibel readings are 34 and 38, respectively, for university subjects 3 and 9. 4. A prominent energy region above F2 is present in both spectrograms. This energy region ranges from the sixth partial (2490 Hz) to the twelfth partial (4980 Hz) at 22 to 30 dB. Figures 106 through l08 are spectrographs showing the formant behavior of mature subjects singing the vowel [e] at g#-4l5 Hz, and display the following information: l. The intensities of the fundamental range from 30 to 38 dB. 1 2. F is absorbed by the fundamental for all three subjects. 3. F2 occurs on the fourth partial (1660 Hz) at 36 dB for all three subjects. 2 is present in all three 4. A prominent energy region above F spectrograms. This energy region ranges from the sixth partial (2490 Hz) to the tenth partial (4150 Hz) at 22 to 34 dB. Freq. 177 .5 dddd—D—I-l-IIN dnuammumoOaMwammummo 22 18 16 14 12 1O OMOOSN 91V 088 9V21 0991 9L02 0672 9062 0289 SCLC 091V 999V 086? 9689 0189 9229 0999 990L OLVL 988L 0088 91L8 0816 Figure l04.-—Spectrograph of university singing the vowel [e] at g#-415 Hz. 9796 0966 9L€01 subject 3 d d unwammummod db 40 8 8 8 S 2 8 8 24 20 18 16 14 10 01'0me Freq. 91V 088 9921 0991 9L02 0602 9062 0288 98L8 091V 999? Figure 105.--Spect singing the vowel [e] at g 178 21 0867 F3 at O1 uqqcoooco d e§a§aaa~:*§$a graph of university subject 9 415 Hz. d «MuammummOA db40 38 36 34 32 30 28 26 24 22 20 18 16 14 12 10 8 6 4 2 0 readaxaasea “W mga§asaaaaa Figure 106.-—Spect singing the vowel [e] at g 179 ... N 086? 5‘3 81 V1 91 91 9689 0189 9229 0999 L1 990L 81 OLVL 9882 61 0088 91L8 0816 graph of mature 415 Hz. 12 22 82 9’96 72 subject SL801 on Part. l.'1l||'l lll db 4O 32 28 26 24 22 20 18 16 14 12 10 creamer: 088 Freq. singing 180 d d dNQbU‘G‘IQ‘DO‘ 21 81 71 91 91 L1 81 61 02 saaggassaggaemaxa" 501‘] 80080) ngmng 010010 01 01001001 01001 Figure lO7.--Spectrograph of mature the vowel [e] at g#-4l5 Hz. 91L8 0816 9V96 0966 92801 subject 4 14 Omem 911? 088 Freq. singing 181 d .5 ddd-I-l-b-A—LN debmmVQCO-DNUOWONQCDO N .... d-‘NN 005be mmwuuoooococo - aaaéaaaaa§§aa§aa§sae§ag Figure l08.--Spectrograph of mature subject 9 the vowel [e] at g#-415 Hz. 182 A Comparison of All Sub%ects Singing_ the Vowel [é]’at g -4lS Hz l. The intensities of the fundamental range from 30 to 38 dB. 2. F1 is absorbed by the fundamental for all subjects. 2 3. F occurs on the fourth partial (1660 Hz) for all subjects. Decibel readings range from 34 to 40. 4. High school subject 10 shows a region of energy above F2 covering over twenty partials at 10 to 23 dB. All university and mature subjects show a prominent energy region above F2. This energy region ranges from approximately the sixth partial (2490 Hz) to the twelfth partial (4980 Hz). Decibel readings at peak intensities range‘from 22 to 34. Vowel [i] at Pitchgf;415 Hz Figures l09 and llO are spectrographs showing the formant behavior of high school subjects singing the vowel [i] at g#-415 Hz, and display the following information: 1. The intensities of the fundamentals are 36 and 40 dB. l 2. F is absorbed by the fundamental for both subjects. 3. F2 occurs on the fourth partial (l660 Hz) at 23 dB for high school subject 9. F2 occurs on the fifth partial (2075 Hz) at 36 dB for high school subject l0. 4. A prominent energy region is present in the spectrogram of high school subject 10. This energy region shows two high intensity peaks occurring on the sixth partial (2490 Hz) and the ninth partial (3735 Hz), both at 32 dB. An energy region above F2 is not present in the spectrogram of high school subject 9. db 40 38 36 34 32 30 28 26 Freq. 183 d ... d-I—h-b-b-A—D-INNNMK,” P drowbmmammo-evowbmaivoocoodwuam art. For’ bm-b-DNN wwbaam OI \INNQQQDCD d d N O {AN-50110 M 05(3de 0 U1 5 \1 N0») 01 ‘10 01‘)” d (a) 01 U" 001 0" 01001001001 01 U1 a Figure lD9.--Spectrograph of high school subject 9 singing the vowel [i] at g#-415 Hz. 184 1 z c v s 9 L 8 6 or 1 21 81 v1 91 91 L1 81 61 oz 12 ca 85 vz s 3’ 3 A an d d N N 8 8 (a) A b 5 01 g O1 8 \J N N a: on CD CD 8 d Fre . d N g C b \J - 01 (D M O 5 00 g \I - C q “83 388338883383333 asfieg Figure llO.--Spectrograph of high school subject 10 singing the vowel [i] at g#-415 Hz. 185 Figures lll and 112 are spectrographs showing the formant behavior of university subjects singing the vowel [i] at g#-415 Hz, and display the following information: 1. The intensities of the fundamentals are 40 dB for both subjects. 2. F1 is absorbed by the fundamental for both subjects. 3. F2 occurs on the fourth partial (1660 Hz) at 28dB for both subjects. 4. A prominent energy region above F2 is present in the spectrograms of both subjects. University subject l shows high intensity peaks on the sixth partial (2490 Hz) and the tenth partial (4150 Hz) at 24 and 18 dB, respectively. University subject 3 shows a high intensity peak on the sixth partial at 26.5 dB. Figures 113 through 115 are spectrographs showing the formant behavior of mature subjects singing the vowel [i] at g#-415 Hz, and display the following information: 1. The intensities of the fundamental range from 32 to 39 dB. 1 2. F is absorbed by the fundamental for all three subjects. 3. F2 occurs on the fourth partial (1660 Hz) for all three subjects. Decibel readings are 38, 32, and 38, respectively, for mature subjects 3, 4, and 9. 