Citation
Correlation of glottal area function within three registers as revealed through measurement of ultra-high-speed photographs and photoelectric glottographs, by R. Joyce Redus Harden

Material Information

Title:
Correlation of glottal area function within three registers as revealed through measurement of ultra-high-speed photographs and photoelectric glottographs, by R. Joyce Redus Harden
Creator:
Harden, Ruby Joyce Redus, 1929-
Publication Date:
Language:
English
Physical Description:
xiii, 104 leaves. : illus. ; 28 cm.

Subjects

Subjects / Keywords:
Area function ( jstor )
Falsetto ( jstor )
Glottal consonants ( jstor )
Phonation ( jstor )
Photographic film ( jstor )
Signals ( jstor )
Trucks ( jstor )
Vocal cords ( jstor )
Vocal registers ( jstor )
Vowels ( jstor )
Dissertations, Academic -- Speech -- UF ( lcsh )
Speech thesis Ph. D ( lcsh )
Genre:
bibliography ( marcgt )
non-fiction ( marcgt )

Notes

Thesis:
Thesis--University of Florida.
Bibliography:
Bibliography: leaves 95-102.
General Note:
Typescript.
General Note:
Vita.

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
This item is presumed in the public domain according to the terms of the Retrospective Dissertation Scanning (RDS) policy, which may be viewed at http://ufdc.ufl.edu/AA00007596/00001. The University of Florida George A. Smathers Libraries respect the intellectual property rights of others and do not claim any copyright interest in this item. Users of this work have responsibility for determining copyright status prior to reusing, publishing or reproducing this item for purposes other than what is allowed by fair use or other copyright exemptions. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder. The Smathers Libraries would like to learn more about this item and invite individuals or organizations to contact the RDS coordinator (ufdissertations@uflib.ufl.edu) with any additional information they can provide.
Resource Identifier:
022918882 ( ALEPH )
14181281 ( OCLC )

Downloads

This item has the following downloads:


Full Text















Correlation of Glottal Area Function Within Three
Registers as Revealed Through Measurement of
Ultra-High-Speed Photographs and
Photo-Electric Glottographs














By

R. JOYCE REDUS HARDEN


A DISSERTATION PRESENTED TO THE GPADUATE COUNCIL OF
THE UNIVERSITY OF FLORIDA IN PARTIAL
FULFILLMENT OF THE REQUIREINTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY




UNIVERSITY OF FLORIDA 1972



























































Copyright by
R. Joyce Redus Harden
1972
















ACKNOWLEDGMENTS


The author wishes to express deep appreciation to her committee chairman, Dr. G. Paul Moore, for his suggestions and contributions to the ideas expressed herein. In addition, she expresses sincere gratitude to her other committee members, Drs. Edward C. Hutchinson, Norman N. Markel, and Madelaine M. Ramey, for their continuous encouragement and valuable guidance during the preparation and execution of the experiment and also throughout the author's training program at the University of Florida.

The author wishes to acknowledge the invaluable assistance of Messrs. David A. Campbell, Robert P. Idzikowski, and Russell E. Pierce in preparing the instrumentation for the study.

Further, she expresses appreciation to Miss Elizabeth L. Allen, Mrs. Mallory W. Iles, Mr. Robert L. Iles, and Dr. Howard B. Rothman for their assistance in evaluating the production of falsetto register by the subjects participating in the experiment.


iii









In conclusion, a word of gratitude and appreciation is expressed to an understanding and cooperative family whose tolerance and encouragement as well as participation as willing subjects made the completion of this research possible.

This research was supported in part by the National Institutes of Health grant NB-06459.

















TABLE OF CONTENTS


ACKNOWLEDGMENTS .... ............

LIST OF TABLES ...... ............

LIST OF FIGURES .... ............

ABSTRACT ... .......... ......

CHAPTER

I INTRODUCTION AND STATEMENT OF


II

III


IV

V

APPENDIX


PROBLEM ..... ..........

PROCEDURES .... .........

RESULTS .... ...........

DISCUSSION. ..........

SUMMARY AND CONCLUSIONS . . .


Page



.... vii


. .. . viii . . . . xi


THE



. . . . . . 15

. . . . .. . 27

. . . . .. . 58

. . . . .. . 65


A PHOTOGRAPHIC SEQUENCE ... ..........

B FORM USED IN THE JUDGMENT OF SUBJECT
PRODUCTION OF FALSETTO REGISTER .......

C PROCEDURE FOR MEASURING GLOTTAL AREA
FUNCTION ..... .................

D AMPLITUDE AND AREA CURVES DERIVED THROUGH
GLOTTOGRAPHIC AND PHOTOGRAPHIC FRA4E-BYFRAME ANALYSIS ..... ..............


70



* 72


74 78









TABLE OF CONTENTS (continued)


APPENDIX

E


FUNDAMENTAL FREQUENCIES . . ........


F RELATIVE INTENSITY ..... ............

BIBLIOGRAPHY ......... .....................

BIOGRAPHICAL SKETCH ....... ............ .....


Page


91 93 95

103
















LIST OF TABLES


Table Page

1 Summary table of correlation between
photo-electric glottographic and photographic methods of analysis
for modal register ... ............. ... 40

2 Summary table of averaged indications of
open quotient and speed quotient for each
subject in modal and falsetto register
phonation derived through analysis of photographic film and photo-electric
glottograms ...... ............... . 49

3 Summary table of correlation between
photo-electric glottographic and photographic methods of analysis
for falsetto register .. ........... . 50

4 Summary table of correlation between photo-electric glottographic and photographic methods of analysis
for vocal fry register ............. ... 56

5 Fundamental frequencies produced by five
subjects used in the present investigation. The measures are reported in Hz
for the sustained vowel /i/ produced in
vocal fry, modal, and falsetto registers. . 92.

6 Relative intensity levels of phonation for the five subjects used in the present
investigation. The measures are reported in dB for the vocal fry, modal,
and falsetto registers ............. ... 94


vii
















LIST OF FIGURES


Figure Paqe

1 Block diagram of equipment used to obtain
the voice signals and simultaneous photographic and glottographic records of
glottal area functions for each subject . 16

2 Three-dimensional drawing of subject
training station ... .......... ....23

3 Plot of glottal area function and photoelectric glottograph for one normalized
cycle. Subject 1, modal register ..... . 30

4 Plot of glottal area function and photoelectric glottograph for one normalized
cycle. Subject 3, modal register ..... . 32

5 Plot of glottal area function and photoelectric glottograph for one normalized
cycle. Subject 2, modal register ..... . 35

6 Plot of glottal area function and photoelectric glottograph for one normalized
cycle. Subject 4, modal register ..... . 36

7 Plot of glottal area function and photoelectric glottograph for one normalized
cycle. Subject 5, modal register .....39

8 Plot of glottal area function and photoelectric glottograph for one normalized
cycle. Subject 1, falsetto register. . . . 41

9 Plot of glottal area function and photoelectric glottograph for one normalized
cycle. Subject 2, falsetto register. . . . 42


viii









LIST OF FIGURES (continued)


Figure Page

10 Plot of glottal area function and photoelectric glottograph for one normalized
cycle. Subject 3, falsetto register. . . . 43

11 Plot of glottal area function and photoelectric glottograph for one normalized
cycle. Subject 4, falsetto register. . .. 44

12 Plot of glottal area function and photoelectric glottograph for one normalized
cycle. Subject 5, falsetto register. . . . 45

13 Plot of glottal area function and photoelectric glottograph for one normalized
cycle. Subject 3, vocal fry register . . . 51

14 Plot of glottal area function and photoelectric glottograph for one normalized
cycle. Subject 4, vocal fry register . .. 52

15 Diagram illustrating the measurement procedures used in data reduction for the
analysis of glottal area function .....77

16 Comparison of photo-glottographic and photographic frame-by-frame analysis for the vowel /i/ produced in modal
register for Subject 1 .......... ....79

17 Comparison of photo-glottographic and photographic frame-by-frame analysis for the vowel /i/ produced in modal
register for Subject 2 ... .......... ... 80

18 Comparison of photo-glottographic and photographic frame-by-frame analysis for the vowel /i/ produced in modal
register for Subject 3 ............. ... 81

19 Comparison of photo-glottographic and photographic frame-by-frame analysis for the vowel /i/ produced in modal
register for Subject 4 ........... ...82









LIST OF FIGURES (continued)


Figure Page

20 Comparison of photo-glottographic and
photographic frame-by-frame analysis for the vowel /i/ produced in modal
register for Subject 5 ............. ... 83

21 Comparison of photo-glottographic and
photographic frame-by-frame analysis
for the vowel /i/ produced in falsetto
register for Subject 1 ............. ... 84

22 Comparison of photo-glottographic and photographic frame-by-frame analysis
for the vowel /i/ produced in falsetto
register for Subject 2 ............. ... 85

23 Comparison of photo-glottographic and photographic frame-by-frame analysis
for the vowel /i/ produced in falsetto
register for Subject 3 ............. ... 86

24 Comparison of photo-glottographic and photographic frame-by-frame analysis
for the vowel /i/ produced in falsetto
register for Subject 4 ............. ... 87

25 Comparison of photo-glottographic and photographic frame-by-frame analysis
for the vowel /i/ produced in falsetto
register for Subject 5 ............. ... 88

26 Comparison of photo-glottographic and photographic frame-by-frame analysis
for the vowel /i/ produced in vocal fry
register for Subject 3 ............. ... 89

27 Comparison of photo-glottographic and photographic frame-by-frame analysis for the vowel /i/ produced in vocal
fry register for Subject 4 ...... ....90









Abstract of Dissertation Presented to the Graduate Council
of the University of Florida in Partial Fulfillment
of the Requirements for the Degree of Doctor of Philosophy

CORRELATION OF GLOTTAL AREA FUNCTION WITHIN THREE
REGISTERS AS REVEALED THROUGH MEASUREMENT OF
ULTRA-HIGH-SPEED PHOTOGRAPHS AND
PHOTO-ELECTRIC GLOTTOGRAPHS

By

R. Joyce Redus Harden

June, 1972

Chairman: Dr. G. Paul Moore Major Department: Speech

Based on the hypothesis that measurement of the quantity of light passing through the glottal area should reveal the same information derived from measurement of the discernible space between the vocal folds, the major purpose of this investigation was to determine the validity of the photo-electric glottograph as a measuring device as compared with a frame-by-frame analysis of ultra-high-speed photographic film. A second purpose was to describe and compare the characteristic glottal configurations in falsetto, modal, and vocal fry registration of pre-adolescent, adolescent, post-adolescent, and adult subjects as revealed by the two methods of investigation. To achieve these purposes, simultaneous recordings of the visual image of the vocal folds, the voice signal and the deviations sensed by the phototransducer during falsetto, vocal fry, and modal xi








phonation were obtained from five subjects. These subjects had met the following criteria: (1) they showed no evidence or history of voice disorders or laryngeal pathology; (2) they were capable of complete anterior-posterior exposure of the vocal folds; (3) they were capable of phonating in vocal fry, modal, and falsetto registers; and (4) they presented five types of laryngeal development (pre-adolescent, adolescent, post-adolescent, adult male and adult female).

Recordings were made during sustained phonation of the vowel /i/ for each of the registers investigated. Intraoral voice recordings made during photography were subjected to a fundamental frequency and intensity analysis.

Measurements of glottal area function were made on six

consecutive cycles from the steady-state portion of each film. Glottal area was determined in the laryngeal photographs by standard polar planimeter methods. The amplitude and frequency deviations produced by the oscillographic trace of the phototransducer were measured from the trace representing the 5k Hz signal. Correlation between the normalized curves produced by the two methods was determined and confidence intervals established.

The results indicated that the photo-electric glottograph represents glottal area function as revealed in highspeed photography. In addition, the use of the glottographic xii









method of investigation in conjunction with photographic procedures may lead to more complete understanding of the phase relationship between the upper and lower margins of the folds. The vibratory patterns exhibited by two subjects during the production of falsetto register suggest the participation of myoelastic factors.


xiii
















CHAPTER I


INTRODUCTION AND STATEMENT OF THE PROBLEM


Introduction

The nature of vocal fold vibration has in the past been the subject of wide differences of opinion (Musehold, 1897; West, 1926; Metzger, 1928; Tarnoczy, 1951). Presently, it is accepted that the vocal folds vibrate synchronously at a frequency that determines the fundamental pitch of the tone being produced, alternately allowing momentary outflow and stoppage of the air stream which is forced up from the lungs below by the muscles of exhalation (Moore, 1937b; van den Berg, 1958, 1960). The quality of the sound produced by the speaker is dependent on the manner in which the vocal folds vibrate.

Researchers have employed various methods for observation of the activity of the laryngeal mechanism over the past 186 years. Moore (1937a) presented an historical accounting of methods used in vocal fold research. Black (1951), Luchsinger and Arnold (1965), and Peterson (1958) have outlined the procedures used in current investigations.

1









With the advent of the laryngeal mirror, stroboscopic techniques, radiographic procedures, spectroscopic analysis, normal- and ultra-high-speed photography, and glottography, much is now known of the normal vibratory patterns in the different frequency ranges of the human voice.

Ultra-high-speed photography has afforded a reliable method for study of the complex vibratory movements of the vocal folds in the normal larynx (Farnsworth, 1940; Moore et al., 1962). Moore (1937b) reported that the opening of the glottis begins anterior to the mid-point of the folds and progresses in a posterior direction. This information contradicted conclusions derived earlier from stroboscopic photography on the nature of vibratory closure. Brackett's (1947) analysis of high-speed photographs of the vibrations of the vocal folds of two male subjects revealed differences in the proportion of time occupied by each of three phases of the vibratory cycle, closed, opening, and closing. Not only were these differences noted between subjects but also for different performances by the same subject. These results were confirmed by Smith (1954). The expected time ratio of three phases was postulated by van den Berg (1968a) who stated that the glottis should remain closed for twothirds of each period for low tones; whereas the closing time should decrease with rising pitch. This conclusion led








to a definition of the opening quotient of the glottis as representing the time occupied by each opening in relation to the total time of each period. Thus, low tones show a small opening quotient and high tones show an increasing opening quotient up to 1.0 when the glottis remains open throughout the cycle for falsetto tones. Information on vibratory patterns associated with different types of phonation is now available (Arnold, 1957; Moore, 1968a; van den Berg, 1958, 1960; Vennard, 1960). The phase relationship of the upper and lower margins of the vocal folds has aided in the physiological description of register phenomena (Miller, 1959).

Although the meaning of the term "voice registers"

appears to have no completely acceptable definition (Morner et al., 1963), there appear to be distinctive acoustic, physiological, and perceptual attributes characteristic for each of three ranges found in the normal production of voice (van den Berg, 1968b). Hollien and Michel (1968) included in their definition of "register" these criteria: (1) each shall be composed of a series of consecutive fundamental frequencies of a similar quality and (2) there should be little or no overlap of frequencies between registers. It was also found that subjective agreement between judges could be made for register change. More recently, Hollien (1971) has proposed three new terms in the definition of








three major vocal registers. Synonymous with vocal fry, pulse register occupies the lowest range of frequencies within the individual phonation range. Modal register designates that range of frequencies correlated with normal, chest and head, or low, mid, and high. This register is composed of those fundamental frequencies that are normally used in speaking and singing. Loft register is that register ordinarily designated as falsetto. This register includes the higher fundamental frequencies within an individual vocal range. For the purposes of this investigation, it seems appropriate to rely upon two of the more traditional terms,consequently, the registers will be designated as vocal fry, modal, and falsetto.

The study of vocal registration has been directed not only to the phonation of the adult but also to that of the child in an effort to define and compare various register productions appropriate for the vocal mechanism in changing stages of maturation. Investigators have established criteria for three vocal registers in the adult for both sexes (Luchsinger and Arnold, 1965; Pronovost, 1942; Snedicor, 1951; Colton, 1969; Hollien and Moore, 1960; Hollien and Michel, 1968; Hollien et al., 1971). Sedlackova (1961) found that the phonation ranges of children demonstrated the presence of three registers. Her findings were obtained









by means of stroboscopic motion pictures. The appearance of the vocal folds was found to be like that of the adult mechanism within the production of chest, mid, and head register. Luchsinger and Arnold (1965) and van Oordt and Drost (1963) have also defined the frequency range and register production of the voice in children. They found that falsetto and modal registers may be qualitatively differentiated.

Motion picture studies (Fletcher, 1958) and radiographic procedures (Russell, 1931) have resulted in understanding of the physiologic changes manifested by the vocal folds within three registers. Vocal fry, described as a succession of laryngeal vibrations of low frequency (Hollien et al., 1966), has been the subject of radiographic investigations. The vocal folds appear to be compact and shorter in length in production of this particular register (Hollien et al., 1969). The results obtained from motion picture studies indicate that the vibratory cycle may be monophasic or biphasic. Further, the initial excursion of the folds presenting a biphasic pattern appears to be 25 percent smaller in amplitude than does the second. Apparently the plateaus noted in the open phase are related to the longitudinal opening waves which travel along the length of the vocal folds (Moore and von Leden, 1958; Timcke et al., 1959).