4. A prominent energy region above F2 is present in all three spectrograms. This energy region ranges from the sixth partial (2490 Hz) to the eleventh partial (4565 Hz) with intensity peaks ranging from 22 to 32 dB. db 8 8 833 8 8 8 24 22 20 18 16 14 12 onumm 088 2 Freq. m singing 186 dd-l—I—D—b-b-LN Mwbmmummo-nnuammuoowo 338881 Part. assessesgagee§§zgmssasa 0" 01001001001 01001001001 0'1 01 a Figure lll.--Spectrograph of university subject 1 the vowel [i] at g#-415 Hz. Freq. 187 dd-I—b-D-D—b—L-I-DN dnwbmmumcoo-nnubmmuoomo FOFfl 12 22 82 92 92 '0 U a P’ creamer: 91L8 0816 9996 088 9721 0991 9L02 0672 9062 0288 98L8 091v 999v 0867 9689 0189 9229 0799 9902 OLVL 988L 0088 0966 SL801 Figure llZ.--Spectrograph of university subject 3 singing the vowel [i] at g#-415 Hz. ii! 188 dd dnwbmmummod domum wmbm mdwo mudm mwoo wage vomm mmbo mwnm mmdo wwwm ammo 5mmm admo wuwm wwuo Mmom Mbmo Moqm 2mmo .me mwo Adm Freq. Figure l13.--Spectrograph of mature subject 3 #-415 Hz. the vowel [i] at g singing 189 evadeeaassaameegisaasswea Freq mgbgumonwgmmgaw gumgaggmu momomom 01001001001061 0: 010; Figure ll4.--Spectrograph of mature subject 4 singing the vowel [i] at g#-415 Hz. 190 d ......d-s-a-I—LNN NNN unuammummadnwbmmummo-aauam Part. 8 6 4 2 0 d-DNN 01 \luu com d We 3§§§as3§§§$§s§§§as§§ae§§§ Figure llS.--Spectrograph of mature subject 9 singing the vowel [i] at g#-415 Hz. 191 A Comparison of All Subjects Singing the Vowel [j]¥atgg#-4l5 Hz l. The intensities of the fundamental range from 32 to 40 dB. 2. F1 is absorbed by the fundamental for all subjects. 3. F2 occurs on the fourth partial (1660 Hz) for all subjects except high school subject 10, with decibel readings ranging from 23 to 33. F2 occurs on the fifth partial (2075 Hz) at 36 dB for high school subject 10. 2 is present in all 4. A prominent region of energy above F spectrograms except that of high school subject 9. This energy region ranges from approximately the sixth partial (2490 Hz) to the eleventh partial (4565 Hz) at 18 to 32 dB. #-415 Hz Vowel [o]at Pitch 9 Figures 116 and 117 are spectrographs showing the formant behavior of high school subjects singing the vowel [0] at g#-415 Hz, and display the following information: 1. The intensity of the fundamental is 34 dB for both sub- jects. 2. F1 is absorbed by the fundamental for both subjects. 3. F2 occurs on the second partial (830 Hz) for both sub— jects. Decibel readings are 38 and 40, respectively, for high school subjects 9 and l0. 4. A prominent region of energy above F2 is present in both spectrograms. This energy region ranges from the sixth partial (2490 Hz) to the twenty-fifth partial (l0375 Hz) at ll to 26 dB. 192 d ‘wammflmwc -O N & \l-‘UICD N hm ”83888888888888 §aa Figure 116.--Spectrograph of high singing the vowel [0] at g#-415 Hz. ges§ga - u «so a school subject 9 193 d a d-I—D-D-Dd-fi—ON uzubmmumoo-onuammuoocoo db 40 38 36 34 32 30 28 26 24 22 20 18 16 14 12 10 owbmoo 9128 0816 088 9921 0991 9202 0692 9062 0288 9828 0919 9999 0869 9689 0189 9229 0999 9902 0292 9882 0088 Freq. 9996 0966 92801 Figure ll7.--Spectrograph of high school subject 10 singing the vowel [0] at g#—415 Hz. 194 Figures 118 and 119 are spectrographs showing the formant #-415 Hz, behavior of university subjects singing the vowel [0] at g and display the following information: 1. The intensity of the fundamental is 38 dB for both sub- jects. 2. F1 is absorbed by the fundamental for both subjects. 2 occurs on the second partial (830 Hz) for both sub- 3. F jects. Decibel readings are 40 and 38, respectively, for university subjects 3 and 9. 4. A prominent energy region is present in both spectrograms. High intensity peaks occur on the seventh partial (2905 Hz) and the ninth partial (3735 Hz) for both subjects, with decibel readings ranging from 24 to 33. Figures 120 through 122 are spectrographs showing the formant behavior of mature subjects singing the vowel [0] at g#-415 Hz, and display the following information: 1. The intensities of the fundamental range from 31 to 35 dB. 1 2 2. F 3. F is absorbed by the fundamental for all three subjects. occurs on the second partial (830 Hz) for all three subjects. Decibel readings are 34, 34, and 40, respectively, for mature subjects 3, 9, and 10. 4. A prominent energy region above F2 is present in all three spectrograms. This energy region shows intensity peaks between the fifth partial (2075 Hz) and the eighth partial (3320 Hz), with decibel readings ranging from 22 to 32. db 8 8 833 010301 Freq. 195 dd dawbmmumoo-swwbmmucoco 919 088 9921 0991 9202 0692 9062 0288 9828 0919 9999 Figure 118.--Spect singing the vowel [0] at g dddddd‘d kilo «are h) 4:02:50! Part. 02 0869 9689 0189 9229 0999 9902 0292 9882 0088 9128 0816 9996 0966 92801 gograph of university subject 3 -415 Hz. 0'0th 00 8 singing Freq. 196 aunngmwnobmgm \INNQmCD‘DCD-b NGO‘ mudgmg N oumwuagmo bQN‘DOng m #10 MNWSHQ 8“ 01001001001 01001001001001 01901 a: Figure ll9.-—Spectrograph of university subject 9 the vowel [0] at g#-415 Hz. 197 db 40 8 8 8 8 8 24 22 20 18 16 14 12 10 8 6 4 2 0 bm-t-hNN 005550101 uuuoo (D .. we asstae§§aeaee2§§agae§§§§§ Figure 120.--Spectrograph of mature subject 3 singing the vowel [0] at g#-415 Hz. 198 db 4O 32 28 26 24 22 16 14 12 10 8 6 4 2 0 bmaduugwwbAbmgmmuuummmmsa F awumoa tau-40108 ”2°58 u-nm o req. moamqwauwcfimm an my gdgbgw momo omomowomomom m 01 6’1 . Figure lZl.--Spectrograph of mature subject 9 singing the vowel [0] at g#-415 Hz. 199 c-I-l-D-Id-O—I-DN dnwammqmmodwubmmummo 14 Ome9 Freq. 9 9 088 9921 0991 9202 0692 9062 0288 9828 0919 9999 0869 9689 0189 9229 0999 9902 0292 882 0088 9128 0816 9996 0966 92801 Figure 122.