The modal register has been equated with the natural

range, and chest or mid register (van den Berg, 1960). Fink and Kirschner (1959) observed that the configuration of the vocal and ventricular folds should be smoothly curved and symmetrical in production of modal register. They also indicated that the medial surface of the folds should be convex during phonation. This observation was confirmed by Sovak et al. (1971). Laminagraphic studies of the vocal folds revealed that as the fundamental frequency of phonation is raised the vocal folds tend to.elevate progressively. The angle of tilt also appears to increase as pitch elevates except in production of frequencies within the falsetto register (Hollien and Curtis, 1962).

Normal speed photography reveals that in the modal

register there is systematic lengthening of the vocal folds as frequency of phonation elevates (Hollien and Moore, 1960; Hollien, 1962). In addition, this lengthening may proceed in a stair step fashion. Ultra-high-speed motion pictures (Timcke et al., 1958) revealed that the opening quotient and speed quotient are similar in head and chest register. Increases in intensity appear to effect a faster opening quotient whereas the open quotient is inversely proportional to the intensity of the sound. Pitch changes do not appear to effect a change in either speed quotients or open quotients.








Moore (1968a) has cautioned that the initial vibratory movements of the vocal folds are usually extremely small and may be limited to only a portion of the glottal margin. Further, a medial motion may commence in a lateral direction. Vibration of the folds may appear with an open glottis when the diameter is not exceedingly large (van den Berg et al., 1957).

Subjective perceptual agreement is possible in judging production of the falsetto range (Rubin and Hirt, 1960; Hollien and Michel, 1968). Physiologically, the margins of the vocal folds have been found to be rather thin and pointed with vibration confined to this area (van den Berg, 1958). Rubin and Hirt (1960) described three patterns of vibratory activity for falsetto production. These were classified as open chink, closed chink, and damping. Furthermore, they found that as the untrained singer progressed through the phonatory range to the falsetto, this production elicited turbulent, chaotic vocal fold activity. It was concluded that aerodynamic forces contributed more to production of the falsetto than did myoelastic properties. This conclusion was verified by Hollien et al. (1971).

Motion picture studies and radiographic techniques (Griesman, 1943) have been successful in clarifying some aspects of larnygeal function. Unfortunately, due to the








mechanics of these methods, physiological studies have been limited to the investigation of sustained vowel sounds or to non-speech activities. Further limitations are imposed not only by the cumbersome aspect of the equipment but also by the necessity for great subject cooperation. Camera techniques necessitate the use of subjects who are able to tolerate a laryngeal mirror in the pharynx and who are capable of full anterior-posterior exposure of the vocal folds.

Two relatively new techniques are being employed in the investigation of laryngeal activity during connected speech. The use of a fiber optics system for photographic iivestigation is still in its infancy. Investigators (Sawashima and Hirose, 1968; Sawashima et al., 1969) have devised a fiber optics cable that may be inserted through the nasal passages. A 16 mm camera is attached to the eyepiece of the image guide with illumination provided by a 150 watt incandescent lamp. Motion pictures with a frame rate of 24 per second are thus possible. Unfortunately, there is the distinct probability of subject discomfort. The fiber optics cable placed in the nasal passage presents a diameter of approximately 6 mm. In addition, the illumination provided by the light source is not sufficient for ultra-high-speed exposure.

During the last 55 years, considerable interest has been directed to a new method for recording the movements








of the larynx (Spencer, 1917). The method has been termed glottography or laryngography. Researchers have proposed that through use of this procedure, direct examination of vocal fold movement in connected speech without undue subject discomfort is possible.

Hartmann and Wullstein (1938) were the first investigators to make any extensive use of a photo-electric cell for sensing vibrations of the vocal folds. They associated the results of the glottogram with those of the phonogram using a double beam oscilloscope. The waves produced by the glottogram resembled a series of positive sinusoidal waves with the exception that the opening phase appeared to be more gradual than did the closing phase. More recently, improvements have been made on this photoelectric glottographic technique (Sonesson, 1959; Ohala, 1966, 1967; Coleman and Wendahl, 1968; Sawashima, 1969; Lisker et al., 1969; Lindqvist, 1969; Frokjaer-Jensen, 1970).

Fabre (1957) and Decroix and Dujardin (1958) used an electronic process based on the principle of a high frequency current modulated by a low frequency current. The movements of the larynx of the subject, placed in an impedance bridge, were thus recorded in the shape of electrical variations. The primary objection to this method of








analysis proved to be electrode placement and size. Fourcin and Abberton (1971) have modified the electrodes and their placement in their Laryngograph. The circuitry of this device is postulated to allow for self compensation for speaker impedance variation and is responsive only to fast changes resulting from vibration of the vocal folds in ordinary phonation rather than gross laryngeal displacement.

An ultrasonographic method for studying the vibratory movements of the vocal folds has been developed by Minifie et al. (1967). Hertz et al. (1970) found that the ultrasonic glottograms presented the same appearances as that obtained with the photo-glottographic method. The technique makes use of short ultrasound-pulses delivered to the surface of the neck. As echoes are received from the vocal folds, ultrasonograms are developed depicting the vibratory amplitude of the vocal folds. Although the gross appearance of movement has been found to be similar to that produced by the photo-glottographic method, the authors suggest that refinement of the equipment is still indicated before exactness of vibratory patterns can be expected.

Lindqvist (1970) used an inverse filtering technique to obtain glottograms of vocal fold activity. Volume velocity waveforms were used to derive information.








Calculations of the opening, closing, and closed phases were made.

Investigators have made comparison between results

derived from one type of glottograph to another type, and found that, while there is similarity of wave form configuration, there is no one-to-one relationship (Fant et al., 1967; K6ster and Smith, 1970; Frokjaer-Jensen, 1970). Apparently, some variation of results may be due to registration change as well as subject subcutanous fat deposit differences in the throatarea.

Phonograms and glottograms have been compared in an

attempt to relate glottal area function to results depicted by the data derived from these methods (Fabre and Frei, 1959; Van Michel, 1966; Dolansky and Tjernlund, 1967; Vallancien and Faulhaber, 1967; Lebrun and Hasquin-Deleval, 1971). Great differences in wave form have been noted. It has been concluded that some of the differences may result from the peculiar ability of the vocal folds to vibrate without producing voice.

Photography has proved to be a reliable and valid

means for obtaining information of glottal area function; therefore, it would appear reasonable to make comparison of wave forms derived by frame-by-frame analysis to those resulting from the use of glottographic techniques (Zemlin,








1959). Sonesson (1960) was one of the first investigators to make such a comparison. The comparison was made between glottograph wave forms and wave forms developed through measurement of the distance between the medial margins of the vocal folds. The results indicated that the open quotient developed through these methods was identical. He suggested that the phase difference between the lower and upper edges of the vocal folds could be represented by the closed interval of the glottogram. This would seem to indicate that amplitude of the curve should be evaluated. Unfortunately, the curves derived were not obtained through simultaneous recording of glottographic and photographic procedures.

Great similarity of wave forms produced by the two

techniques have been found by other investigators (Sawashima et al., 1968; Van Michel et al., 1970; Lindqvist and Lubker, 1970). Unfortunately, the method of measurement remains unknown.

Two investigations made by Wendahl and Coleman (1967, 1968) provoke some question of the validity of the information derived through glottographs. Their frame-by-frame analysis of glottal function derived by polar planimeter measurement produced information on the total area. These data were then compared to the curves simultaneously








developed with the glottograph. The dissimilarity of the wave forms places the validity of the glottographic technique in some jeopardy. The measurements of glottal width rather than glottal area may be one factor influencing the disparity of fit noted in their glottogram data. Another factor that should be considered is the degree of vocal fold exposure. The subjects who participated in the study may not have attained the same degree of glottal area exposure.


Statement of the Problem

Previous studies of glottal area function have made

use of methods that have either been established as capable of deriving valid physiological data or investigators have made assumptions of glottal activity based on data derived from equipment ordinarily utilized in acoustical investigations. To date, measurement procedures employed in the study of the vocal mechanism have not been specifically described and replicated in subsequent investigations. Further, adequate identification of the frequency, intensity, and register of the phonational sample has not been consistent.

Based on the hypothesis that measurement of the

quantity of light passing through the glottal area reveals








the same information derived from measurement of the discernible space between the vocal folds, the major purpose of this study was to determine whether the method of photoelectric glottography is as valid for depicting glottal area function as a frame-by-frame analysis of ultra-highspeed film. A second purpose was to describe and compare the characteristic glottal configurations in falsetto, modal, and vocal fry registers of pre-adolescent, adolescent, post-adolescent, and adult subjects as revealed by the two methods of measuring glottal area function.















CHAPTER II


PROCEDURES


The plan of this study was to obtain simultaneous indications of glottal area function as photographed by means of ultra-high-speed camera equipment and as sensed by the phototransistor of a photo-electric glottograph during the production of falsetto, modal, and vocal fry phonation. Five individuals (1 pre-adolescent male, 1 adolescent male, 1 post-adolescent male, 1 adult male, and 1 adult female) who possessed no evidence or history of voice disorders or laryngeal pathology served as subjects. The photographs and glottograms were obtained during sustained phonation of the vowel /i/ which provides the most complete exposure of the anterior-posterior aspect of the vocal folds. The samples were produced at medium vocal effort for each of the registers. The vocal registration was determined by agreement between the subject and the investigator.


Equipment

A block diagram of the instrumentation employed in the present study is shown in Figure 1. The subject sat on an

15







I I I Ti i g S s e Photographic System Sound Recording I Light Sensing Timing System System j System

Wollensak T AmpeX eBK 5X Amplifier Hewlettoose mpxPackard 5 Kc

Model 301 Model 354 Amp i Square Wave
Douglas Generator
Model 531 .Model 211 AR
oscilloscope + I
" / I Flsh 'Frokjaer-Jensen

Fa ui Type LG900 Lens JGlottograph

IL Timing Light and Larynx
C. -Artificial Larynx Synchronizing FASTAX Light S -ensorPus
e Generator IBruel-Kjaer Condensbr I oblique Microphone #4134 mirrorwith 4-nun probe tube
-] Front Filters I
~f~h~J Lens
G.E. Laryngeal I Lamp Mirror Trigger Button


Figure 1. Block diagram of equipment used *to obtain the voice signals and simultaneous photographic and glottographic records of glottal area functions for each subject








adjustable stool before a fixed laryngeal mirror and introduced the mirror and the probe microphone into his pharynx. The phototransducer was affixed to the neck of the subject in the space between the anterior-inferior boundary of the thyroid cartilage and the anterior-superior boundary of the cricoid cartilage. The neck was then draped with black plastic material to block extraneous light sources and thus achieve maximum sensitivity of the transducer. Following amplification of the voice source by a condenser amplifier and of the glottal area function by the glottograph and external amplifier, each of the two signals were displayed on the face of an oscilloscope. As the subject achieved full exposure of the vocal folds for each register production, ultra-high-speed photographs were made of the vocal folds and of the product of the voice source and glottal area function as displayed on the face of the oscilloscope.

For monitoring phonation, the voice signal was recorded on an Ampex 354 full-track tape recorder. This signal was subsequently subjected to fundamental frequency analysis on an oscillographic writer and to intensity analysis by a graphic level recorder.


Instrumentation for obtaining ultrahigh-speed photographs

Standard Fastax ultra-high-speed photographic procedures






18

were used to obtain simultaneous photographs of the activity of the vocal folds and the photocell monitoring device recorded on one channel of an oscilloscope. This activity was filmed at an exposure rate of approximately 5000 frames per second.

Specifically, the photographic equipment consisted of a model WF 301 Wollensak Goose control unit set in position one to allow the event and the camera to start at the same time. The event is defined as the desired exposure of the laryngeal area. At the end of the time cycle, the event and the camera stopped. A timing impulse control unit made up of a time delay, electro-larynx drive, electro-larynx and a flash/sync generator were used to drive a Heathkit square wave generator thus producing a 5 kc square wave and flashing a time marker for the film. The 5 kc square wave was displayed on one channel of a Douglas 531 four channel oscilloscope.

The side lens of the Fastax camera was focused on the face of the oscilloscope, thereby photographing the wave form generated by the 5 kc square wave as well as those produced by the input of the photo-electric glottograph and the condenser microphone. The Fastax camera was fitted with an extension bellows and a 152 mm telephoto lens. The F stop was set at 3.5. The camera was loaded with Kodak 4 XR film #7277 with double perforations of 0.3000 inch pitch.








Illumination for filming and for glottography was provided by a 2100 watt, 60 volt General Electric projector lamp. Cooled light was directed by lenses and a mirror to a laryngeal mirror in the pharynx from which it was reflected into the larynx. The light that passed through the glottal opening between the vocal folds struck the anterior laryngeal wall subglottally and at the level of the phototransducer located on the external surface of the neck.


Instrumentation for obtaining phototransistor glottographs

Glottographs of glottal area function were obtained by a phototransducer mounted in a polyethylene tube with an internal diameter of 2 mm and an external diameter of 3 mm. The transducer was connected to a DC amplifier housed in a Frokjaer-Jensen Photo-electric Glottograph, type LG900, by means of a short shielded cable. This, in turn, was connected to a five times operational amplifier.

In total darkness, the phototransducer produces an input voltage to the amplifier of 0.7 /uV. In position over a closed glottis the input voltage is 3.7 /uV. A fully opened glottis yields an input voltage of 6 mV. The input level varies only up to 3 mV under conditions of normal phonation. Amplification of the glottography system is linear within plus or minus 2 percent in the normal working range of 0-3 mV








input voltage. The power gain is 80 dB maximum. By means of a Philips Stroboscope type 9103, the frequency range of the glottograph was tested and found to be from DC-10,000 Hz. The signal from glottograph output A with a 4 volts/67 mA maximum in a 60 ohm load was input to a five times operational amplifier and thence to one channel of the Douglas 531 four channel oscilloscope. The output voltage selector was set at maximum.

In the present study, the phototransistor was affixed with adhesive tape to the neck of the subject at a point rostral to the superior boundary of the anterior aspect of the cricoid cartilage. The upper and lower length of the transistor was indicated by a line drawn on the neck of the subject. This procedure was used to afford a position monitor for the transistor at the beginning of each film sequence. The neck was then draped with black plastic material to exclude all extraneous light. Measurement devices

To provide a base line response from the glottograph

and a millimeter scale for glottal area measurement, a modification of the procedure developed by Hollien and Moore (1960) was employed. Immediately following the photographic procedure for each register and without any adjustment of the lens, a millimeter scale was photographed. This technique








permitted measurement not only of vocal fold function but also served as a "no response" line for measuring glottograph response.


Experimental Procedure

Subjects

Five individuals were selected for the investigation.

The subjects volunteered to perform the tasks and were chosen on the basis of the following criteria: (1) they showed no evidence or history of voice disorders or laryngeal pathology; (2) they were capable of complete anterior-posterior exposure of the vocal folds; (3) they were capable of phonating in vocal fry, modal, and falsetto registers; and (4) they presented five types of laryngeal development (pre-adolescent, adolescent, post-adolescent, adult male, and adult female). Training Procedure

To provide an experimental environment for subject task practice, a training station was devised. This consisted of a fixed laryngeal mirror, a light source,. and a subject viewing apparatus. The light source consisted of a high intensity lamp with the beam directed onto an ophthalmoscope mirror. The reflected light of the ophthalmoscope mirror was adjusted to illuminate the fixed laryngeal mirror. The subject viewing apparatus included a transparent mirror placed at a 450 angle which directed the image of the vocal folds to a mirror 20 inches above the base of the station.






22

This mirror was inclined at a variable angle, thus allowing the subject to monitor vocal fold activity and duplicating the experimental environment. The back of the station support box was equipped with a two-way mirror to permit viewing of the larynx of the subject by the investigator. A three-dimensional drawing of the station may be seen in Figure 2.


Tasks

After the investigator affixed the phototransducer for recording glottograph traces, the subject sat on an adjustable stool before a fixed laryngeal mirror and introduced the laryngeal mirror and probe microphone into his pharynx. Each subject was then instructed to produce a sample of the vowel /i/ in the register under investigation. The investigator observed the image of the subject's vocal folds in the viewing mechanism of the camera and checked the accuracy of the phonation range. The traces of the output of the microphone and the glottograph were observed for adequate illumination and placement on the face of the oscilloscope. When a full exposure of the vocal folds was visualized the investigator indicated that the subject should sustain the phonatory posture and the controls for the camera and the electrolarynx were triggered by the remote switch. Thus,



















Hinged mirror for subjects to view - camera training


Back side has two-way mirrorfor subject and observer viewing


6Y Ophthalmoscope










Light source Laryngeal mirror







Figure 2. Three-dimensional drawing of subject training station








simultaneous indications of laryngeal activity as viewed through the camera, sensed by the phototransducer, and recorded by the microphone, were photographed in each of three registers. After each phonation was filmed, the millimeter grid was brought into focus and photographed as the subject viewed the adducted vocal folds with a hand-held laryngeal mirror. Appendix A contains the recording itinerary followed throughout.


Falsetto Range

Following the experimental tasks, verification of the production of falsetto register was confirmed for each subject. Four phoneticians and the investigator listened to the production of the complete vocal range of each subject. The experimental sample was then reproduced immediately after the vocal range was presented. Agreement between the judges was found to be unanimous for each sample. The judgment form is reproduced in Appendix B.