--Spectrograph of mature subject l0 singing the vowei [0] at g#-415 Hz. 200 A Comparison of All Sub'ects Singing the Vowel [o] at g -415 Hz l. The intensities of the fundamental range from 31 to 38 dB. 2.' F1 is absorbed by the fundamental for all subjects. 3. F2 occurs on the second partial (830 Hz) for all subjects. Decibel readings range from 34 to 40. 2 is present in all 4. A prominent energy region above F spectrograms. A region of high intensity peaks ranging from the fifth partial (2075 Hz) to the ninth partial (3735 Hz) is registered by the university and mature subjects, with decibel readings ranging from 22 to 33. The high school subjects show an energy region from the sixth partial (2490 Hz) to the twenty-fifth partial (10375 Hz) at 11 to 26 dB. Vowel [u] at Pitch]g#-415 Hz Figures 123 and 124 are spectrographs showing the formant behavior of high school subjects singing the vowel [u] at g#-415 Hz, and display the following information: 1. The intensities of the fundamentals are 38 and 40 dB. 2. F1 is absorbed by the fundamental for both subjects. 2 occurs on the second partial (830 Hz) for both sub- 3. F jects. High school subject 9 shows a very low second formant of 22 dB, whiie high school subject 10 has a decibel reading of 38. 2 is present in the spec- 4. A prominent energy region above F trogram of high school subject 10. liufiiintensity peaks occur on the sixth partial (2490 Hz) and the eighth partial (3320 Hz) at 29 and 26 dB, respectively. In addition, a region of energy from the 201 db 40 14 ouamm b F req. 5'1 m \1 d U1 8’ 9996 0966 088 9921 0991 9202 0692 9062 0288 9828 0919 9999 0869 9689 0189 9229 0999 9902 0292 9882 0088 92801 Figure 123.--Spectrograph of high school subject 9 singing the vowel [u] at g#-415 Hz. 14 0105010 Freq. 202 h) 3888c» Part. bm-b-IMPO wwbbamgm NNNW HUN80b (ON-50108 N 045009 0105 ‘10 Mummoo HM 01‘108 (”00001000000001001001 Figure 124.--Spectrograph of high singing the vowel [u] at g#-415 Hz. 9128 0816 9996 0966 92801 school subject 10 203 fifteenth partial (6225 Hz) to beyond the twenty-fifth partial, with a maximum intensity of 27 dB, occurs in the spectrogram of high school 2 is not evident in the spectro- subject 10. An energy region above F gram of high school scubject 9. Figures 125 and 126 are spectrographs showing the formant behavior of university subjects singing the vowel [u] at g#-415 Hz, and display the following information: 1. The intensity of the fundamental is 40 dB for both sub- jects. 2. F1 is absorbed by the fundamental for both subjects. 2 occurs on the second partial (830 Hz) for both sub- 3. F jects: Decibel readings are 40 and 36, respectively, for university subjects 3 and 9. 4. A prominent region of energy is present in the spectrogram of university subject 9. The energy region ranges from the fifth partial (2075 Hz) to the tweifth partial (4980 Hz) with a maximum decibel reading of 30. University subject 3 shows a very weak and insignificant region of energy above F2. Figures 127 and 128 are spectrographs showing the formant behavior of mature subjects singing the vowel [u] at g#-415 Hz, and display the foTlowing information: 1. The intensities of the fundamentals are 39 and 40 dB. 1 2 2. F 3. F is absorbed by the fundamental for both subjects. occurs on the second partial (830 Hz) for both sub- jects. Decibel readings are 30 and 34, respectively, for mature subjects 4 and 10. db Freq. 204 FOF.‘ wammumoé‘fififii‘aa: 40 38 36 34 32 30 28 26 24 22 20 18 16 14 12 10 8 6 4 2 0 «50¢de (055301 a: \1 isaeaei§eei§e§a§i Figure 125.--Spectrograph of singing the vowe1 [u] at g#-415 Hz. §§3§9§8§ assessed university subject 3 Freq. 08.15010 35" singing 9921 Figure 126.--Spectrograph of 0991 9202 0692 9062 0288 9828 205 Part. Abbmgm “##0000 d #0108 N comedy-n? 0 U100 #10 01‘108-‘8 0 00100100100100! 0'1 0'1 3 university subject 9 the vowel [u] at g#-415 Hz. Ffiri’ db 40 38 36 34 32 3O 28 26 24 22 20 18 16 14 12 0105010 088 3 Freq. m singing 206 dad—Addddddn wammumoo-onuamaiuoocpo ddNN to 55501 \INN b§fl00~080 d” 01‘] 01 01001001 0'1 01001001001 Figure 127.--Spectrograph of mature the vowel [u] at g#-415 Hz. 12 22 2 2 9128 0816 9996 0966 subject 92801 db F 8 8 813 singing 207 dd—bd—t-n—L—A—A-IM gfinwbmmumoodvoubmmuoomo 9921 0991 9202 0692 9062 0288 9828 0919 9999 0869 9689 0189 9229 0999 9902 0292 9882 0088 Figure 128.--Spectrograph of mature the vowe1 [u] at g#-415 Hz. NNNNN ‘10me 01010 -I 38§§§ subject 10 Part. 208 4. A prominent energy region above F2 is present in both spectrograms. Both subjects show intensity peaks on the sixth partial (2490 Hz) and the seventh partial (2905 Hz). Decibel readings range from 22 to 34. A Comparison of All Subjects Singing the Vowel [u] at g?-415 Hz l. The intensities of the fundamental range from 38 to 40 dB. 2. F1 is absorbed by the fundamental for all subjects. 3. F2 occurs on the second partial (830 Hz) for a11 subjects. Decibel readings range from 30 to 40 with the exception of high school subject 9, who shows a low second formant of 22 dB. 2 is present in the 4. A prominent region of energy above F spectrograms of most subjects, although mature subjects show the only similar patterns in this energy region. Both mature subjects show intensity peaks on the sixth partial (2490 Hz) and the seventh par- tial (2905 Hz) with decibel readings ranging from 22 to 34. The data presented in this chapter reveal information con- cerning formant behavior. The fol1owing chapter will outline the conclusions concerning the formant behavior of the subjects studied in this investigation, relate the conclusions to that of previous research, and recommend suggestions for future research. CHAPTER V SUMMARY, CONCLUSIONS, AND RECOMMENDATIONS Summar The problem of the study was to discover whether there are similar patterns of formant behavior in the spectra of sung vowels judged to be excellent in quality in tenor voices. The study attempted to discover the formants present in a given vowe1, any similarity in patterns present, and compare the results with those of previous research. The study included the vowels [a], [e], [i], [o], and [u]. The voices studied were high school, university, and mature tenor voices selected by qualified choral directors and applied voice teachers at various high schools, universities, and cities in Michigan. Thirty singers (10 high school voices, 10 university voices, and 10 mature voices) sang each vowel on the pitches e-l65 Hz, b-246 Hz, and g#-415 Hz. The singers were recorded on the same equipment, at the same volume level, and the same distance from the microphone to insure differences in vocal quality which might have resulted from uneven signal strengths. The recording of each subject took place in an environment similar to that of a voice studio. The sample used for spectrographic analysis was selected by 20 adjudicators from the Michigan School Vocal Association's list of $010 adjudicators. 209 210 For selecting the sample of voices to be analyzed, the taped samples were prepared in random order on separate tapes for each of the three pitches and for each of the age groups--a total of nine tapes. Aural instructions on the tape kept the adjudicator informed. . as to the sample, pitch, and vowel performed. Each adjudicator listened to only three tapes at one time to avoid any possible fatigue factor. The adjudication took place on three consecutive days for each adjudicator. Only those vowel samples receiving a simple majority of excel- lent ratings (i.e., 11 or more), with no fair or poor ratings, were used for spectrographic analysis. The number of samples selected as excellent within a given age level varied according to vowel sound and pitch. Selected samples were prepared into tape loops and converted into spectrographs at the Audiology and Speech Sciences Laboratory at Michigan State University. The spectrographs were analyzed and reproduced graphically, indicating the frequencies, intensities, and locations of the fundamental and first two formants for each subject selected. Hoyt‘s Reliability Coefficient was calculated at the Computer Center at Michigan State University for each vowel selected at each of the three pitches to determine adjudicator reliability. The resulting print-out shows reliability scores (r) and standard error of measurement scores (SE). 211 Conclusions Adjudicator evaluations were found to be highly consistent. Results showed reliability coefficients ranging from .9698 to .9819 for all vowels analyzed. » The relationships between the fundamental and F1 were as follows: 1. For the vowel [a], F1 had a frequency of 660 Hz on pitch e-l65 Hz, 738 Hz on pitch b-246 Hz, and 830 Hz on pitch g#-4l5 Hz. F1 occurred on partial frequencies for all subjects analyzed. The relative intensity of F1 increased as the fundamental frequency increased. 2. For the vowel [e], F1 had a frequency of 495 Hz on pitch e-l65 Hz, and 492 Hz on pitch b-246 Hz. F1 was absorbed by the fundamental on pitch g#-415 Hz. F1 occurred on partial frequencies for all subjects analyzed. The relative intensity of F1 remained approximately the same for a11 three pitches. 3. For the vowel [i], F1 had a frequency of 330 Hz on pitch e-l65 Hz, and was absorbed by the fundamental on pitches b-246 Hz and g#-4l5 Hz. The relative intensity of F1 remained approximately the same for all three pitches. F1 occurred on partial frequencies for all subjects analyzed. 4. For the vowel [o], F1 had a frequency of 495 Hz on pitch 1 e-165 Hz and 492 Hz for pitch -246 Hz. F was absorbed by the funda- 1 occurred on partial frequencies for all 1 mental on pitch g#-415 Hz. F subjects analyzed. The relative intensity of F remained approxi- mately the same for all three pitches. 212 5. For the vowel [u], F1 had a frequency of 330 Hz for pitch 1 e-l65 Hz and 492 Hz for pitch b-246 Hz. F was absorbed by the funda- #-415 Hz. F1 occurred on partial frequencies for mental on pitch 9 all subjects analyzed. The relative intensity ofF1 remained ,. approximately the same for all three pitches. The relationships between the fundamental and F2 were as fol- lows: 1. For the vowel [a], F2 had a frequency of 990 Hz on pitch e-165 Hz for all subjects except high school subject 3 and mature subject 3. F2 occurred at 1155 Hz for these two subjects. F2 had a frequency of 984 Hz on pitch b-246 Hz for all subjects except mature 2 subject 3, who showed an F frequency of 1230 Hz. F2 had a frequency of 1245 Hz on pitch g#-415 Hz for all subjects except university subject 3 and mature subject 10. F2 appeared to be combined with F1 2 occurred on partial frequencies for 2 (830 Hz) for these subjects. F all subjects analyzed. The relative intensity of F remained the same for all three pitches. 