Data Analysis

Glottal area function was determined by traditional

polar planimeter methods with total area of glottal opening in successive film frames plotted as a function of time. This necessitated measuring the activity of the vocal folds for each frame in a cycle three times and dividing by three








to obtain a high degree of accuracy. The use of a Kodak Photo-Optical Data Analyzer Model 224 allowed for stopframe study of each frame for each segment of the film used in the study. Measurement techniques are outlined in Appendix C.

The output of the phototransducer for the glottal cycle was analyzed by measuring the deviation of the oscilloscope trace of the glottograph from the base line established by the 5k Hz signal. The drafting instruments developed and employed for this measurement procedure are described in Appendix C.

Amplitude and area curves were developed as a function of time and plotted for each experimental condition. Measurements were made on six complete cycles selected from the 50- to 60-foot portions on the 100-foot films. The curves developed through analysis of the data are presented in Appendix D.

Frequency of phonation was derived for each subject in each experimental condition by means of input from the Ampex tape recorder to the Honeywell model 1508A Visicorder. Waveto-wave measurement of the output of each sample was made by measuring the wave from peak-to-peak and dividing this by the time interval selector of the Visicorder to obtain the period. Frequency in Hz was computed by the formula F = l/T.






26

A summary table of frequency of phonation for each individual during each experimental condition may be seen in Appendix E.

The relative intensity levels were computed for each phonational sample. The output of the Ampex 352 dual channel, full-track tape recorder was coupled to a Bruel and Kjaer 2305 Graphic Level Recorder which recorded the sample signal at a mid-scale with a 50-dB potentiometer. The measures were made relative to this scale. These results are summarized in Table 6 in Appendix F.

To determine the relationship between the wave forms. developed by measurement of the glottal area function derived through analysis of the photographic film and by measurement of the deviation of the quantity of light sensed by the photo-electric glottograph, the curves were normalized and a Pearson product-moment correlation coefficient was computed for each sample for each subject. In addition, confidence intervals were developed for each computation. To determine the degree of approximation to glottal area function produced by the data derived from the photo-electric glottograph within three registers, a comparison was also made between the correlation coefficients developed for the samples obtained from each subject.















CHAPTER III


RESULTS


The major purpose of this study was to determine

whether photo-electric glottography was as valid for depicting glottal area function as frame-by frame analysis of ultra-high-speed film. The two investigative techniques were compared in three vocal registers (vocal fry, modal, and falsetto) produced by five normal larynges representing different ages and sexes (pre-adolescent male, adolescent male, post-adolescent male, adult male, and adult female). The raw data were acquired by recording glottal area vibrational variaEions simultaneously in high-speed motion pictures and in photo-electric glottogram oscillograph traces. The five subjects phonated the vowel /i/ at two fundamental frequency regions, one in each of two registers, modal and falsetto. In vocal fry registration, it was possible to obtain only two satisfactory samples of subjects phonating the vowel /i/ at appropriate fundamental frequencies. The reader should be aware of the difficulty of vocal fold exposure in this register (Timcke et al., I959).

27 .






28

One sample of each experimental condition was obtained, the criterion measures were register production, vocal fold exposure, relative intensity, and phototransducer activity for each sample in each condition.

It should be noted that the glottal area function and

the fundamental frequency of phonation obtained in the present study varied from subject to subject as well as from register to register. For example, Subjects One and Three were found to demonstrate incomplete glottal approximation in modal register, whereas, te other subjects did achieve complete vocal fold closure in this register. This manner of approximation in modal register is in disagreement with that found by other investigators (van den Berg, 1958, 1960; Vennard, 1960; Moore, 1968a). Further, the frequency of phonation in each of the registers differed from subject to subject. However, in each sample of register production the frequency produced by the individual subject was appropriate for the age and sex of that subject (Luchsinger and Arnold, 1965; Sedlackova, 1961; van Oordt and Drost, 1963). To identify -the developmental stage and sex of each of the subjects, the following designations were used: Subject. One, adult male; Subject Two, adult female; Subject Three, post-adolescent male; Subject Four, adolescent male; and Subject Five, pre-adolescent male.








Correlation of Glottal Area Function Depicted
by Photo-electric Glottographs and UltraHigh-Speed Pictures

Modal register

The measurements of glottal area function derived from the frame-by-frame analysis were normalized for each of the subjects. These data were then compared to the normalized data derived simultaneously from the signal transmitted by the phototransducer of the photo-electric glottograph. The amplitude and area curves thus obtained for each of the five subjects producing modal register phonation are shown in Figures 3-7. These figures and Table 1 indicate that the phototransducer output and glottal area function measurements correlate significantly at the .05 level of confidence.

In view of the variability exhibited by the subjects

during the production of the modal register frequencies, no definitive statement is possible about the ratio of the opening, closing, and closed phases of the vibratory cycle. It should be noted, however, that there appear to be several modes of vibration appropriate to this register.

Similarity of closure was noted between Subjects One

and Three. Figures 3 and 4 clearly indicate that the frameby-frame analysis and the glottogram show no complete closure for either of the two subjects. In addition, the two methods define the open quotient and speed quotient













fo 210 Hz

Intensity 32 dB re 10 mV Photographic Glottographic----------


100







z
H z w
024 0 F4
0










0


0 FRAMES


Figure 3. Plot of glottal area function and photo-electric glottograph for one normalized cycle. Subject 1, modal register.








equally well. The vibratory cycle characteristic of Subject One began with a lateral movement of the upper margins of the vocal folds. Discernible opening began just anterior to the middle one-third of the vocal folds and proceeded toward the anterior and posterior commissures. As the upper margins completed their lateral displacements, the lower margins could be seen approaching the mid-line at the posterior one-third of the vocal folds. This closing movement proceeded synchronously to the anterior commissure. The upper margins began their adductory movement at the middle one-third and the closing gesture continued to the anterior commissure. As shown in Table 2, the open quotient was found to be .88, while the speed quotient was 1.0.

The opening phase noted in the vibratory cycle produced by Subject Three was found to show incomplete closure .as indicated in Figure 4. Initial opening began at the posterior commissure. The phase relationship of the upper and lower margins of the vocal folds presented an appearance


1open Quotient will be defined by the formula:
Fraction of cycle during which glottis is open
Duration of entire cycle
Speed Quotient will be defined by the formula:
S.Q. T-ime of abduction or lateral excursion
Time of adduction of medial excursion

(Timcke et al., 1959).













fo 180 Hz

Intensity 37 dB re 10 mV Photographic Glottographic----------


100







0
H

0

0



U







0


FRAMES 20


Figure 4. Plot of glottal area function and photo-electric glottograph for one normalized cycle. Subject 3, modal register.








similar to that of Subject One. During the opening phase, the only difference noted between the two was that the lower margins of the vocal folds were not visible. On closing, the upper margins appeared to move laterally while the lower lip of the vocal folds presented a synchronous closing movement originating at the posterior commissure. As this adductive movement was completed, the upper lip moved to the mid-line. During the closed phase, the mass of the thyroarytenoid muscles seemed to compress toward the midline. This movement continued into the beginning of the following opening gesture. A stiffening2 of the edge of the vocal folds then appeared to occur. The open quotient was found to be .61 with a speed quotient of 2.0 as shown in Table 2. These figures were acquired through analysis of the averaged data derived by the two methods of investigation.

The vibratory pattern characteristic of Subject Two

was found to employ a movement pattern more similar to that described by Timcke et al. (1958), in their discussion of the vocal fold margins. The opening commenced at the


2Stiffening will be operationally defined as the
appearance of a compact, dense, and fixed positioning of the vocalis musculature (van den Berg, 1958).









posterior commissure with the upper margins of the vocal folds separating first. This opening then seemed to progress to the middle one-third. The main body of the thyroarytenoid musculature appeared to move laterally away from the mid-line. Complete abduction to the anterior commissure was then achieved exposing the lower margins of the folds which seemed to be similar in contour but of smaller total area than that encompassed by the upper margins. This activity was followed by complete opening of the lower margins of the vocal folds., On closing, both vocal folds appeared to move synchronously toward the mid-line where they contacted each other simultaneously throughout their entire borders. During the closed phase, movement of the thyroarytenoid muscles toward the mid-line was observed. Figure 5 and Table 2 allow a definition of the open quotient to be .70, while the speed quotient may be calculated to be 1.75 as analyzed by an average of the two methods. The speed quotient suggests that the intensity of phonation was medium to loud (Timcke et al., 1958) which was confirmed by reference to the recorded sound.

The action noted in the vibratory pattern for Subject Four was found to be grossly similar to that noted for Subject Two. This similarity may be noted by a comparison of Figures 5 and 6. Moreover, the stiffening posture of the












f 240 Hz

Intensity 38 dB re 10 mV Photographic Glottographic----------


100






Z
H Z
W 0


0


rX0
0


0 FRAMES


Figure 5. Plot of glottal area function and photo-electric glottograph for one normalized cycle. Subject 2, modal register.












fo 170 Hz

Intensity 35 dB re 10 mV

Photographic Glottographic----------


100







Z
H Z 0 0


0.4 0


0 FRAMES 20


Figure 6. Plot of glottal area function and photo-electric glottograph for one normalized cycle. Subject 4, modal register.






37

vocal folds noted in the vibratory pattern of Subject Three was apparent during the opening phase for Subject Four. Of interest to the investigator was the initial opening gesture in which a massing of the upper margins of the vocal folds seemed to continue for several frames into the opening phase. However, as the closing phase began, the vertical expanse of the marginal surfaces of the vocal folds presented a somewhat concave appearance. Again, the initial adductory movement of the folds began at the posterior commissure. For this subject, analysis of the averaged data obtained from the phototransducer and the photographic film revealed the open quotient to be .57 with a speed quotient of 1.40 as shown in Figure 6 and Table 2.

The activity of the vocal folds of Subject Five differed somewhat from that of the other subjects. Specifically, during the closing gesture, the lower margins of the vocal folds approached initially at the posterior commissure. Although this gesture was not completed, the anterior portion of the folds did then adduct. Contrary to other observations made during the production of modal register, the upper margins of the folds began the adductory movement at the anterior commissure. Adduction was then characterized by symmetrical, progressive, medial movement of the vocal folds terminating in complete glottal closure. Some lateral








movement of the thyroarytenoid muscle mass was noted. Figure 7 and Table 2 illustrate that the two methods of investigation indicated an open quotient of 1.0 with a speed quotient of 1.5 for this subject.

To determine the relationship between the two methods

used in determining glottal area function, a Pearson productmoment correlation coefficient (Hays, 1963) and 95 percent confidence intervals (Walker and Lev, 1953) for each correlation were calculated. The results of this analysis are presented in Table 1. As may be seen from inspection of the table, the correlation between the photographic and glottographic methods for evaluating glottal area function in the modal register is significant. Falsetto register

The results of analysis of the normalized amplitude and area curves derived through measurement of the glottal area function from the developed photographic film and the traces representing the deviations of the quantity of light sensed by the phototransducer for the five subjects producing phonational frequencies within their falsetto register range are shown in Figures 8-12. A significant correlation was found to exist between the results of the analysis obtained from the two methods of measurement.












fo 360 Hz

Intensity 43 dB re 10 mV Photographic Glottographic----------


100




0
z
H
z 04
0


r.)









0


0 FRAMES 20


Figure 7. Plot of glottal area function and photo-electric glottograph for one normalized cycle. Subject 5, modal register.









Table 1. Summary table of correlation between photo-electric glottographic and photographic methods of analysis for modal register



Confidence Intervals

Subject r df Lower Limit Upper Limit


1 .90* 16 .747 .963 2 .92* 16 .795 .970 3 .97* 19 .926 .987 4 .98* 18 .949 .955 5 .96* 8 .836 .991


*Significant at the .05 level











f 490 Hz

Intensity 40 dB re 10 mV

Photographic Glottographic----------


Figure 8. Plot of glottal area function and photo-electric glottograph for one normalized cycle. Subject 1, falsetto register.













f 680 Hz

Intensity 25 dB re 10 mV Photographic Glottographic----------


100









0

0
0

U E-4







0


FRAMES 20


Figure 9. Plot of glottal area function and photo-electric glottograph for one normalized cycle. Subject 2, falsetto register.













f 300 Hz
0
Intensity 36 dB re 10 mV Photographic Glottographic----------


100







Z
H 0

0 U C4



04


0 FRAMES


Figure 10. Plot of glottal area function and photo-electric glottograph for one normalized cycle. Subject 3, falsetto register.











f 580 Hz Intensity 30 dB re 10 mV Photographic Glottographic----------


100






0
Z
H
Z



0



ra4






0


0 20
FRAMES


Figure 11. Plot of glottal area function and photo-electric glottograph for one normalized cycle. Subject 4, falsetto register.












f 760 Hz
0

Intensity 42 dB re 10 mV Photographic Glottographic----------


100








H a4
0

0


E-4
Z




(0


0 FRAMES


Figure 12. Plot of glottal area function and photo-electric glottograph for one normalized cycle. Subject 5, falsetto register.









There was great variability in the frequency of phonation produced by the subjects during the execution of this register. In addition, there appeared to be three modes of vibration exhibited by the subjects of this study.

Subjects One, Two, and Three seemed to present similar modes of vibration. There was no complete closure of the vocal folds during the "closed" phase. For these individuals, an opening at the posterior commissure was apparent. This is similar to the type of vibrational pattern described by Rubin and Hirt (1960) as open chink. For these subjects, opening was initiated at the middle one-third and was rapidly continued to the anterior commissure. On closing, the folds approached first at the anterior commissure and completed their closing gesture with contact of the middle one-third of the vocal folds. The figures associated with the glottal area function for each of the subjects reveal that the open quotient and the speed quotient were both found to be 1.0 (Table 2). This suggests that the intensity of production in this register was medium to high and that there was considerable tension of the vocal folds as predicted by Timcke et al. (1958) and Rubin and Hirt (1960).

The vibratory pattern for Subject Four differs from that previously described in the production of falsetto register in that, during the opening phase, the folds









were seen to approximate at the middle one-third with an opening present at the anterior and posterior one-thirds of the vocal folds. Complete opening was then achieved with an apparent stiffening of the margins of the folds. Closure was accomplished in much the same manner as that of the aforementioned subjects. An inspection of Figure 11 and Table 2 reveals an open quotient of 1.0 with a speed quotient of .75.

A most unusual pattern of vibration was noted in the phonatory mechanism for Subject Five. Although no closed phase was accomplished, the opening first perceived was at the middle one-third of the vocal fold expanse rather than at the posterior portion as had been noted in the previous patterns. In addition, the opening phase showed an abduction of the anterior portion extending to the anterior commissure. On closing, the posterior one-third made adductory movements followed by adduction of the anterior portion of the vocal folds. Again, there appeared to be stiffening of the margins of the vocal folds. However, this process seemed to progress segmentally throughout the length of the vibrating vocal folds. It is evident from the results displayed in Figure 12 and Table 2 that the opening quotient for this subject was 1.0 with a speed quotient of .66.








A Pearson product-moment correlation coefficient

and 95 percent confidence intervals for each correlation were calculated. The results depicted in Table 3 indicate that the correlation between the two methods of investigation is significant for the falsetto register. This finding is in agreement with the suggestion made by K~ster and Smith (1970). The glottograms derived in falsetto register are more representative of glottal area function than in registers encompassing the lower phonational ranges.


Vocal fry register

For vocal fry register, the resulting normalized

amplitude and area curves produced by two subjects are shown in Figures 13 and 14. Although the five subjects participating in the investigation were capable of producing vocal fry phonation, for only two of the subjects were measurable film images obtained. Analysis of the measurement of the glottal area function in the frame-by-frame display of the developed film and the deviation of the trace produced by the sensitivity of the phototransducer revealed that the correlation between the two methods of investigation is not as exact in vocal fry register as in the modal and falsetto








Table 2. Summary table of averaged indications of open quotient and speed quotient for each subject in modal and falsetto register phonation derived through analysis of photographic film and photo-electric glottograms* Modal Register Falsetto Register Subject O.Q. S.Q. O.Q. S.Q.


1 .88 1.00 1.00 1.00 2 .70 1.75 1.00 1.00 3 .61 2.00 1.00 1.00 4 .57 1.40 1.00 .75 5 1.00 1.50 1.00 .66


*No figures for the open quotient or speed quotient for vocal fry phonation were included due to lack of protocols for this register.










Table 3. Summary table of correlation between photo-electric glottographic and photographic methods of analysis for falsetto register Confidence Intervals

Subject r df Lower Limit Upper Limit


1 .95* 7 .775 .989 2 .98* 6 .889 .997 3 .94* 15 .838 .979 4 .98* 7 .905 .995 5 .97* 5 .807 .995


*Significant at the .05 level













f 10 Hz

Intensity 37 dB re 10 mV Photographic Glottographic----------


100






o I.
I I

-4 .
o, I ;J L I " '














0
10 20 30 40 50 60 70 80 90 FRAMES







Figure 13. Plot of glottal area function and photo-electric glottograph for one normalized cycle. Subject 3, vocal fry register.











f 20 Hz

Intensity 34 dB re 10 mV

Photographic Glottographic----------


100






0
H z w 0


z




o 0


10 20 30 40 50


FRAMES






Figure 14. Plot of glottal area function and photo-electric glottograph for one normalized cycle. Subject 4, vocal fry register.








registers. However, it should be noted that a striking similarity exists between the derived curves.