2. For the vowel [e], F2 had a frequency of 1320 Hz for mature subject 4, 1485 Hz for mature subjects 3 and 10, 1650 Hz for high school subject 10 and both university subjects, and 1815 Hz for high school subjects 3 and 4 and mature subject 1 on pitch e—l65 Hz. F2 had a frequency of 1475 Hz on pitch b-246 Hz and 1660 Hz on pitch g#-415 Hz. F2 occurred on partial frequencies for all subjects analyzed. The relative intensity of F2 remained approximately the same for all three pitches. 213 3. For the vowel [i], F2 had a frequency of 1815 Hz on pitch e-165 Hz for all subjects except high school subjects 3 and 4. F2 occurred at 2145 Hz for these two subjects. F2 had a frequency of 2 1722 Hz for all subjects on pitch b-246 Hz. F had a frequency of 1660 Hz on pitch g#-415 Hz for all subjects except high school subject 10, who showed an F2 frequency of 2075 Hz. F2 occurred on partial frequencies for all subjects analyzed. The relative intensity of F2 increased slightly as the fundamental frequency increased. 4. For the vowel [o], F2 had a frequency of 660 Hz for all subjects except high school subjects 4 and 10, and university subject 4 on pitch e-165 Hz. F2 occurred at 825 Hz for these three subjects. 2 F had a frequency of 738 Hz for all subjects except university 2 subject 4 and mature subject 9 on pitch b-246 Hz. F occurred at 984 Hz for these two subjects. F2 had a frequency of 830 Hz on pitch g#-415 Hz for all subjects. F2 occurred on partial frequencies for all subjects analyzed. The relative intensity of F2 increased slightly as the fundamental frequency increased. 5. For the vowel [u], F2 appeared to be combined with F1 (330 hz) for all subjects except high school subject 3 and mature subject 10 on pitch e-l65 Hz. F2 occurred at 495 Hz for these two subjects. 2 F had a frequency of 738 Hz on pitch b-246 Hz for all subjects except high school subject 5 and mature subject 9. F2 occurred at 984 Hz for these two subjects. F2 # had a frequency of 830 Hz for all -415 Hz. F2 occurred on partial frquencies for 2 subjects on pitch 9 all subjects analyzed. The relative intensity of F remained approxi- mately the same for all three pitches. 214 1 2 The relationships between F and F were as follows: 1. For the vowel [a], the relative intensity of F.l increased 2 slightly as the fundamental frequency increased, while F remained constant for all three pitches. 2. For the vowel [e], F1 was absorbed by the fundamental as 2 the pitch increased, while the relative intensity of F remained the same for all three pitches. 3. For the vowel [i], F1 was absorbed by the fundamental and 2 increased slightly as the fundamental frequency the intensity of F increased. 4. For the vowel [o], F1 was absorbed by the fundamental 2 increased slightly as the fundamental fre- and the intensity of F quency increased. 5. For the vowel [u], F1 was absorbed by the fundamental as the fundamental frequency increased, while the intensity of F2 remained constant for all three pitches. Formants occurred on partial frequencies for all subjects. This was true, even though formants occasionally occurred on different partials for a given vowe1 or pitch. F1 was consistent in location for all subjects on all vowels -and pitches. Not one variation occurred in the study. F2 was not as consistent in location, especially on the lower pitches. The location of F2 differed for two subjects singing the vowel [a] at e-165 Hz. F2 occurred on four different frequencies for the vowel [e] at e-165 Hz. The location of F2 differed for two subjects singing the vowel [i], three subjects singing the vowel [o], 215 and one subject singing the vowel [u] at e-l65 Hz. For pitch b—246 Hz, the location of F2 differed for one subject singing the vowel [a], two subjects singing the vowel [o], and two subjects singing the vowel 2 [u]. There was total consistency in location of F for the vowels #-415 Hz, the location of F2 dif- [e] and [i] at b-246 Hz. For pitch g fered for two subjects singing the vowel [a] and one subject singing the vowel [i]. All subjects were consistent in the location of F2 for the vowels [e], [o], and [u] at g#-415 Hz. Thus, only 17, out of 120 vowels analyzed, differed in formant location. There was little indication that the age of the singer created basic differences in formant behavior. Formant intensities were relatively similar for all subjects. There was a noticeable variation in the intensities of the fundamentals, especially between high school and mature subjects. A prominent region of energy above F2 was present in most spectrograms. This energy region was not consistent among subjects and failed to be as prominent as either F1 or F2. Results Related to Previous Research The present investigation found that formants always occurred on partial frequencies. This disagrees with Appelman53 but corroborates 55 Sullivan56 found the formants to the findings of Jones54 and Wash. be in the same general location, but not exactly related to partials in all cases. 53 54 Jones, op. cit. Sullivan, op cit. . w Appelman, op. cit. 55Hash, op. cit. 56 216 The formant locations and frequencies in the present investi- gation agree with those found by Hash for the vowels [a], [i], and [u]. Hash found the intensity of F2 remained constant at lower pitches and increased as the pitch increased. The present findings disagree in that (l) for the vael [a], the intensity of F2 remained constant for all three pitches, and (2) for the vowel [u] the intensity of F2 F2 remained constant for all three pitches. did increase slightly in intensity as the fundamental frequency increased for the vowel [i]. The fixed formants for vowels, without regard to fundamental pitch, given by Appelman do not agree with the locations found in the present investigation. For example, Appelman states that for the vowel [a], F1 is 700 Hz and F2 is 1200 Hz. While some subjects showed F1 and F2 frequencies in this general range, the majority of subjects registered F1 and F2 frequencies that varied over 100 Hz or more with the frequencies of Appelman. Recommendations for Future Research The present study dealt with three different age levels, five vowels, and three pitches. Due to the magnitude of Chapter IV, it is recommended that future research be limited to one age level at a time. Previous studies dealt with professional or mature subjects. Based on the results of this study, it is recommended that more research be initiated dealing with younger voices in all voice classi- fications. The present study dealt with vowels judged to be excellent. It is recommended that carefully controlled research be initiated 217 comparing the formant behavior of vowels judged to be poor with vowels judged to be excellent. As far as is known, this is the first study of its kind deal- ing with the tenor voices. It is recommended that research be con- tinued with the tenor voice and expanded to include all voice classifications, vowels, and pitches. Due to the opinions of previous researchers and the lack of consistency of the patterns registered, the present investigation did not deal with the third formant. Future research in this area will be difficult until more sophisticated measuring devices are invented. Part of the present problem is the fact that the Bruel and Kjaer 2107 frequency analyzer scans each of the six spectrum scanning stages in the same amount of time. The analyzer is very accurate in the first two stages (20 to 63 Hz, and 63 to 200 Hz). However, the fourth stage (630 to 2000 Hz) and fifth stage (2000 to 6300 Hz) are not measured as accurately since a wider range of frequencies is measured in the same amount of time taken to measure stage one. This accounts for the lack of clarity in energy regions above F2. The present investigation dealt with the solo voice. Choral educators are constantly working for vowel unity and achieve success in this area by utilizing vocal exercises in unison. The design of this investigation could easily be applied to a unison choral sound for a thorough analysis of formant behavior. Music educators might benefit from exploring and utilizing the scientific equipment and research available to them. APPENDICES 218 APPENDIX A VOICE EXPERIMENT 219 APPENDIX A VOICE EXPERIMENT A. Adjudicator Information Name: Title: Place of Employment: B. Preliminary Information 9 minutes pencil, test paper, earphones Earphones - 3 Speakers - 7 Time of Test: Materials Needed: Listening Level on Volume Dial: Tape Speed: 7-1/2 ips. Put the earphones on and start the tape for complete instructions. C. Instructions: TEST Most Natural Sound Degree of Difference Sample 1: 1 (A) A great deal of difference 1 (B) A moderate difference Equal Very little difference Sample 2: 2 (A) A great deal of difference 2 (B) A moderate difference Equal Very little difference Sample 3: 3 (A) A great deal of difference 3 (B) A moderate difference Equal Very little difference Sample 4: 4 (A) A great deal of difference 4 (B) A moderate difference Equal Very little difference (REMOVE EARPHONES) 220 Sample 5: 5 (A) 5 (B) Equal 6 (A) 6 (B) Equal Sample 7: 7 (A) 7 (B) Equal Sample 6: Sample 8: 8 (A) 8 (B) Equal Please express your preference for listening either earphones or speakers. 221 Ill lll III III Check below: Earphones Speakers A great deal of difference A moderate difference Very little difference A great deal of difference A moderate difference Very little difference A great deal of difference A moderate difference Very little difference A great deal of difference A moderate difference Very little difference to the tape through APPENDIX B LETTER TO DIRECTORS 222 APPENDIX B LETTER TO DIRECTORS Dear I am asking your assistance in gathering data for my doctoral dissertation concerning the study of quality in tenor voices. This is an acoustical study entitled, "The Formant Behavior of Vowels in Selected Tenor Voices." There is presently little scientific research available to choral and applied voice teachers that gives them a concrete basis on which to judge vowel quality. It is the purpose of this study to dis- cover formant patterns present in high school, university, and mature tenor voices that may lead to positive judgments regarding quality in a voices. I am requesting that you recommend any high school, college, or professional tenors that you consider to be exceptionally excellent voices. The data to be collected could be gathered in the course of one day (or evening) on your campus. The data to be collected consists of a selected voice singing the five basic vowels on three different pitches, covering lower voice, middle voice, and upper voice. This would require approximately 20 minutes of each individual selected for the study. I would appreciate hearing from you in the very near future. Please indicate your reactions to this study, your willingness to cooperate, names or number of students willing and qualified to partici- pate, and a suggested date and time that would be convenient for me to come to your