The vibratory pattern displayed in Figure 13 in the

production of vocal fry was regularly biphasic for Subject Three. Production of a multiphasic vibratory cycle is in agreement with the results of earlier investigations (Timcke et al., 1959; Moore and von Leden, 1958). The initial opening excursion was seen at the anterior commissure followed slowly by opening to the middle one-third of the vocal fold margin. There was an adductory motion leaving a narrow eliptical opening to the posterior commissure. This posture was maintained until the second excursion of the folds appeared. A lateral motion was visualized, followed by complete opening of the folds. Closing was characterized by a rapid closure of the anterior portion of the folds followed by complete anterior-posterior closure. The closed phase allowed for visualization of the participation of the ventricular folds. It should be noted that in no-, instance was the activity of the vocal folds obscured by the participation of the ventricular folds. As the margins of the folds closed, there appeared to be massing of the marginal area. The ventricular folds gradually receded into a completely abducted position. This observation is in agreement with the results reported in a recent laminagraphic






54

investigation (Allen and Hollien, in press) which suggested that the ventricular folds operate in conjunction with the vocal folds in vocal fry phonation and, further, that the travelling wave from the true folds is continued into this mass.

The time ratio and area of the two opening gestures noted in the single vibratory cycle are of interest. The initial excursion occupied twice the time period noted for the second while the area measurement indicated an increase of 25 percent for the second excursion. The closed period accounted for approximately two-thirds of the entire period.

For Subject Four, the production of vocal fry registration was characterized by a vibratory cycle demonstrating a triple opening sequence and an occasional vibrational cycle presenting a single opening pattern. The initial excursion was evidenced by a small circular opening at the posterior commissure followed by a narrowing of this area. The second excursion demonstrated a triangular configuration; the smaller opening appearing at the posterior commissure. This opening was then modified to that of a narrow ellipse. The final abduction apparently involved a greater length of the folds and appeared to be symmetrically elliptical in shape. A rapid closure was then accomplished. The single opening pattern was seen only on occasion and may not have been






55

representative of the subject's normal phonatory posturing in vocal fry phonation.

More participation of the ventricular folds was noted in the production of this register by Subject Four than by Subject Three. It should be noted also that for Subject Four, the vocal folds seemed to be covered by a considerable amount of mucus. The action of the ventricular folds was similar to that of Subject Three. Although protocols for defining the open quotient and speed quotient for a single cycle have not been established for vocal fry register, it is of interest to note the quotients for the excursions of the folds visualized in one vibratory cycle. In the photographic film analysis, Subject Four displayed an open quotient for the first excursion of 1.0 with a speed quotient of 3.43. The second excursion presented an open quotient of 1.0 with a speed quotient of 1.09; whereas the third abductory posture indicated an open quotient of

1.0 and a speed quotient of .60. The closed phase for the entire cycle occupied about one-third of the entire cycle.

The results of computation of a Pearson-r between the two methods of investigation, together with 95 percent confidence intervals for each correlation, are shown in Table 4. Indications are that, although correlation does










Table 4. Summary table of correlation between photo-electric glottographic and photographic methods of analysis for vocal fry register Confidence Intervals

Subject r df Lower Limit Upper Limit



3 .65* 128 .153 .943 4 .84* 78 .761 .895


*Significant at the .05 level









exist, the phototransducer is not capable of sensing the same information as that derived visually from photographic film.

For the five subjects used in the present investigation, it may be concluded that the photo-electric glottograph is as valid a method for depicting glottal area function as a frame-by-frame analysis of ultra-high-speed film in the modal and falsetto registers. In addition, although the correlation found between the two methods of investigation is not as high in the vocal fry register, the photo-electric glottograph does appear to be capable of generating reasonably approximate information.
















CHAPTER IV


DISCUSSION


The subjects in the present investigation utilized

essentially normal laryngeal mechanisms (as determined by the subject selection criteria) to produce samples of vocal fry, modal, and falsetto phonation which were analyzed through measurements obtained from a light sensing device and a high-speed camera. The results indicate that the two methods of analysis are highly correlated in their ability to depict similarly shaped area and amplitude curves for the modal and falsetto registers.. In addition, although the statistical analysis of the correlation between the curves developed by the two methods in vocal fry phonation did not indicate a one-to-one relation, there was striking similarity in the overall configuration.


Glottographic Response and Vibratory Pattern in Modal Register

Prior to this study, few studies had been made employing simultaneous recordings of vocal fold activity through the utilization of glottography and photography. Moreover, 58






59

the investigators who did obtain such simultaneous recordings (Coleman and Wendahl, 1967, 1968) found that the photoglottograph was limited in providing accurate indications of glottal area function. The present study disagrees with those findings in the modal register. Although it was apparent that the phototransducer signaled the opening phase earlier than the occurrence of this phase became visible, this is not in disagreement with the suggestion of Sonesson (1960). He surmised that the glottograph could be useful in defining the phase relationship of the lower and upper edges of the vocal folds. It is possible that such complex glottal configurations can cause the differences noted in the results of this study. Timcke et al. (1958) found that, although the shape of the amplitude curves produced by the upper and lower margins of the vocal folds was similar, there was a definite phase shift between the two movements. The upper margin was seen to lag behind the lower margin in the closing phase of the vibratory cycle. This finding would seem to be verified by the results of this study.

Variations in the vibratory patterns produced by the subjects of this study suggest that in the modal register there is no completely predictable mode of vibration. Brackett (1947) found not only differences between subjects but also intra-individual differences. A careful comparison








of individual vocal fold movement and vocal intensity confirmed the hypothesis set forth by van den Berg et al. (1957) that the Bernoulli effect, i.e., the negative pressure in the glottal air, is enhanced with an increase in tone intensity and air flow thereby drawing the vocal folds more forcibly toward each other.

Observations also supported the myoelastic-aerodynamic theory of phonation. The vocal fold margins of two of the subjects participating in the study demonstrated an active longitudinal tensing. van den Berg (1968a) reports that this contraction is expected in sustained phonation of a given fundamental frequency. The concave appearance of the vocal fold margins visualized in the phonation of one subject suggests an active participation of the vocal muscles.


Glottographic Response and Vibratory
Pattern in Falsetto Register

The correlation between the measurements of glottal area produced by photo-electric glottographs and by the photographic film in falsetto register found in this study tend to conform to the conclusions stated by Koster and Smith (1970). They found that in phonations in the higher frequencies the glottograph provided a more accurate representation of glottal area function than in the lower frequencies. The present results indicate that the








glottographs are significantly related to glottal area function in the falsetto register. It may be surmised that the thin shape of the glottal margins creates single, relatively simple glottal folds in which there is no change of phase relationship between an upper and lower margin. Thus, it could be expected that the light passing between the vocal folds would not be impeded by vocal fold tissue during the opening and closing phases of the vibratory cycle.

Although Rubin and Hirt (1960) indicated three basic patterns of vibratory activity in the falsetto, only that designated as open chink was observed in this investigation. However, two subjects presented vibratory patterns that have not previously been described.

A narrow opening at the posterior and anterior commissures with closure at the middle one-third of the vocal fold expanse characterized the vibratory pattern for one of the subjects. Contrary to this mode of vibration, another of the subjects achieved closure that may be described as "tight" at the posterior and anterior commissure. The opening thus accomplished only encompassed the middle onethird of the vocal folds.

In all cases, the appearance of the vocal folds did

correspond with that suggested by other investigators (van den Berg, 1960; Luchsinger and Arnold, 1965). The vocal









folds were indeed elongated and appeared to be thin in mass. There also seemed to be an obvious stiffening of the vocal fold margins which suggests a longitudinal tension in the vocal ligaments without an active vibratory participation of the vocal muscles. However, the visualization of adduction of the middle one-third of the vocal folds suggests that myoelastic properties may have been active in at least this individual's production of falsetto register. The appearance of tight closure of the anterior and posterior commissures with resulting open chink at the middle one-third may possibly be explained by myoelastic properties. It was obvious that aerodynamic properties were predominant in the falsetto production of the remaining subjects.


Glottographic Response and Vibratory
Pattern in Vocal Fry Register

Previous investigators of the vibratory pattern for vocal fry phonation have obtained photographic film of phonation that should be considered as high fry or fry mixed with modal (Timcke et al., 1959; Moore and von Leden, 1958). The samples obtained in this investigation were of fundamental frequencies more appropriate to this register (Allen and Hollien, in press; Hollien et al., 1966; Hollien and Michel, 1968; Hollien et al., 1969). Although the results for this register are based on the production of only two subjects, it was concluded that this information









was valid. The magnitude of the correlations indicated inexact replication of.vocal fold activity by the photoelectric glottograph as compared with the analysis of photographic film. However, the curves representing the variations of quantity of light and the glottal area function do seem to bear a striking similarity.

Information on the physiologic parameters of vocal fry phonation have not been completed. However, the glottograms resulting from this study resemble those analyzed by Ohala (1966) and Fourcin and Abberton (1971). These investigators termed the vocal production as creaky voice.

The vibratory pattern exhibited by one of the subjects was like that defined by Moore and von Leden (1958) as dicrotic dysphonia. The biphasic excursions of the vocal folds were found to represent amplitudes similar to those suggested by those investigators, i.e., the first was approximately 25 percent smaller in amplitude than the second.

The vibratory pattern characteristic of the fourth

subject was unusual in that it appeared to represent three definite excursions of the vocal folds. Each of the excursions seemed to be about 25 percent larger than the preceding one. It has been suggested (Hollien et al., 1966) that there may exist various modes of vibration for this register.









Hollien (1971) has postulated that myoelastic properties are predominant in vocal fry register. Visualization of the laryngeal mechanism reveals that this must indeed be the case. However, due to the complexity of relationships of the structures involved in the production of this register, much more extensive investigation of the vocal phenomenon is indicated before aerodynamic properties can be dismissed as contributory to the control of change of frequency.
















CHAPTER V


SUMMARY AND CONCLUSIONS


The purpose of this investigation was to determine the validity of the information concerning vocal fold activity derived through measuring the quantity of light passing through the glottis by comparing these data with measurements of glottal area taken from the photographic images. A second purpose was to describe and compare the characteristic glottal configurations in falsetto, modal, and vocal fry registers of pre-adolescent, adolescent, post-adolescent, and adult subjects as revealed by the two methods of measuring glottal area function. Based on the hypothesis that measurement of the quantity of light passing through the glottal area should reveal the same information as that derived from measurement of the discernible space between the vocal folds, simultaneous recordings on photographic film were obtained of the image of the vocal folds, the voice signal, and the deviations sensed by the phototransducer during vocal fry, modal, and falsetto phonation. Recordings were made during sustained phonation of the vowel

65






66

/i/ at frequencies appropriate to the vocal fry, modal, and falsetto registers. Deviations of the quantity of light were sensed by a phototransducer attached to the space between the thyroid and cricoid cartilages. The fundamental frequency and intensity of phonation were recorded on magnetic tape from the acoustic signal input to a probe tube microphone inserted into the oral cavity of the subject. The visual image of the vocal fold activity was photographed from the fixed laryngeal mirror introduced into the pharyix. The voice signal, deviations of the light quantity and a 5k Hz monitoring signal were displayed on the face of an oscilloscope and photographed simultaneously with the mirrored laryngeal activity by an ultra-high-speed camera. The voice signal recorded on the tape recorder was subsequently submitted to fundamental frequency and intensity analysis. Glottal area measurements were made from the photographic film by polar planimeter methods and by measuring the deviations of the light quantity. The data were normalized and submitted to statistical analysis in order to determine the correlation between the glottal area curves depicted by the photo-electric glottograph and the frame-by-frame analysis of the visual image in each of three registers.

From the investigation the following findings were obtained:








1. The measurements of the light deviations sensed

by the photo-electric glottograph during laryngeal vibrations correlate significantly at the .05 level

of significance with the photographic film frameby-frame measurements of the visual image in modal

and falsetto register.

2. The glottograms resulting from measurement of vocal

fry phonation approximate those derived from photographic film measurement, however the correlation

between the two curves is not as high in this register as is the case in modal and falsetto registers.

3. Inter-individual and intra-individual differences

of vibratory patterns of phonation were found in

vocal fry phonation.

4. During modal range phonation, there appeared to be

no established vibratory pattern for all subjects.

5. Aerodynamic properties may not completely account

for the vocal fold activity observed in the subjects of this study in the production of falsetto

register.

It appears, therefore, that the photo-electric glottograph is a valid method of determining glottal area function. This conclusion, which is also supported by data obtained using a photo-electric glottograph and separate photographic film (Sonesson, 1960), may result in more









complete understanding of the complex relationship of the upper and lower margins of the vocal folds. It might also result in a correlation between acoustically and physiologically derived data, i.e., through an inverse filtering process it may be possible to separate from the acoustically derived data valid information about the glottal area function.

The presence of at least two examples of variations from the vibratory pattern expected in falsetto register suggests a need for collecting a large and representative sample for obtaining normal values of the vibratory pattern of the vocal folds. In addition, the contribution o.f myoelastic properties should be examined in this register.

The subjects participating in this investigation

represented a wide age differential and it was found to be possible to obtain laryngeal photographs of the preadolescent, adolescent, and post-adolescent aged individuals without undue difficulty. Therefore, this suggests that photographic investigation of the phonatory mechanism at different stages of maturation may result in more adequate understanding of the phenomenon of voice change.



































APPENDICES



































APPENDIX A

PHOTOGRAPHIC SEQUENCE











Photographic Sequence Followed Throughout the Experiment


Phototransducer attached to space between thyroid and cricoid cartilages.

1. Vocal fry register /i/

2. Millimeter grid photographed as the phototransducer transmits a "no response" signal to the
oscilloscope.

3. Modal register /i/

4. Millimeter grid photographed as the phototransducer transmits a "no response" signal to the
oscilloscope.

5. Falsetto register /i/

6. Millimeter grid photographed as the phototransducer
transmits a "no response" signal to the oscilloscope.


































APPENDIX B

FORM USED IN THE JUDGMENT OF SUBJECT
PRODUCTION OF FALSETTO REGISTER











JUDGMENT FORM USED BY PHONETICIANS


Please indicate with a check mark (X), if you judge this sample of phonation to be within the subject's falsetto range.



Subject Yes No

1 2 3 4





































APPENDIX C

PROCEDURE FOR MEASURING GLOTTAL AREA FUNCTION








PROCEDURES FOR MEASURING GLOTTAL AREA FUNCTION

A. Photographic film image. The Kodak Photo-Optical Data Analyzer and an auxiliary viewing system were utilized in the detailed study of glottal area function for each subject in each experimental condition. After the film image passed through the lens system, it was reflected upward by a mirror placed at a 70 degree angle to a second mirror placed above the main body of the analyzer at a 15 degree angle. This mirror, in turn, redirected the image downward to a layout pad situated 66 inches from the mirror. The linear enlargement ratio of the picture on the film to the projected image was 1:27. The area of the vocal fold opening was traced onto the layout pad and glottal area function of the vocal folds was determined by traditional polar planimeter methods. The space between the vocal folds was outlined by the tracer point of the Keuffel and Esser model 620005 Compensating Polar Planimeter for each frame. A fine line was drawn across the periphery of the glottal area to mark the starting and finishing point. The tracer point was placed precisely at the beginning line and the scale set to zero. The outline of the opening was then circumscribed to the starting point and a reading made. Second and third circuits of the area were made with a measurement reading being made at the end of each circuit. These three readings









were then averaged to obtain area measurement in square centimeters.


B. Glottographic images. Immediately following the tracing of the glottal area, the frame displaying the glottographic signal corresponding to the first measured frame of the photographic film was determined. The signal from the electro larynx and the signal from the flash/sync generator allowed for identification of corresponding frames. The lines produced by the input of the phototransducer to the oscilloscope were measured from the parallel 5k Hz lines at a constant position for each photographic film frame. A straight line was scribed on a layout pad and placed under the image of the 5k Hz line and used as a base line. This line also served as an aid in the alignment of the parallel guide instruments employed in drafting the deviations produced by the glottograph and the voice traces. The position for measuring was chosen for consistent clearness throughout the series and held constant. A diagram of the measuring devices used in the data reduction for this investigation may be seen in Figure 15.






























Images






Thread guides


Figure 15. Diagram illustrating the measurement procedures used in data reduction for the analysis of glottal area function




































APPENDIX D

AMPLITUDE AND AREA CURVES DERIVED THROUGH GLOTTOGRAPHIC
AND PHOTOGRAPHIC FRAME-BY-FRAME ANALYSIS








I I I I I I I I
Time in Milliseconds


I I I
0


25


Glottographic


I I P I J' I'I''I IIIIIIII
Flames


Photographic


Figure 16. Comparison of photo-glottographic and photographic frame-by-frame analysis for the vowel /i/ produced in modal register for Subject 1








I I I I I I I i I I I J1 1 , I1 1 I I 1 1 1 I 1 1 1 I I
0 Time in Milliseconds 40




Glottographic


Frames


Photographic








Figure 17. Comparison of photo-glottographic and photographic frame-by-frame analysis for the vowel /i/ produced in modal register for Subject 2








I I I I I I I I I I I I I I I I I I I I I I I I j I
0 Time in Milliseconds 30 Glottographic


Frames


Photographic


Figure 18. Comparison of photo-glottographic and photographic frame-by-frame analysis for the vowel /i/ produced in modal register for Subject 3








I I ;I III I I I I 1 111 ,1 i i l 1 1, 1i 1 1 1 1 1 1 1i 1, 1i 1 1 1 1 11 1 1, 1 1 1, 1 1 1 I I I I I I I I
0 Time in Milliseconds 70 Glottographic


Frames

Photographic


Figure 19. Comparison of photo-glottographic and photographic frame-by-frame analysis for the vowel /i/ produced in modal register for Subject 4


I








1i i I I I IJ I I I I i I I i I I I I I I I I I I Ii I I I 8 0 0 Time in Milliseconds 80



Glottographic


Frames


Photographic


Figure 20. Comparison of photo-glottographic and photographic frame-by-frame analysis for the vowel /i/ produced in modal register for Subject 5







I I I , , I I I I I I I I I I


I' I


Time in Milliseconds


Glottographic


1il I l I I II Frames


. i


Photographic


Figure 21. Comparison of photo-glottographic and photographic frame-by-frame analysis for the vowel /i/ produced in falsetto register for Subject 1


20


' � i


I I I







I iI I I i I iI


0


Time in Milliseconds Glottographic


I . I aI S II I. I I I I I I


Frames

Photographic


Figure 22. Comparison of photo-glottographic and photographic frame-by-frame analysis for the vowel /i/ produced in falsetto register for Subject 2








I I I I


Time in Milliseconds Glottographic


Frames


Photographic


Figure 23. Comparison of photo-glottographic and photographic frame-by-frame analysis for the vowel /i/ produced in falsetto register for Subject 3








D T 1 I I 0 Time in Milliseconds 7


Glottographic


III''" ' ll I I II
Frames

Photographic


Figure 24. Comparison of photo-glottographic and photographic frame-by-frame analysis for the vowel /i/ produced in falsetto register for Subject 4




Full Text

PAGE 1

Correlation of Glottal Area Function Within Three Registers as Revealed Through Measurement of Ultra-High-Speed Photographs and Photo-Electric Glottographs By R. JOYCE REDUS HARDEN A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 1972

PAGE 2

Copyright by Joyce Redus Harden 1972

PAGE 3

ACKNOWLEDGMENTS The author wishes to express deep appreciation to her committee chairman, Dr. G. Paul Moore, for his suggestions and contributions to the ideas expressed herein. In addition, she expresses sincere gratitude to her other committee members, Drs . Edward C. Hutchinson, Norman N. Market, and Madelaine M. Ramey, for their continuous encouragement and valuable guidance during the preparation and execution of the experiment and also throughout the author's training program at the University of Florida. The author wishes to acknowledge the invaluable assistance of Messrs. David A. Campbell, Robert P. Idzikowski, and Russell E. Pierce in preparing the instrumentation for the study. Further, she expresses appreciation to Miss Elizabeth L. Allen, Mrs. Mallory W. lies, Mr. Robert L. lies, and Dr. Howard B. Rothman for their assistance in evaluating the production of falsetto register by the suojects participating in the experiment. iii

PAGE 4

In conclusion, a word of gratitude and appreciation is expressed to an understanding and cooperative family whose tolerance and encouragement as well as participation as willing subjects made the completion of this research possible . This research was supported in part by the National Institutes of Health grant NB-06459. IV

PAGE 5

TABLE OF CONTENTS Page ACKNOWLEDGMENTS iii LIST OF TABLES vii LIST OF FIGURES viii ABSTRACT xi CHAPTER I INTRODUCTION AND STATEMENT OF THE PROBLEM 1 II PROCEDURES 15 III RESULTS 2 7 IV DISCUSSION 58 V SUMMARY AND CONCLUSIONS 65 APPENDIX A PHOTOGRAPHIC SEQUENCE 70 B FORM USED IN THE JUDGMENT OF SUBJECT PRODUCTION OF FALSETTO REGISTER 72' C PROCEDURE FOR MEASURING GLOTTAL AREA FUNCTION 74 D AMPLITUDE AND AREA CURVES DERIVED THROUGH GLOTTOGRAPHIC AND PHOTOGRAPHIC FRAME-BYFRAME ANALYSIS 78 v

PAGE 6

TABLE OF CONTENTS (continued) APPENDIX Paqe E FUNDAMENTAL FREQUENCIES 91 > F RELATIVE INTENSITY 93 BIBLIOGRAPHY 95 BIOGRAPHICAL SKETCH 103 vi

PAGE 7

LIST OF TABLES Table 1 Summary table of correlation between photo-electric glottographic and photographic methods of analysis for modal register 2 Summary table of averaged indications of open quotient and speed quotient for each subject in modal and falsetto register phonation derived through analysis of photographic film and photo-electric glottograms 3 Summary table of correlation between photo-electric glottographic and photographic methods of analysis for falsetto register 4 Summary table of correlation between photo-electric glottographic and photographic methods of analysis for vocal fry register 5 Fundamental frequencies produced by five subjects used in the present investigation. The measures are reported in Hz for the sustained vowel /i/ produced in vocal fry, modal, and falsetto registers. . 6 Relative intensity levels of phonation for the five subjects used in the present investigation. The measures are reported in dB for the vocal fry, modal, and falsetto registers vii Page 40 49 50 56 92 94

PAGE 8

LIST OF FIGURES Figure Page 1 Block diagram of equipment used to obtain the voice signals and simultaneous photographic and glottographic records of glottal area functions for each subject .• . 16 2 Three-dimensional drawing of subject training station 23 3 Plot of glottal area function and photoelectric glottograph for one normalized cycle. Subject 1, modal register 30 4 Plot of glottal area function and photoelectric glottograph for one normalized cycle. Subject 3, modal register 32 5 Plot of glottal area function and photoelectric glottograph for one normalized cycle. Subject 2, modal register 35 6 Plot of glottal area function and photoelectric glottograph for one normalized cycle. Subject 4, modal register 36 7 Plot of glottal area function and photoelectric glottograph for one normalized cycle. Subject 5, modal register 39 8 Plot of glottal area function and photoelectric glottograph for one normalized cycle. Subject 1, falsetto register. ... 41 9 Plot of glottal area function and photoelectric glottograph for one normalized cycle. Subject 2, falsetto register. ... 42 viii

PAGE 9

LIST OF FIGURES (continued) Figure 10 11 12 13 14 15 16 17 18 19 Plot of glottal area function and photoelectric glottograph for one normalized cycle. Subject 3, falsetto register. . . . Plot of glottal area function and photoelectric glottograph for one normalized cycle. Subject 4, falsetto register. . . . Plot of glottal area function and photoelectric glottograph for one normalized cycle. Subject 5, falsetto register. . . . Plot of glottal area function and photoelectric glottograph for one normalized cycle. Subject 3, vocal fry register . . . Plot of glottal area function and photoelectric glottograph for one normalized cycle. Subject 4, vocal fry register . . . Diagram illustrating the measurement procedures used in data reduction for the analysis of glottal area function Comparison of photo-glottographic and photographic frame-by-frame analysis for the vowel /i/ produced in modal register for Subject 1 Comparison of photo-glottographic and photographic frame-by-frame analysis for the vowel /i/ produced in modal register for Subject 2 Comparison of photo-glottographic and photographic frame-by-frame analysis for the vowel /i/ produced in modal register for Subject 3 Comparison of photo-glottographic and photographic frame-by-frame analysis for the vowel /i/ produced in modal register for Subject 4 Page 43 44 45 51 52 77 79 80 81 82 IX

PAGE 10

LIST OF FIGURES (continued) Figure Page 20 Comparison of photo-glottographic and photographic frame-by-frame analysis for the vowel /i/ produced in modal register for Subject 5 83 21 Comparison of photo-glottographic and photographic frame-by-frame analysis for the vowel /i/ produced in falsetto register for Subject 1 84 22 Comparison of photo-glottographic and photographic frame-by-frame analysis for the vowel /i/ produced in falsetto register for Subject 2 85 23 Comparison of photo-glottographic and photographic frame-by-frame analysis for the vowel /i/ produced in falsetto register for Subject 3 86 24 Comparison of photo-glottographic and photographic frame-by-frame analysis for the vowel /i/ produced in falsetto register for Subject 4 87 25 Comparison of photo-glottographic and photographic frame-by-frame analysis for the vowel /i/ produced in falsetto register for Subject 5 88 26 Comparison of photo-glottographic and photographic frame-by-frame analysis for the vowel /i/ produced in vocal fry register for Subject 3 89 27 Comparison of photo-glottographic and photographic frame-by-frame analysis for the vowel /i/ produced in vocal fry register for Subject 4 90 x

PAGE 11

Abstract of Dissertation Presented to the Graduate Council of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy CORRELATION OF GLOTTAL AREA FUNCTION WITHIN THREE REGISTERS AS REVEALED THROUGH MEASUREMENT OF ULTRA-HIGH-SPEED PHOTOGRAPHS AND PHOTO-ELECTRIC GLOTTOGRAPHS By R. Joyce Redus Harden June, 1972 Chairman: Dr. G. Paul Moore Major Department: Speech Based on the hypothesis that measurement of the quantity of light passing through the glottal area should reveal the same information derived from measurement of the discernible space between the vocal folds, the major purpose of this investigation was to determine the validity of the photo-electric glottograph as a measuring device as compared with a frame-by-frame analysis of ultra-high-speed photographic film. A second purpose was to describe and compare the characteristic glottal configurations in falsetto, modal, and vocal fry registration of pre-adolescent, adolescent, post-adolescent, and adult subjects as revealed by the two methods of investigation. To achieve these purposes, simultaneous recordings of the visual image of the vocal folds, the voice signal and the deviations sensed by the phototransducer during falsetto, vocal fry, and modal xi

PAGE 12

phonation were obtained from five subjects. These subjects had met the following criteria: (1) they showed no evidence or history of voice disorders or laryngeal pathology; (2) they were capable of complete anterior-posterior exposure of the vocal folds; (3) they were capable of phonating in vocal fry, modal, and falsetto registers; and (4) they presented five types of laryngeal development (pre-adolescent, adolescent, post-adolescent, adult male and adult female) . Recordings were made during sustained phonation of the vowel /i/ for each of the registers investigated. Intraoral voice recordings made during photography were subjected to a fundamental frequency and intensity analysis. Measurements of glottal area function were made on six consecutive cycles from the steady-state portion of each film. Glottal area was determined in the laryngeal photographs by standard polar planimeter methods. The amplitude and frequency deviations produced by the oscillographic trace of the phototransducer were measured from the trace representing the 5k Hz signal. Correlation between the normalized curves produced by the two methods was determined and confidence intervals established. The results indicated that the photo-electric glottograph represents glottal area function as revealed in highspeed photography. In addition, the use of the glottographic xii

PAGE 13

method of investigation in conjunction with photographic procedures may lead to more complete understanding of the phase relationship between the upper and lower margins of the folds. The vibratory patterns exhibited by two subjects during the production of falsetto register suggest the participation of myoelastic factors. xm

PAGE 14

CHAPTER I INTRODUCTION AND STATEMENT OF THE PROBLEM Introduction The nature of vocal fold vibration has in the past been the subject of wide differences of opinion (Musehold, 1897; West, 1926; Metzger, 1928; Tarnoczy, 1951) . Presently, it is accepted that the vocal folds vibrate synchronously at a frequency that determines the fundamental pitch of the tone being produced, alternately allowing momentary outflow and stoppage of the air stream which is forced up from the lungs below by the muscles of exhalation (Moore, 1937b; van den Berg, 1958, 1960). The quality of the sound produced by the speaker is dependent on the manner in which the vocal folds vibrate. Researchers have employed various methods for observa-' tion of the activity of the laryngeal mechanism over the past 186 years. Moore (1937a) presented an historical accounting of methods used in vocal fold research. Black (1951), Luchsinger and Arnold (1965), and Peterson (1958) have outlined the procedures used in current investigations. 1

PAGE 15

With the advent of the laryngeal mirror, stroboscopic techniques, radiographic procedures, spectroscopic analysis, normaland ultra-high-speed photography, and glottography , much is now known of the normal vibratory patterns in the different frequency ranges of the human voice. Ultra-high-speed photography has afforded a reliable method for study of the complex vibratory movements of the vocal folds in the normal larynx (Farnsworth, 1940; Moore al ., 1962). Moore (1937b) reported that the opening of the glottis begins anterior to the mid-point of the folds and progresses in a posterior direction. This information contradicted conclusions derived earlier from stroboscopic photography on the nature of vibratory closure. BrackettÂ’s (1947) analysis of high-speed photographs of the vibrations of the vocal folds of two male subjects revealed differences in the proportion of time occupied by each of three phases of the vibratory cycle, closed, opening, and closing. Not only were these differences noted between subjects but also for different performances by the same subject. These results were confirmed by Smith (1954) . The expected time ratio of three phases was postulated by van den Berg (1968a) who stated that the glottis should remain closed for twothirds of each period for low tones; whereas the closing time should decrease with rising pitch. This conclusion led

PAGE 16

3 to a definition of the opening quotient of the glottis as representing the time occupied by each opening in relation to the total time of each period. Thus, low tones show a small opening quotient and high tones show an increasing opening quotient up to 1.0 when the glottis remains open throughout the cycle for falsetto tones. Information on vibratory patterns associated with different types of phonation is now available (Arnold, 1957; Moore, 1968a; van den Berg, 1958, 1960; Vennard, 1960). The phase relationship of the upper and lower margins of the vocal folds has aided in the physiological description of register phenomena (Miller, 1959) . Although the meaning of the term "voice registers" appears to have no completely acceptable definition (Morner et al . , 1963), there appear to be distinctive acoustic, physiological, and perceptual attributes characteristic for each of three ranges found in the normal production of voice (van den Berg, 1968b). Hollien and Michel (1968) included in their definition of "register" these criteria: (1) each shall be composed of a series of consecutive fundamental frequencies of a similar quality and (2) there should be little or no overlap of frequencies between registers. It was also found that subjective agreement between judges could be made for register change. More recently, Hollien (1971) has proposed three new terms in the definition of

PAGE 17

4 three major vocal registers. Synonymous with vocal fry, pulse register occupies the lowest range of frequencies within the individual phonation range. Modal register designates that range of frequencies correlated with normal, chest and head, or low, mid, and high. This register is composed of those fundamental frequencies that are normally used in speaking and singing. Loft register is that register ordinarily designated as falsetto. This register includes the higher fundamental frequencies within an individual vocal range. For the purposes of this investigation, it seems appropriate to rely upon two of the more traditional terms, consequently, the registers will be designated as vocal fry, modal, and falsetto. The study of vocal registration has been directed not only to the phonation of the adult but also to that of the child in an effort to define and compare various register productions appropriate for the vocal mechanism in changing stages of maturation. Investigators have established criteria for three vocal registers in the adult for both sexes (Luchsinger and Arnold, 1965; Pronovost, 1942; Snedicor, 1951; Colton, 1969; Hollien and Moore, 1960; Hollien and Michel, 1968; Hollien et al . , 1971). Sedlackova (1961) found that the phonation ranges of children demonstrated the presence of three registers. Her findings were obtained

PAGE 18

5 by means of stroboscopic motion pictures. The appearance of the vocal folds was found to be like that of the adult mechanism within the production of chest, mid, and head register. Luchsinger and Arnold (1965) and van Oordt and Drost (1963) have also defined the frequency range and register production of the voice in children. They found that falsetto and modal registers may be qualitatively differentiated. Motion picture studies (Fletcher, 1958) and radiographic procedures (Russell, 1931) have resulted in understanding of the physiologic changes manifested by the vocal folds within three registers. Vocal fry, described as a succession of laryngeal vibrations of low frequency (Hollien et al . , 1966), has been the subject of radiographic investigations . The vocal folds appear to be compact and shorter in length in production of this particular register (Hollien et al . , 1969). The results obtained from motion picture studies indicate that the vibratory cycle may be monophasic or biphasic. Further, the initial excursion of the folds presenting a biphasic pattern appears to be 25 percent smaller in amplitude than does the second. Apparently the plateaus noted in the open phase are related to the longitudinal opening waves which travel along the length of the vocal folds (Moore and von Leden, 1958; Timcke et al . , 1959).

PAGE 19

6 The modal register has been equated with the natural range, and chest or mid register (van den Berg, 1960) . Fink and Kirschner (1959) observed that the configuration of the vocal and ventricular folds should be smoothly curved and symmetrical in production of modal register. They also indicated that the medial surface of the folds should be convex during phonation. This observation was confirmed by Sovak et al . (1971) . Laminagraphic studies of the vocal folds revealed that as the fundamental frequency of phonation is raised the vocal folds tend to. elevate progressively. The angle of tilt also appears to increase as pitch elevates except in production of frequencies within the falsetto register (Hollien and Curtis, 1962). Normal speed photography reveals that in the modal register there is systematic lengthening of the vocal folds as frequency of phonation elevates (Hollien and Moore, 1960; Hollien, 1962). In addition, this lengthening may proceed in a stair step fashion. Ultra-high-speed motion pictures (Timcke et al . , 1958) revealed that the opening quotient and speed quotient are similar in head and chest register. Increases in intensity appear to effect a faster opening quotient whereas the open quotient is inversely proportional to the intensity of the sound. Pitch changes do not appear to effect a change in either speed quotients or open quotients.