campus to collect the data. It would be especially helpful if you would send me the names and addresses of any professional tenors throughout the state that you consider truly excellent performers. I anticipate some problems in lining up an adequate sample in this category. Your assistance in this matter is greatly appreciated. I look forward to your reply in the near future. Sincerely, Paul w. Schultz 223 APPENDIX C ADJUDICATOR'S INSTRUCTIONS 224 APPENDIX C ADJUDICATOR'S INSTRUCTIONS As you listen to the tapes you will hear tenor voices singing the five basic vowels at three different pitch levels. Each vowel will last for six seconds and will be repeated after a pause of three sec- onds. There will be ten different voices performing each vowel at each of the three pitches. .After you hear a given vowel and its repetition by an individual voice, you will indicate your judgment of the quality of that voice according to the following rating scale: (l) Excellent, (2) Good, (3) Fair, (4) Poor. Please make a prompt judgment, place it in the apprOpriate square on the answer sheet, and do not change your judgment once you have recorded it on the answer sheet. Please disregard all clicks or noises on the tape that are not a part of the vowel production. Make your judgement on the steady state portion of the vowel. Do not consider thg attack or release of that vowel. Any vowel produced at the pitch 9 -4l5 Hz that is, in your Opinion, sung falsetto will be rated as Poor. Aural instructions will guide you on the tape as to vowel sound, pitch, and sample letters. High school, university, and mature tenor voices are used in this study. Evaluate each category based upon your experience and musical values. Use your own standards for each level but, under no circumstances, will you compare the three levels. Instruction will be given as to the operation of the tape recorder and changing of tapes. Once you begin a tape, do not stop it under any circumstances. You will complete three tapes at this sitting, and the remainder on two separate days. Vowels are indicated on the answer sheet according to the Inter- national Phonetic Alphabet as follows: [a] as in calm [e] as in pay [i] as in beet [o] as in tone [u] as in boot 225 APPENDIX D ADJUDICATOR MASTER SCORE CHARTS 226 227 ADJUDICATOR MASTER SCORE CHART e-165 HZ [a] Vowel: Pitch: High School Subjects Adjudicators lO 10 ll 12 l3 14 15 16 l7 l8 T9 20 228 ADJUDICATOR MASTER SCORE CHART e-165 Hz [91 Vowel: Pitch: High School Subjects Adjudicators 1O 1O 11 12 13 14 15 16 17 18 19 20 10 [i] e-l65 Hz 229 High School Subjects Vowel: Pitch: ADJUDICATOR MASTER SCORE CHART Adjudicators l0 ll l2 l3 14 15 l6 l7 18 l9 03 20 230 ADJUDICATOR MASTER SCORE CHART [0] e-165 Hz Vowel: Pitch: High School Subjects Adjudicators 10 1O 11 12 13 14 15 16 17 18 19 20 231 ADJUDICATOR MASTER SCORE CHART [U] e-165 Hz Vowel: Pitch: High School Subjects Adjudicators 10 10 11 12 13 14 15 16 17 18 19 20 232 ADJUDICATOR MASTER SCORE CHART [a] e-l65 Hz University Subjects Vowel: Pitch: Adjudicators 1O 1O 11 12 13 14 15 16 17 18 19 20 233 ADJUDICATOR MASTER SCORE CHART e-165 Hz [e] Vowel: Pitch: University Subjects Adjudicators 10 1O 11 12 13 14 15 16 17 18 19 20 234 ADJUDICATOR MASTER SCORE CHART [i] e-165 Hz University Subjects Vowel: Pitch: Adjudicators 10 1O 11 12 13 14 15 16 17 18 19 2O 235 ADJUDICATOR MASTER SCORE CHART e-l65 Hz [0] Vowel: Pitch: University Subjects Adjudicators 1O 1O 11 12 13 14 15 16 17 18 19 20 236 ADJUDICATOR MASTER SCORE CHART [U] e-l65 Hz University Subjects Vowel: Pitch: Adjudicators 10 10 11 12 13 14 15 16 17 18 19 20 237 ADJUDICATOR MASTER SCORE CHART e-l65 Hz Mature Subjects [a] Vowel: Pitch: Adjudicators 10 1O 11 12 13 14 15 16 17 18 19 20 238 ADJUDICATOR MASTER SCORE CHART [e] e-165 Hz Vowel: Pitch: Mature Subjects Adjudicators 10 1O 11 12 13 14 15 16 17 18 19 20 239 ADJUDICATOR MASTER SCORE CHART e-165 Hz [1'] Vowel: Pitch: Mature Subjects Adjudicators 10 1O 11 12 13 14 15 16 17 18 19 20 240 ADJUDICATOR MASTER SCORE CHART [0] e-l65 Hz Vowel: Pitch Mature Subjects Adjudicators 10 1O 11 12 13 14 15 16 17 18 19 20 241 ADJUDICATOR MASTER SCORE CHART e-165 Hz Mature Subjects [U] Vowel: Pitch: Adjudicators 1O 1O 11 12 13 14 15 16 17 18 19 20 242 ADJUDICATOR MASTER SCORE CHART [a] b-246 Hz Vowel: Pitch: High School Subjects Adjudicators 1O 1O 11 12 13 14 15 16 17 18 19 20 243 ADJUDICATOR MASTER SCORE CHART 46 Hz e Vowel: Pitch: High School Subjects Adjudicators 10 10 11 12 13 14 15 16 17 18 19 20 244 ADJUDICATOR MASTER SCORE CHART [i] b-246 Hz Vowel: Pitch: High School Subjects Adjudicators 1O 2. 10 11 12 13 14 15 16 17 18 19 20 245 ADJUDICATOR MASTER SCORE CHART [0] b-246 Hz Vowel: Pitch: High School Subjects Adjudicators 10 1O 11 12 13 14 15 16 17 18 19 20 246 ADJUDICATOR MASTER SCORE CHART [U] b-246 Hz Vowel: Pitch: High School Subjects Adjudicators 10 1O 11 12 13 14 15 16 17 18 19 20 247 ADJUDICATOR MASTER SCORE CHART 46 Hz .142 [.D Vowel: Pitch University Subjects Adjudicators 1O 1O 11 12 13 14 15 16 17 18 19 2O 248 ADJUDICATOR MASTER SCORE CHART [e] b-246 Hz Vowel: Pitch: University Subjects Adjudicators 10 1O 11 12 13 14 15 16 17 18 19 20 249 ADJUDICATOR MASTER SCORE CHART b-246 Hz University Subjects [1'] Vowel: Pitch: Adjudicators 10 1O 11 12 13 14 15 16 17 18 19 20 250 ADJUDICATOR MASTER SCORE CHART b-246 Hz University Subjects [0] Vowel: Pitch: Adjudicators 10 10 11 12 13 14 15 16 17 l8 19 20 251 ADJUDICATOR