PAGE 20

7 Moore (1968a) has cautioned that the initial vibratory movements of the vocal folds are usually extremely small and may be limited to only a portion of the glottal margin. Further, a medial motion may commence in a lateral direction. Vibration of the folds may appear with an open glottis when the diameter is not exceedingly large (van den Berg et al . , 1957) . Subjective perceptual agreement is possible in judging production of the falsetto range (Rubin and Hirt, 1960; Hollien and Michel, 1968) . Physiologically, the margins of the vocal folds have been found to be rather thin and pointed with vibration confined to this area (van den Berg, .1958) . Rubin and Hirt (1960) described three patterns of vibratory activity for falsetto production. These were classified as open chink, closed chink, and damping. Furthermore, they found that as the untrained singer progressed through the phonatory range to the falsetto, this production elicited turbulent, chaotic vocal fold activity. It was concluded that aerodynamic forces contributed more to production of the falsetto than did myoelastic properties. This conclusion was verified by Hollien et al . (1971) . Motion picture studies and radiographic techniques (Griesman, 1943) have been successful in clarifying some aspects of larnygeal function. Unfortunately, due to the

PAGE 21

8 mechanics of these methods, physiological studies have been limited to the investigation of sustained vowel sounds or to non-speech activities. Further limitations are imposed not only by the cumbersome aspect of the equipment but also by the necessity for great subject cooperation. Camera techniques necessitate the use of subjects who are able to tolerate a laryngeal mirror in the pharynx and who are capable of full anterior-posterior exposure of the vocal folds. Two relatively new techniques are being employed in the investigation of laryngeal activity during connected speech. The use of a fiber optics system for photographic investigation is still in its infancy. Investigators (Sawashima and Hirose, 1968; Sawashima et a 1 . , 1969) have devised a fiber optics cable that may be inserted through the nasal passages A 16 mm camera is attached to the eyepiece of the image guide with illumination provided by a 150 watt incandescent lamp. Motion pictures with a frame rate of 24 per second are thus possible. Unfortunately, there is the distinct probability of subject discomfort. The fiber optics cable . placed in the nasal passage presents a diameter of approximately 6 mm. In addition, the illumination provided by the light source is not sufficient for ultra-high-speed exposure During the last 55 years, considerable interest has been directed to a new method for recording the movements

PAGE 22

9 of the larynx (Spencer, 1917) . The method has been termed glottography or laryngography . Researchers have proposed that through use of this procedure, direct examination of vocal fold movement in connected speech without undue subject discomfort is possible. Hartmann and Wullstein (1938) were the first investigators to make any extensive use of a photo-electric cell for sensing vibrations of the vocal folds. They associated the results of the glottogram with those of the phonogram using a double beam oscilloscope. The waves produced by the glottogram resembled a series of positive sinusoidal waves with the exception that the opening phase appeared to be more gradual than did the closing phase. More recently, improvements have been made on this photoelectric glottographic technique (Sonesson, 1959; Ohala, 1966, 1967; Coleman and Wendahl, 1968; Sawashima, 1969; Lisker et al . , 1969; Lindqvist, 1969; Frokjaer-Jensen, 1970) . Fabre (1957) and Decroix and Dujardin (1958) used an electronic process based on the principle of a high frequency current modulated by a low frequency current. The movements of the larynx of the subject, placed in an impedance bridge, were thus recorded in the shape of electrical variations. The primary objection to this method of

PAGE 23

10 analysis proved to be electrode placement and size. Fourcin and Abberton (1971) have modified the electrodes and their placement in their Laryngograph . The circuitry of this device is postulated to allow for self compensation for speaker impedance variation and is responsive only to fast changes resulting from vibration of the vocal folds in ordinary phonation rather than gross laryngeal displacement . An ultrasonographic method for studying the vibratory movements of the vocal folds has been developed by Minifie gjL-.al • (1967) . Hertz et al . (1970) found that the ultrasonic glottograms presented the same appearances as that obtained with the photo-glottographic method. The technique makes use of short ultrasound-pulses delivered to the surface of the neck. As echoes are received from the vocal folds, ultrasonograms are developed depicting the vibratory amplitude of the vocal folds. Although the gross appearance of movement has been found to be similar to that produced by the photo-glottographic method, the authors suggest that refinement of the equipment is still indicated before exactness of vibratory patterns can be expected. Lindqvist (1970) used an inverse filtering technique to obtain glottograms of vocal fold activity. Volume velocity waveforms were used to derive information.

PAGE 24

11 Calculations of the opening, closing, and closed phases were made. Investigators have made comparison between results derived from one type of glottograph to another type, and found that, while there is similarity of wave form configuration, there is no one-to-one relationship (Fant et_al., 1967; Koster and Smith, 1970; Frokjaer-Jensen, 1970) . Apparently, some variation of results may be due to registration change as well as subject subcutanous fat deposit differences in the throat. area. Phonograms and glottograms have been compared in an attempt to relate glottal area function to results depicted by the data derived from these methods (Fabre and Frei, 1959; Van Michel, 1966; Dolansky and Tjernlund, 1967; Vallancien and Faulhaber, 1967; Lebrun and Hasquin-Deleval , 1971) . Great differences in wave form have been noted. It has been concluded that some of the differences may result from the peculiar ability of the vocal folds to vibrate without producing voice. Photography has proved to be a reliable and valid means for obtaining information of glottal area function; therefore , it would appear reasonable to make comparison of wave forms derived by frame-by-frame analysis to those resulting from the use of glottographic techniques (Zemlin,

PAGE 25

12 1959) . Sonesson (1960) was one of the first investigators to make such a comparison. The comparison was made between glottograph wave forms and wave forms developed through measurement of the distance between the medial margins of the vocal folds. The results indicated that the open quotient developed through these methods was identical. He suggested that the phase difference between the lower and upper edges of the vocal folds could be represented by the closed interval of the glottogram. This would seem to indicate that amplitude of the curve should be evaluated. Unfortunately, the curves derived were not obtained Â’ through simultaneous recording of glottographic and photographic procedures . Great similarity of wave forms produced by the two techniques have been found by other investigators (Sawashima et al . , 1968; Van Michel et al . , 1970; Lindqvist and Lubker, 1970) . Unfortunately, the method of measurement remains unknown. Two investigations made by Wendahl and Coleman (1967, 1968) provoke some question of the validity of the information derived through glottographs . Their frame-by-frame analysis of glottal function derived by polar planimeter measurement produced information on the total area. These data were then compared to the curves simultaneously

PAGE 26

13 developed with the glottograph. The dissimilarity of the wave forms places the validity of the glottographic technique in some jeopardy. The measurements of glottal width rather than glottal area may be one factor influencing the disparity of fit noted in their glottogram data. Another factor that should be considered is the degree of vocal fold exposure. The subjects who participated in the study may not have attained the same degree of glottal area exposure . Statement of the Problem Previous studies of glottal area function have made use of methods that have either been established as capable of deriving valid physiological data or investigators have made assumptions of glottal activity based on data derived from equipment ordinarily utilized in acoustical investigations. To date, measurement procedures employed in the study of the vocal mechanism have not been specifically described and replicated in subsequent investigations. Further, adequate identification of the frequency, intensity, and register of the phonational sample has not been consistent . Based on the hypothesis that measurement of the quantity of light passing through the glottal area reveals

PAGE 27

14 the same information derived from measurement of the discernible space between the vocal folds, the major purpose of this study was to determine whether the method of photoelectric glottography is as valid for depicting glottal area function as a frame-by-frame analysis of ultra-highspeed film. A second purpose was to describe and compare the characteristic glottal configurations in falsetto, modal, and vocal fry registers of pre-adolescent, adolescent, post-adolescent, and adult subjects as revealed by the two methods of measuring glottal area function.

PAGE 28

CHAPTER II PROCEDURES The plan of this study was to obtain simultaneous indications of glottal area function as photographed by means of ultra-high-speed camera equipment and as sensed by the phototransistor of a photo-electric glottograph during the production of falsetto, modal, and vocal fry phonation. Five individuals (1 pre-adolescent male, 1 adolescent male, 1 P os t-adolescent male, 1 adult male, and 1 adult female) who possessed no evidence or history of voice disorders or laryngeal pathology served as subjects. The photographs and glottograms were obtained during sustained phonation of the vowel /i/ which provides the most complete exposure of the anterior-posterior aspect of the vocal folds. The samples were produced at medium vocal effort for each of the registers . The vocal registration was determined by agreement between the subject and the investigator. Equipment A block diagram of the instrumentation employed in the present study is shown in Figure 1. The subject sat on an 15

PAGE 29

16 Figure 1. Block diagram of equipment used to obtain the voice signals and simultaneous photographic and glottographic records of glottal area functions for each subject

PAGE 30

17 adjustable stool before a fixed laryngeal mirror and introduced the mirror and the probe microphone into his pharynx. The phototransducer was affixed to the neck of the subject in the space between the anterior-inferior boundary of the thyroid cartilage and the anterior-superior boundary of the cricoid cartilage. The neck was then draped with black plastic material to block extraneous light sources and thus achieve maximum sensitivity of the transducer. Following amplification of the voice source by a condenser amplifier and of the glottal area function by the glottograph and external amplifier, each of the two signals were displayed on the face of an oscilloscope. As the subject achieved full exposure of the vocal folds for each register production, ultra-high-speed photographs were made of the vocal folds and of the product of the voice source and glottal area function as displayed on the face of the oscilloscope. For monitoring phonation, the voice signal was recorded on an Ampex 354 full-track tape recorder. This signal was subsequently subjected to fundamental frequency analysis on an oscillographic writer and to intensity analysis by a graphic level recorder. Instrumentation for obtaining ultra high-speed photographs Standard Fastax ultra -high-speed photographic procedures

PAGE 31

18 were used to obtain simultaneous photographs of the activity of the vocal folds and the photocell monitoring device recorded on one channel of an oscilloscope. This activity was filmed at an exposure rate of approximately 5000 frames per second . Specifically, the photographic equipment consisted of a model WF 301 Wollensak Goose control unit set in position one to allow the event and the camera to start at the same time. The event is defined as the desired exposure of the laryngeal area. At the end of the time cycle, the event and the camera stopped. A timing impulse control unit made up of a time delay, electro-larynx drive, electro-larynx and a flash/sync generator were used to drive a Heathkit square wave generator thus producing a 5 kc square wave and flashing a time marker for the film. The 5 kc square wave was displayed on one channel of a Douglas 531 four channel oscilloscope. The side lens of the Fastax camera was focused on the face of the oscilloscope, thereby photographing the wave form generated by the 5 kc square wave as well as those produced by the input of the photo-electric glottograph and the condenser microphone. The Fastax camera was fitted with an extension bellows and a 152 mm telephoto lens. The F stop was set at 3.5. The camera was loaded with Kodak 4 XR film #7277 with double perforations of 0.3000 inch pitch.

PAGE 32

19 Illumination for filming and for glottography was provided by a 2100 watt, 60 volt General Electric projector lamp. Cooled light was directed by lenses and a mirror to a laryngeal mirror in the pharynx from which it was reflected into the larynx. The light that passed through the glottal opening between the vocal folds struck the anterior laryngeal v/all subglottally and at the level of the phototransducer located on the external surface of the neck. Instrumentation for obtaining phototransistor glottographs Glottographs of glottal area function were obtained by a phototransducer mounted in a polyethylene tube with an internal diameter of 2 mm and an external diameter of 3 mm. The transducer was connected to a DC amplifier housed in a Frok jaer-Jensen Photo-electric Glottograph, type LG900, by means of a short shielded cable. This, in turn, was connected to a five times operational amplifier. In total darkness, the phototransducer produces an input voltage to the amplifier of 0 . 7 /U V. In position over a closed glottis the input voltage is 3 . 7 ^uV. A fully opened glottis yields an input voltage of 6 mV. The input level varies only up to 3 mV under conditions of normal phonation. Amplification of the glottography system is linear within plus or minus 2 percent in the normal working range of 0-3 mV

PAGE 33

20 input voltage. The power gain is 80 dB maximum. By means of a Philips Stroboscope type 9103, the frequency range of the glottograph was tested and found to be from DC-10,000 Hz. The signal from glottograph output A with a 4 volts/67 mA maximum in a 60 ohm load was input to a five times operational amplifier and thence to one channel of the Douglas 531 four channel oscilloscope. The output voltage selector was set at maximum. In the present study, the phototransistor was affixed with adhesive tape to the neck of the subject at a point rostral to the superior boundary of the anterior aspect of the cricoid cartilage. The upper and lower length of the transistor was indicated by a line drawn on the neck of the subject. This procedure was used to afford a position monitor for the transistor at the beginning of each film sequence. The neck was then draped with black plastic material to exclude all extraneous light. Measurement devices To provide a base line response from the glottograph and a millimeter scale for glottal area measurement, a modification of the procedure developed by Hollien and Moore (1960) was employed. Immediately following the photographic procedure for each register and without any adjustment of the lens, a millimeter scale was photographed. This technique

PAGE 34

21 permitted measurement not only of vocal fold function but also served as a "no response" line for measuring glottograph response. Experimental Procedure Subjects Five individuals were selected for the investigation. The subjects volunteered to perform the tasks and were chosen on the basis of the following criteria: (1) they showed no evidence or history of voice disorders or laryngeal pathology; (2) they were capable of complete anterior-posterior exposure of the vocal folds; (3) they were capable of phonating in vocal fry, modal, and falsetto registers; and (4) they presented five types of laryngeal development (pre-adolescent, adolescent, post-adolescent, adult male, and adult female). Training Procedure To provide an experimental environment for subject task practice, a training station was devised. This consisted of a fixed laryngeal mirror, a light source,, and a subject viewing apparatus. The light source consisted of a high intensity lamp with the beam directed onto an ophthalmoscope mirror. The reflected light of the ophthalmoscope mirror was adjusted to illuminate the fixed laryngeal mirror. The subject viewing apparatus included a transparent mirror placed at. a 45° angle which directed the image of the vocal folds to a mirror 20 inches above the base of the station.

PAGE 35

22 This mirror was inclined at a variable angle, thus allowing the subject to monitor vocal fold activity and duplicating the experimental environment. The back of the station support box was equipped with a two-way mirror to permit viewing of the larynx of the subject by the investigator. A three-dimensional drawing of the station may be seen in Figure 2 . Tasks After the investigator affixed the phototransducer for recording glottograph traces, the subject sat on an adjustable stool before a fixed laryngeal mirror and introduced the laryngeal mirror and probe microphone into his pharynx. Each subject was then instructed to produce a sample of the vowel /i/ in the register under investigation. The investigator observed the image of the subject's vocal folds in the viewing mechanism of the camera and checked the accuracy of the phonation range. The traces of the output of the microphone and the glottograph were observed for adequate illumination and placement on the face of the oscilloscope. When a full exposure of the vocal folds was visualized the investigator indicated that the subject should sustain the phonatory posture and the controls for the camera and the electrolarynx were triggered by the remote switch. Thus,

PAGE 36

Figure 2. Three-dimensional drawing of subject training station

PAGE 37

24 * simultaneous indications of laryngeal activity as viewed through the camera, sensed by the phototransducer, and recorded by the microphone, were photographed in each of three registers. After each phonation was filmed, the millimeter grid was brought into focus and photographed as the subject viewed the adducted vocal folds with a hand-held laryngeal mirror. Appendix A contains the recording itinerary followed throughout. Falsetto Range Following the experimental tasks, verification of the production of falsetto register was confirmed for each subject. Four phoneticians and the investigator listened to the production of the complete vocal range of each subject. The experimental sample was then reproduced immediately after the vocal range was presented. Agreement between the judges was found to be unanimous for each sample. The judgment form is reproduced in Appendix B. Data Analysis Glottal area function was determined by traditional polar plan.imeter methods with total area of glottal opening in successive film frames plotted as a function of time. This necessitated measuring the activity of the vocal folds for each frame in a cycle three times and dividing by three

PAGE 38

25 to obtain a high degree of accuracy. The use of a Kodak Photo-Optical Data Analyzer Model 224 allowed for stopframe study of each frame for each segment of the film used in the study. Measurement techniques are outlined in Appendix C. The output of the phototransducer for the glottal cycle was analyzed by measuring the deviation of the oscilloscope trace of the glottograph from the base line established by the 5k Hz signal. The drafting instruments developed and employed for this measurement procedure are described in Appendix C. Amplitude and area curves were developed as a function of time and plotted for each experimental condition. Measurements were made on six complete cycles selected from the 50to 60-foot portions on the 100-foot films. The curves developed through analysis of the data are presented in Appendix D. Frequency of phonation was derived for each subject in each experimental condition by means of input from the Ampex tape recorder to the Honeywell model 1508A Visicorder. Waveto-wave measurement of the output of each sample was made by measuring the wave from peak-to-peak and dividing this by the time interval selector of the Visicorder to obtain the period. Frequency in Hz was computed by the formula F = 1/T.