MASTER SCORE CHART 46 Hz [b Vowel: Pitch University Subjects Adjudicators 10 1O 11 12 13 14 15 16 17 18 19 20 252 ADJUDICATOR MASTER SCORE CHART [a] b-246 Hz Vowel: Pitch: Mature Subjects Adjudicators 1O 10 11 12 13 14 15 16 17 18 19 20 253 ADJUDICATOR MASTER SCORE CHART b-246 Hz Mature Subjects [6] Vowel: Pitch: Adjudicators 1O 10 11 12 13 14 15 16 17 18 19 20 254 ADJUDICATOR MASTER SCORE CHART [i] b-246 Hz Vowel: Pitch: Mature Subjects Adjudicators 10 10 11 12 13 14 15 16 17 18 19 20 255 ADJUDICATOR MASTER SCORE CHART [0] b-246 Hz Vowel: Pitch: Mature Subjects Adjudicators 10 1O 11 12 13 14 15 16 17 18 19 20 256 ADJUDICATOR MASTER SCORE CHART b-246 Hz Mature Subjects [111 Vowel: Pitch: Adjudicators 1O 10 11 12 13 14 15 16 17 18 19 20 257 ADJUDICATOR MASTER SCORE CHART [a] g#-415 Hz Vowel: Pitch: High School Subjects Adjudicators 10 10 11 12 13 14 15 16 17 18 19 2O 258 ADJUDICATOR MASTER SCORE CHART Vowel: [5] Pitch: g -415 Hz High School Subjects Adjudicators 10 1O 11 12 13 14 15 16 17 18 19 2O 259 ADJUDICATOR MASTER SCORE CHART [)1 Pitch: g -415 Hz Vowel: High School Subjects Adjudicators 10 1O 11 12 13 14 15 16 17 18 19 2O 260 ADJUDICATOR MASTER SCORE CHART Vowel: [3] Pitch: g -4l5 Hz High School Subjects Adjudicators 1O 10 11 12 13 14 15 16 17 18 19 20 261 ADJUDICATOR MASTER SCORE CHART Vowel: [y] Pitch: g -4l5 Hz High School Subjects Adjudicators 1O 1O 11 12 13 14 15 16 17 18 19 20 262 ADJUDICATOR MASTER SCORE CHART Vowel: [a] Pitch: g -415 Hz University Subjects Adjudicators 1O 1O 11 12 13 14 15 16 17 18 19 20 263 ADJUDICATOR MASTER SCORE CHART [e] g#-415 Hz Vowel: Pitch: University Subjects Adjudicators 1O 1O 11 12 13 14 15 16 17 18 19 20 264 ADJUDICATOR MASTER SCORE CHART [i] g#-415 Hz Vowel: Pitch: University Subjects Adjudicators 10 1O 11 12 13 14 15 16 17 18 19 20 265 ADJUDICATOR MASTER SCORE CHART Vowel: [3] Pitch: g -415 Hz — ---.-.-——‘-- University Subjects Adjudicators 1O 1O 11 12 13 14 15 16 17 18 19 20 266 ADJUDICATOR MASTER SCORE CHART Vowel: [y] Pitch: g -4l5 Hz University Subjects Adjudicators 10 1O 11 12 13 14 15 16 17 18 19 20 10 Mature Subjects 267 [a] g#-415 Hz Pitch: ADJUDICATOR MASTER SCORE CHART Vowel: 1O 12 13 11 Adjudicators 2 14 15 16 17 18 19 20 268 ADJUDICATOR MASTER SCORE CHART Vowel: [g] Pitch: g -415 Hz Mature Subjects Adjudicators 1O 10 11 12 13 14 15 16 17 18 19 20 269 ADJUDICATOR MASTER SCORE CHART [i] g#-415 Hz Mature Subjects Vowel: Pitch: Adjudicators 1O 10 11 12 13 14 15 16 17 18 19 2O 270 ADJUDICATOR MASTER SCORE CHART g#-415 Hz [0] Vowel: Pitch: Mature Subjects Adjudicators 10 1O 11 12 13 14 15 16 17 18 19 20 I! {...l 1‘! lll [II-III." III. [III III 271 ADJUDICATOR MASTER SCORE CHART g#-415 Hz Mature Subjects [U] Vowel: Pitch: Adjudicators 10 10 11 12 13 14 15 16 17 18 19 20 BIBLIOGRAPHY 272 .‘l‘IIIIVIII-ll I| 1| l‘ll-[llllli BIBLIOGRAPHY Aikin, William A. The Voice. New York: Longmans, Green and Company, 1951. Appelman, D. Ralph. The Science of Vocal Pedagogy. Bloomington: Indiana University Press, 1967. Backus, John. The Acoustical Foundations of Music. New York: W. W. Norton and Company, Inc., 1969. Bartholomew, Wilmer T. Acoustics of Music. New York: Prentice-Hall, Inc., 1942. Borchers, Orville J. "The Relation Between Intensity and Harmonic Structure in Voice," Psychological Record, III (April, 1939), 59-67. Castle, William E. The Effect of Selective Narrow-Band Filtering_on the Perception of Certain English Vowels. The Hague: Mouton and Company, 1964. Delattre, Pierre. "Vowel Color and Voice Quality: An Acoustic Articu- latory Comparison," National Association of Teachers of Singing Bulletin (October, 1958), 4-7. Erickson, Carl I. "The Basic Factors in the Human VoiceJ'Psychological Monographs, XXXVI (February, 1927), 88-92. Fletcher, Harvey. Speech and Hearing_in Communication. New York: Van Nostrand Company, Inc., 1953. Gray, Giles Wilkeson, and Wise, Claude Merton. The Bases of Speech. 3rd ed. New York: Harper and Brothers, 1959. Helmholtz, Hermann L. F. Sensations of Tone. 4th ed. New York: Dover Publications, 1954. Holmes, F. Lincoln. "An Experimental Study of Individual Vocal Quality," Quarterly Journal of Speech, XVI (October, 1930), 351. Jones, J. Loren. "A Cinefluorographic and Spectrographic Analysis of the Effect of Velum Positions on Sung Vowels." D.M.Ed. disserta- tion, Indiana University, 1971. 273 274 Paget, Sir Richard A. Human Speech. New York: Harcourt, Brace and Company, 1930. Potter, Ralph K.; Kopp, George A.; and KOpp, Harriet Green. Visible Speech. New York: Dover Publications, Inc., 1966. Scripture, Edward Wheeler. "Analysis and Interpretation of Vowel Tracks," Journal of the Acoustical Society of America, V (February, 1933), 148-152. Sullivan, Ernest G. "An Experimental Study of the Relationships between Physical Characteristics and Subjective Evaluation of Male Voice Quality in Singing." Ph.D. dissertation, Indiana University, 1956. Turabian, Kate L. A Manual for Writers of Term Papers, Theses, and Dissertations. 3rd ed., rev. London: The University of Chicago Press, 1967. Vennard, William. Singing, the Mechanism and the Technic. 3rd ed., rev. Ann Arbor: Edwards Brothers, Inc., 1964. Wash, Nathaniel Hubert. "The Formant Behavior of the Vowels [a], [i], and [u] in Baritone Voices in Relation to Different Voice Ranges." Ph.D. dissertation, Michigan State University, 1971. . Winckel, Fritz. Music, Sound and Sensation. New York: Dover Publi- cations, Inc., 1967. II. 4 I {I {14‘