PAGE 39

26 A summary table of frequency of phonation for each individual during each experimental condition may be seen in Appendix E. The relative intensity levels were computed for each phonational sample. The output of the Ampex 352 dual channel, full-track tape recorder was coupled to a Bruel and Kjaer 2305 Graphic Level Recorder which recorded the sample signal at a mid-scale with a 50-dB potentiometer. The measures were made relative to this scale. These results are summarized in Table 6 in Appendix F. To determine the relationship between the wave forms developed by measurement of the glottal area function derived through analysis of the photographic film and by measurement of the deviation of the quantity of light sensed by the photo-electric glottograph, the curves were normalized and a Pearson product-moment correlation coefficient was computed for each sample for each subject. In addition, confidence intervals were developed for each computation. To determine the degree of approximation to glottal area function produced by the data derived from the photo-electric glottograph within three registers, a comparison was also made between the correlation coefficients developed for the samples obtained from each subject.

PAGE 40

CHAPTER III RESULTS The major purpose of this study was to determine whether photo-electric glottography was as valid for depicting glottal area function as frame-by frame analysis of ultra-high-speed film. The two investigative techniques were compared in three vocal registers (vocal fry, modal, and falsetto) produced by five normal larynges representing different ages and sexes (pre-adolescent male, adolescent male, post-adolescent male, adult male, and adult female). The raw data were acquired by recording glottal area vibrational variations simultaneously in high-speed motion pictures and in photo-electric glottogram oscillograph traces. The five subjects phonated the vowel /i/ at two fundamental frequency regions, one in each of two registers, modal and falsetto. In vocal fry registration, it was possible to obtain only two satisfactory samples of subjects phonating the vowel /i/ at appropriate fundamental frequencies. The reader should be aware of the difficulty of vocal fold exposure in this register (Timcke et al . , 1959). 27

PAGE 41

28 One sample of each experimental condition was obtained, the criterion measures were register production, vocal fold exposure, relative intensity, and phototransducer activity for each sample in each condition. It should be noted that the glottal area function and the fundamental frequency of phonation obtained in the present study varied from subject to subject as well as from register to register. For example. Subjects One and Three were found to demonstrate incomplete glottal approximation in modal register, whereas, the other subjects did achieve complete vocal fold closure in this register. This manner of approximation in modal register is in disagreement with that found by other investigators (van den Berg, 1958, 1960; Vennard, 1960; Moore, 1968a) . Further, the frequency of phonation in each of the registers differed from subject to subject. However, in each sample of register production the frequency produced by the individual subject was appropriate for the age and sex of that subject (Luchsinger and Arnold, 1965; Sedlackova, 1961; van Oordt and Drost, 1963). To identify the developmental stage and sex of each of the subjects, the following designations were used: Subject One, adult male; Subject Two, adult female; Subject Three, post-adolescent male; Subject Four, adolescent male; and Subject Five, pre-adolescent male.

PAGE 42

29 Correlation of Glottal Area Function Depicted by Photo-electric Glottographs and UltraHigh-Speed Pictures Modal register The measurements of glottal area function derived from the frame-by-frame analysis were normalized for each of the subjects. These data were then compared to the normalized data derived simultaneously from the signal transmitted by the phototransducer of the photo-electric glottograph. The amplitude and area curves thus obtained for each of the five subjects producing modal register phonation are shown in Figures 3-7. These figures and Table 1 indicate that the phototransducer output and glottal area function measurements correlate significantly at the .05 level of confidence. In view of the variability exhibited by the subjects during the production of the modal register frequencies, no definitive statement is possible about the ratio of the opening, closing, and closed phases of the vibratory cycle. It should be noted, however, that there appear to be several modes of vibration appropriate to this register. Similarity of closure was noted between Subjects One and Three. Figures 3 and 4 clearly indicate that the frameby-frame analysis and the glottogram show no complete closure for either of the two subjects. In addition, the two methods define the open quotient and speed quotient

PAGE 43

30 f Q 210 Hz Intensity 32 dB re 10 mV Photographic Glottographic Figure 3. Plot of glottal area function and photo-electric glottograph for one normalized cycle. Subject 1, modal register .

PAGE 44

31 equally well.^ The vibratory cycle characteristic of Subject One began with a lateral movement of the upper margins of the vocal folds. Discernible opening began just anterior to the middle one-third of the vocal folds and proceeded toward the anterior and posterior commissures. As the upper margins completed their lateral displacements, the lower margins could be seen approaching the mid-line at the posterior one-third of the vocal folds. This closing movement proceeded synchronously to .the anterior commissure. The upper margins began their adductory movement at the middle one-third and the closing gesture continued to the anterior commissure. As shown in Table 2, the open quotient was found to be .88, while the speed quotient was 1.0. The opening phase noted in the vibratory cycle produced by Subject Three was found to show incomplete closure .as indicated in Figure 4. Initial opening began at the posterior commissure. The phase relationship of the upper and lower margins of the vocal folds presented an appearance ^Open Quotient will be defined by the formula: Fraction of cycle during which glottis is open vj # v/ . Duration of entire cycle Speed Quotient will be defined by the formula: S Q Time of abduction or lateral excursion Time of adduction of medial excursion (Timcke et al . , 1959).

PAGE 45

PERCENT OF OPENING 32 f G 180 Hz Intensity 37 dB re 10 mV Photographic GXottographic Figure 4. Plot of glottal area function and photo-electric glottograph for one normalized cycle. Subject 3, modal register .

PAGE 46

33 similar to that of Subject One. During the opening phase, the only difference noted between the two was that the lower margins of the vocal folds were not visible. On closing, the upper margins appeared to move laterally while the lower lip of the vocal folds presented a synchronous closing movement originating at the posterior commissure. As this adductive movement was completed, the upper lip moved to the mid-line. During the closed phase, the mass of the thyroarytenoid muscles seemed to compress toward the midline . This movement continued into the beginning of the following opening gesture. A stiffening 2 of the edge of the vocal folds then appeared to occur. The open quotient was found to be .61 with a speed quotient of 2.0 as shown in Table 2. These figures were acquired through analysis of the averaged data derived by the two methods of investigation . The vibratory pattern characteristic of Subject Two was found to employ a movement pattern more similar to that described by Timcke et al . (1958) , in their discussion of the vocal fold margins. The opening commenced at the 2 Stiffening will be operationally defined as the appearance of a compact, dense, and fixed positioning of the vocalis musculature (van den Berg, 1958) .

PAGE 47

34 posterior commissure with the upper margins of the vocal folds separating first. This opening then seemed to progress to the middle one-third. The main body of the thyroarytenoid musculature appeared to move laterally away from the mid-line. Complete abduction to the anterior commissure was then achieved exposing the lower margins of the folds which seemed to be similar in contour but of smaller total area than that encompassed by the upper margins. This activity was followed by complete opening of the lower margins of the vocal folds.On closing, both vocal folds appeared to move synchronously toward the mid-line where they contacted each other simultaneously throughout their entire borders. During the closed phase, movement of the thyroarytenoid muscles toward the mid-line was observed. Figure 5 and Table 2 allow a definition of the open quotient to be .70, while the speed quotient may be calculated to be 1.75 as analyzed by an average of the two methods. The speed quotient suggests that the intensity of phonation was medium to loud (Timcke et al . , 1958) which was confirmed by reference to the recorded sound. The action noted in the vibratory pattern for Subject Four was found to be grossly similar to that noted for Subject Two. This similarity may be noted by a comparison of Figures 5 and 6. Moreover, the stiffening posture of the

PAGE 48

35 f 240 Hz Intensity 38 dB re 10 mV Photographic Glottographic Figure 5. Plot of glottal area function and photo-electric glottograph for one normalized cycle. Subject 2, modal register .

PAGE 49

36 f Q 170 Hz Intensity 35 dB re 10 mV Photographic Glottographic Figure 6. Plot of glottal area function and photo-electric glottograph for one normalized cycle. Subject 4, modal register .

PAGE 50

37 vocal folds noted in the vibratory pattern of Subject Three was apparent during the opening phase for Subject Four. Of interest to the investigator was the initial opening gesture in which a massing of the upper margins of the vocal folds seemed to continue for several frames into the opening phase. However, as the closing phase began, the vertical expanse of the marginal surfaces of the vocal folds presented a somewhat concave appearance. Again, the initial adductory movement of the folds began at the posterior commissure. For this subject, analysis of the averaged data obtained from the phototransducer and the photographic film revealed the open quotient to be .57 with a speed quotient of 1.40 as shown in Figure 6 and Table 2. The activity of the vocal folds of Subject Five differed somewhat from that of the other subjects. Specifically, during the closing gesture, the lower margins of the vocal folds approached initially at the posterior commissure. Although this gesture was not completed, the anterior portion of the folds did then adduct. Contrary to other observations made during the production of modal register, the upper margins of the folds began the adductory movement ^t the anterior commissure. Adduction was then characterized by symmetrical, progressive, medial movement of the vocal folds terminating in complete glottal closure. Some lateral

PAGE 51

38 movement of the thyroarytenoid muscle mass was noted. Figure 7 and Table 2 illustrate that the two methods of investigation indicated an open quotient of 1.0 with a speed quotient of 1.5 for this subject. To determine the relationship between the two methods used in determining glottal area function, a Pearson productmoment correlation coefficient (Hays, 1963) and 95 percent confidence intervals (Walker and Lev, 1953) for each correlation were calculated. The results of this analysis are presented in Table 1. As may be seen from inspection of the table, the correlation between the photographic and glottographic methods for evaluating glottal area function in the modal register is significant. Falsetto register The results of analysis of the normalized amplitude and area curves derived through measurement of the glottal area function from the developed photographic film and the traces representing the deviations of the quantity of light sensed by the phototransducer for the five subjects producing phonational frequencies within their falsetto register range are shown in Figures 8-12. A significant correlation was found to exist between the results of the analysis obtained from the two methods of measurement.

PAGE 52

39 f 360 Hz Intensity 43 dB re 10 mV Photographic Glottographic Figure 7. Plot of glottal area function and photo-electric glottograph for one normalized cycle. Subject 5, modal register .

PAGE 53

Table 1. Summary table of correlation between photo-electric glottographic and photographic methods of analysis for modal register 40 0) I — I rtj > QJ P C H CD O C >
PAGE 54

PERCENT OF OPENING 41 f Q 490 Hz Intensity 40 dB re 10 mV Photographic Glottographic Figure 8. Plot of glottal area function and photo-electric glottograph for one normalized cycle. Subject 1, falsetto register.

PAGE 55

PERCENT OF OPENING 42 f 680 Hz o Intensity 25 dB re 10 mV Photographic Glottographic Figure 9. Plot of glottal area function and photo-electric glottograph for one normalized cycle. Subject 2, falsetto register.

PAGE 56

PERCENT OF OPENING 43 f 300 Hz o Intensity 36 dB re 10 mV Photographic Glottographic Figure 10. Plot of glottal area function and photo-electric glottograph for one normalized cycle. Subject 3, falsetto register .

PAGE 57

PERCENT OF OPENING 44 f 580 Hz Intensity 30 dB re 10 mV Photographic Glottographic Figure 11. Plot of glottal area function and photo-electric glottograph for one normalized cycle. Subject 4, falsetto register .

PAGE 58

45 f 760 Hz o Intensity 42 dB re 10 mV Photographic Glottographic Figure 12. Plot of glottal area function and photo-electric glottograph for one normalized cycle. Subject 5, falsetto register .

PAGE 59

46 There was great variability in the frequency of phonation produced by the subjects during the execution of this register. In addition, there appeared to be three modes of vibration exhibited by the subjects of this study. Subjects One, Two, and Three seemed to present similar modes of vibration. There was no complete closure of the vocal folds during the "closed" phase. For these individuals, an opening at the posterior commissure was apparent. This is similar to the type of vibrational pattern described by Rubin and Hirt (1960) as open chink. For these subjects, opening was initiated at the middle one-third and was rapidly continued to the anterior commissure. On closing, the folds approached first at the anterior commissure and completed their closing gesture with contact of the middle one-third of the vocal folds. The figures associated with the glottal area function for each of the subjects reveal that the open quotient and the speed quotient were both found to be 1.0 (Table 2). This suggests that the intensity of production in this register was medium to high and that there was considerable tension of the vocal folds as predicted by Timcke et al . (1958) and Rubin and Hirt (1960). The vibratory pattern for Subject Four differs from that previously described in the production of falsetto register in that, during the opening phase, the folds

PAGE 60

47 were seen to approximate at the middle one-third with an opening present at the anterior and posterior one-thirds of the vocal folds. Complete opening was then achieved with an apparent stiffening of the margins of the folds. Closure was accomplished in much the same manner as that of the aforementioned subjects. An inspection of Figure 11 and Table 2 reveals an open quotient of 1.0 with a speed quotient of .75. A most unusual pattern of vibration was noted in the phonatory mechanism for Subject Five. Although no closed phase was accomplished, the opening first perceived was at the middle one-third of the vocal fold expanse rather than at the posterior portion as had been noted in the previous patterns. In addition, the opening phase showed an abduction of the anterior portion extending to the anterior commissure. On closing, the posterior one-third made adductory movements followed by adduction of the anterior portion of the vocal folds. Again, there appeared to be stiffening of the margins of the vocal folds. However, this process seemed to progress segmentally throughout the length of the vibrating vocal folds. It is evident from the results displayed in Figure 12 and Table 2 that the opening quotient for this subject was 1.0 with a speed quotient of .66.

PAGE 61

48 A Pearson product-moment correlation coefficient and 95 percent confidence intervals for each correlation were calculated. The results depicted in Table 3 indicate that the correlation between the two methods of investigation is significant for the falsetto register. This finding is in agreement with the suggestion made by Foster and Smith (1970) . The glottograms derived in falsetto register are more representative of glottal area function than in registers encompassing the lower phonational ranges. Vocal fry register For vocal fry register, the resulting normalized amplitude and area curves produced by two subjects are shown in Figures 13 and 14. Although the five subjects participating in the investigation were capable of producing vocal fry phonation, for only two of the subjects were measurable film images obtained. Analysis of the measurement of the glottal area function in the frame-by-frame display of the developed film and the deviation of the trace produced by the sensitivity of the phototransducer revealed that the correlation between the two methods of investigation is not as exact in vocal fry register as in the modal and falsetto

PAGE 62

Table 2. Summary table of averaged indications of open quotient and speed quotient for each subject in modal and falsetto register phonation derived through analysis of photographic film and photo-electric glottograms* 49 u 1 P 44 03 CJ o > p o 4h -p g
PAGE 63

e 3. Summary table of correlation between photo-electric glottographic photographic methods of analysis for falsetto register 50 m iH (0 > P cn Ci 0) • • • • • a n, t>
PAGE 64

51 f 10 Hz o Intensity 37 dB re 10 mV Photographic Glottographic FRAMES Figure 13. Plot of glottal area function and photo-electric glottograph for one normalized cycle. Subject 3, vocal fry register .

PAGE 65

PERCENT OF OPENING 52 f 20 Hz Intensity 34 dB re 10 mV Photographic Glottographic Figure 14. Plot of glottal area function and photo-electric glottograph for one normalized cycle. Subject 4, vocal fry register .

PAGE 66

53 registers. However, it should be noted that a striking similarity exists between the derived curves. The vibratory pattern displayed in Figure 13 in the production of vocal fry was regularly biphasic for Subject Three. Production of a multiphasic vibratory cycle is in agreement with the results of earlier investigations (Timcke et al . , 1959; Moore and von Leden, 1958). The initial opening excursion was seen at the anterior commissure followed slowly by opening to the middle one-third of the vocal fold margin. There was an adductory motion leaving a narrow eliptical opening to the posterior commissure. This posture was maintained until the second excursion of the folds appeared. A lateral motion was visualized, followed by complete opening of the folds. Closing was characterized by a rapid closure of the anterior portion of the folds followed by complete anterior-posterior closure. The closed phase allowed for visualization of the participation of the ventricular folds. It should be noted that in no., instance was the activity of the vocal folds obscured by the participation of the ventricular folds. As the margins of the folds closed, there appeared to be massing of the marginal area. The ventricular folds gradually receded into a completely abducted position. This observation is in agreement with the results reported in a recent laminagraphic

PAGE 67

54 investigation (Allen and Hollien, in press) which suggested that the ventricular folds operate in conjunction with the vocal folds in vocal fry phonation and, further, that the travelling wave from the true folds is continued into this mass . The time ratio and area of the two opening gestures noted in the single vibratory cycle are of interest. The initial excursion occupied twice the time period noted for the second while the area measurement indicated an increase of 25 percent for the second excursion. The closed period accounted for approximately two-thirds of the entire period. For Subject Four, the production of vocal fry registration was characterized by a vibratory cycle demonstrating a triple opening sequence and an occasional vibrational cycle presenting a single opening pattern. The initial excursion was evidenced by a small circular opening at the posterior commissure followed by a narrowing of this area. The second excursion demonstrated a triangular configuration; the smaller opening appearing at the posterior commissure. This opening was then modified to that of a narrow ellipse. The final abduction apparently involved a greater length of the folds and appeared to be symmetrically elliptical in shape. A rapid closure was then accomplished. The single opening pattern was seen only on occasion and may not have been

PAGE 68

55 representative of the subject's normal phonatory posturing in vocal fry phonation. More participation of the ventricular folds was noted in the production of this register by Subject Four than by Subject Three. It should be noted also that for Subject Four, the vocal folds seemed to be covered by a considerable amount of mucus. The action of the ventricular folds was similar to that of Subject Three. Although protocols for defining the open quotient and speed quotient for a single cycle have not been established for vocal fry register, it is of interest to note the quotients for the excursions of the folds visualized in one vibratory cycle. In the photographic film analysis, Subject Four displayed an open quotient for the first excursion of 1.0 with a speed quotient of 3.43. The second excursion presented an open quotient of 1.0 with a speed quotient of 1.09; whereas the third abductory posture indicated an open quotient of 1.0 and a speed quotient of .60. The closed phase for the entire cycle occupied about one-third of the entire eye le . The results of computation of a Pearson-r between the two methods of investigation, together with 95 percent confidence intervals for each correlation, are shown in Table 4. Indications are that, although correlation does

PAGE 69

e 4. Summary table of correlation between photo-electric glottographic photographic methods of analysis for vocal fry register 56 -P •H e •H pi p w a) r— I Qj ns cu > m u cu P a H CD in CTi (T, CD • • ai CJ a (D T3 •H +J 4H -H d e O -rH U PI S-l Cl) S O PI n in ID C4-I ms CD CD CN t'' I — I b ms ro c EH CIS CU > (U u * * in ^ ID CD •p o as •n b P CD m ^1* in o cu b p p ns -P G ns o -p MH •P c Cn •H CD *

PAGE 70

57 exist, the phototransducer is not capable of sensing the same information as that derived visually from photographic film. For the five subjects used in the present investigation, it may be concluded that the photo-electric glottograph is as valid a method for depicting glottal area function as a frame-by-frame analysis of ultra -high-speed film in the modal and falsetto registers. In addition, although the correlation found between the two methods of investigation is not as high in the vocal fry register, the photo-electric glottograph does appear to be capable of generating reasonably approximate information.

PAGE 71

CHAPTER IV DISCUSSION The subjects in the present investigation utilized essentially normal laryngeal mechanisms (as determined by the subject selection criteria) to produce samples of vocal fry, modal, and falsetto phonation which were analyzed through measurements obtained from a light sensing device and a high-speed camera. The results indicate that the two methods of analysis are highly correlated in their ability to depict similarly shaped area and amplitude curves for the modal and falsetto registers. In addition, although the statistical analysis of the correlation between the curves developed by the two methods in vocal fry phonation did not indicate a one-to-one relation, there was striking similarity in the overall configuration. Glottoqraphic Response and Vibratory Pattern in Modal Register Prior to this study, few studies had been made employing simultaneous recordings of vocal fold activity through the utilization of glottography and photography. Moreover, 58

PAGE 72

59 the investigators who did obtain such simultaneous recordings (Coleman and Wendahl, 1967, 1968) found that the photoglottograph was limited in providing accurate indications of glottal area function. The present study disagrees with those findings in the modal register. Although it was apparent that the phototransducer signaled the opening phase earlier than the occurrence of this phase became visible, this is not in disagreement with the suggestion of Sonesson (1960) . He surmised that the glottograph could be useful in defining the phase relationship of the lower and upper edges of the vocal folds. It is possible that such complex glottal configurations can cause the differences noted in the results of this study. Timcke et al . (1958) found that, although the shape of the amplitude curves produced by the upper and lower margins of the vocal folds was similar, there was a definite phase shift between the two movements. The upper margin was seen to lag behind the lower margin in the closing phase of the vibratory cycle. This finding would seem to be verified by the results of this study. Variations in the vibratory patterns produced by the subjects of this study suggest that in the modal register there is no completely predictable mode of vibration. Brackett (1947) found not only differences between subjects but also intra-individual differences. A careful comparison

PAGE 73

60 of individual vocal fold movement and vocal intensity confirmed the hypothesis set forth by van den Berg et al . (1957) that the Bernoulli effect, i.e., the negative pressure in the glottal air, is enhanced with an increase in tone intensity and air flow thereby drawing the vocal folds more forcibly toward each other. Observations also supported the myoelastic-aerodynamic theory of phonation. The vocal fold margins of two of the subjects participating in the study demonstrated an active longitudinal tensing. van den Berg (1968a) reports that this contraction is expected in sustained phonation of a given fundamental frequency. The concave appearance of the vocal fold margins visualized in the phonation of one subject suggests an active participation of the vocal muscles. Glottographic Response and Vibratory Pattern in Falsetto Register The correlation between the measurements of glottal area produced by photo-electric glottographs and by the photographic film in falsetto register found in this study tend to conform to the conclusions stated by Koster and Smith (1970) . They found that in phonations in the higher frequencies the glottograph provided a more accurate representation of glottal area function than in the lower frequencies. The present results indicate that the

PAGE 74

61 glottographs are significantly related to glottal area function in the falsetto register. It may be surmised that the thin shape of the glottal margins creates single, relatively simple glottal folds in which there is no change of phase relationship between an upper and lower margin. Thus, it could be expected that the light passing between the vocal folds would not be impeded by vocal fold tissue during the opening and closing phases of the vibratory cycle. Although Rubin and Hirt (1960) indicated three basic patterns of vibratory activity in the falsetto, only that designated as open chink was observed in this investigation. However, two subjects presented vibratory patterns that have not previously been described. A narrow opening at the posterior and anterior commissures with closure at the middle one-third of the vocal fold expanse characterized the vibratory pattern for one of the subjects. Contrary to this mode of vibration, another of the subjects achieved closure that may be described as "tight" at the posterior and anterior commissure. The opening thus accomplished only encompassed the middle onethird of the vocal folds. In all cases, the appearance of the vocal folds did correspond with that suggested by other investigators (van den Berg, 1960; Luchsinger and Arnold, 1965). The vocal

PAGE 75

62 folds were indeed elongated and appeared to be thin in mass. There also seemed to be an obvious stiffening of the vocal fold margins which suggests a longitudinal tension in the vocal ligaments without an active vibratory participation of the vocal muscles. However, the visualization of adduction of the middle one-third of the vocal folds suggests that myoelastic properties may have been active in at least this individual's production of falsetto register. The appearance of tight closure of the anterior and posterior commissures with resulting open chink at the middle one-third may possibly be explained by myoelastic properties. It was obvious that aerodynamic properties were predominant in the falsetto production of the remaining subjects. Glottographic Response and Vibratory Pattern in Vocal Fry Register Previous investigators of the vibratory pattern for vocal fry phonation have obtained photographic film of phonation that should be considered as high fry or fry mixed with modal (Timcke et al . , 1959; Moore and von Leden, 1958) . The samples obtained in this investigation were of fundamental frequencies more appropriate to this register (Allen and Hollien, in press; Hollien et al ., 1966; Hollien and Michel, 1968; Hollien et al ., 1969). Although the results for this register are based on the production of only two subjects, it was concluded that this information

PAGE 76

63 was valid. The magnitude of the correlations indicated inexact replication of .vocal fold activity by the photoelectric glottograph as compared with the analysis of photographic film. However, the curves representing the variations of quantity of light and the glottal area function do seem to bear a striking similarity. Information on the physiologic parameters of vocal fry phonation have not been completed. However, the glottograms resulting from this study resemble those analyzed by Ohala (1966) and Fourcin and Abberton (1971) . These investigators termed the vocal production as creaky voice. The vibratory pattern exhibited by one of the subjects was like that defined by Moore and von Leden (1958) as dicrotic dysphonia. The biphasic excursions of the vocal folds were found to represent amplitudes similar to those suggested by those investigators, i.e., the first was approximately 25 percent smaller in amplitude than the second . The vibratory pattern characteristic of the fourth subject was unusual in that it appeared to represent three definite excursions of the vocal folds. Each of the excursions seemed to be about 25 percent larger than the preceding one. It has been suggested (Hollien et al . , 1966) that there may exist various modes of vibration for this register.

PAGE 77

64 Hollien (1971) has postulated that myoelastic properties are predominant in vocal fry register. Visualization of the laryngeal mechanism reveals that this must indeed be the case. However, due to the complexity of relationships of the structures involved in the production of this register, much more extensive investigation of the vocal phenomenon is indicated before aerodynamic properties can be dismissed as contributory to the control of change of frequency .

PAGE 78

CHAPTER V SUMMARY AND CONCLUSIONS The purpose of this investigation was to determine the validity of the information concerning vocal fold activity derived through measuring the quantity of light passing through the glottis by comparing .these data with measurements of glottal area taken from the photographic images . A second purpose was to describe and compare the characteristic glottal configurations in falsetto, modal, and vocal fry registers of pre-adolescent, adolescent, post-adolescent, and adult subjects as revealed by the two methods of measuring glottal area function. Based on the hypothesis that measurement of the quantity of light passing through the glottal area should reveal the same information as that derived from measurement of the discernible space between the vocal folds, simultaneous recordings on photographic film were obtained of the image of the vocal folds, the voice signal, and the deviations sensed by the phototransducer during vocal fry, modal, and falsetto phonation. Recordings were made during sustained phonation of the vowel 65

PAGE 79

66 /i/ at frequencies appropriate to the vocal fry, modal, and falsetto registers. Deviations of the quantity of light were sensed by a phototransducer attached to the space between the thyroid and cricoid cartilages. The fundamental frequency and intensity of phonation were recorded on magnetic tape from the acoustic signal input to a probe tube microphone inserted into the oral cavity of the subject. The visual image of the vocal fold activity was photographed from the fixed laryngeal mirror introduced into the pharynx. The voice signal, deviations of the light quantity and a 5k Hz monitoring signal were displayed on the face of an oscilloscope and photographed simultaneously with the mirrored laryngeal activity by an ultra-high-speed camera. The voice signal recorded on the tape recorder was subsequently submitted to fundamental frequency and intensity analysis. Glottal area measurements were made from the photographic film by polar planimeter methods and by measuring the deviations of the light quantity. The data were normalized and submitted to statistical analysis in order to determine the correlation between the glottal area curves depicted by the photo-electric glottograph and the frame-by-frame analysis of the visual image in each of three registers. From the investigation the following findings were obtained :

PAGE 80

67 1. The measurements of the light deviations sensed by the photo-electric glottograph during laryngeal vibrations correlate significantly at the .05 level of significance with the photographic film frameby-frame measurements of the visual image in modal and falsetto register. 2. The glottograms resulting from measurement of vocal fry phonation approximate those derived from photographic film measurement, however the correlation between the two curves is not as high in this register as is the case in modal and falsetto registers. 3. Inter-individual and intra-individual differences of vibratory patterns of phonation were found in vocal fry phonation. 4. During modal range phonation, there appeared to be no established vibratory pattern for all subjects. 5. Aerodynamic properties may not completely account for the vocal fold activity observed in the subjects of this study in the production of falsetto register . It appears, therefore, that the photo-electric glottograph is a valid method of determining glottal area function. This conclusion, which is also supported by data obtained using a photo-electric glottograph and separate photographic film (Sonesson, 1960) , may result in more

PAGE 81

68 complete understanding of the complex relationship of the upper and lower margins of the vocal folds. It might also result in a correlation between acoustically and physiologically derived data, i.e., through an inverse filtering process it may be possible to separate from the acoustically derived data valid information about the glottal area function . The presence of at least two examples of variations from the vibratory pattern expected in falsetto register suggests a need for collecting a large and representative sample for obtaining normal values of the vibratory pattern of the vocal folds. In addition, the contribution of myoelastic properties should be examined in this register. The subjects participating in this investigation represented a wide age differential and it was found to be possible to obtain laryngeal photographs of the preadolescent, adolescent, and post-adolescent aged individuals without undue difficulty. Therefore, this suggests that photographic investigation of the phonatory mechanism at different stages of maturation may result in more adequate understanding of the phenomenon of voice change.

PAGE 82

APPENDICES

PAGE 83

APPENDIX A PHOTOGRAPHIC SEQUENCE

PAGE 84

71 Photographic Sequence Followed Throughout the Experiment Phototransducer attached to space between thyroid and cricoid cartilages . 1. Vocal fry register /i/ 2. Millimeter grid photographed as the phototransducer transmits a "no response" signal to the oscilloscope. 3. Modal register /i/ 4. Millimeter grid photographed as the phototransducer transmits a "no response" signal to the oscilloscope. 5. Falsetto register /i/ 6. Millimeter grid photographed as the phototransducer transmits a "no response" signal to the oscilloscope .

PAGE 85

APPENDIX B FORM USED IN THE JUDGMENT OF SUBJECT PRODUCTION OF FALSETTO REGISTER

PAGE 86

73 JUDGMENT FORM USED BY PHONETICIANS Please indicate with a check mark (X) , if you judge this sample of phonation to be within the subject's falsetto range . Subject Yes No 1 2 3 4 5

PAGE 87

APPENDIX C PROCEDURE FOR MEASURING GLOTTAL AREA FUNCTION

PAGE 88

75 PROCEDURES FOR MEASURING GLOTTAL AREA FUNCTION A. Photographic film image. The Kodak Photo-Optical Data Analyzer and an auxiliary viewing system were utilized in the detailed study of glottal area function for each subject in each experimental condition. After the film image passed through the lens system, it was reflected upward by a mirror placed at a 70 degree angle to a second mirror placed above the main body of the analyzer at a 15 degree angle. This mirror, in turn, redirected the image downward to a layout pad situated 66 inches from the mirror. The linear enlargement ratio of the picture on the film to the projected image was 1:27. The area of the vocal fold opening was traced onto the layout pad and glottal area function of the vocal folds was determined by traditional polar planimeter methods. The space between the vocal folds was outlined by the tracer point of the Keuffel and Esser model 620005 Compensating Polar Planimeter for each frame. A fine line was drawn across the periphery of the glottal area to mark the starting and finishing point. The tracer point was placed precisely at the beginning line and the scale set to zero. The outline of the opening was then circumscribed to the starting point and a reading made. Second and third circuits of the area were made with a measurement reading being made at the end of each circuit. These three readings

PAGE 89

76 were then averaged to obtain area measurement in square centimeters . B. Glottographic images. Immediately following the tracing of the glottal area, the frame displaying the glottographic signal corresponding to the first measured frame of the photographic film was determined. The signal from the electro larynx and the signal from the flash/sync generator allowed for identification of corresponding frames. The lines produced by the input of the phototransducer to the oscilloscope were measured from the parallel 5k Hz lines at a constant position for each photographic film frame. A straight line was scribed on a layout pad and placed under the image of the 5k Hz line and used as a base line. This line also served as an aid in the alignment of the parallel guide instruments employed in drafting the deviations produced by the glottograph and the voice traces. The position for measuring was chosen for consistent clearness throughout the series and held constant. A diagram of the measuring devices used in the data reduction for this investigation may be seen in Figure 15 .

PAGE 90

77 cn QJ Cn 0 £ H W ( 1 ) 'O Figure 15. Diagram illustrating the measurement procedures used in data reduction for the analysis of glottal area function

PAGE 91

APPENDIX D AMPLITUDE AND AREA CURVES DERIVED THROUGH GLOTTOGRAPHIC AND PHOTOGRAPHIC FRAME-BY-FRAME ANALYSIS

PAGE 92

Time in Milliseconds 79 in CN O •H b Cb to p (T> O -P -P O I — I o — o •H d) m cr> io •H £ p4 fO 16. Comparison of photo-glottographic and photographic frame-by-frame s for the vowel /i/ produced in modal register for Subject 1

PAGE 93

Time in Milliseconds 80 o o •H b a (0 Cp o -p +j o I — I a •H (D 0) s-i >1 d -n CP 03 •H G Pm 03 17. Comparison of photo-glottographic and photographic frame-by-frame s for the vowel /i/ produced in modal register for Subject 2

PAGE 94

Time in Milliseconds Glottographic 81 n3 •H £ pL| (0 18. Comparison of photo-glottographic and photographic frame-by-frame s for the vowel /i/ produced in modal register for Subject 3

PAGE 95

in Milliseconds 82 Figure 19. Comparison of photo-glottographic and photographic frame-by-frame analysis for the vowel /i/ produced in modal register for Subject 4

PAGE 96

Time in Milliseconds 83 •H 1 d h O' (0 h a fci f0 20. Comparison of photo-glottographic and photographic frame-by-frame s for the vowel /i/ produced in modal register for Subject 5

PAGE 97

Time in Milliseconds 84 o CN O 0 •H b a fO M tn O -P -P o l — I u Figure 21. Comparison of photo-glottographic and photographic frame-by-frame analysis for the vowel /i/ produced in falsetto register for Subject 1

PAGE 98

Milliseconds 85 r00 G *H i G -H tP fO •H G Cm ro 22. Comparison of photo-glottographic and photographic frame-by-frame s for the vowel /i/ produced in falsetto register for Subject 2

PAGE 99

in Milliseconds 8 6 CN O •H a, ns Cn O P +> O o o *H (L) U) P >1 3 -H tn ns •H C ra 23. Comparison of photo-glottographic and photographic frame-by-frame s for the vowel /i/ produced in falsetto register for Subject 3

PAGE 100

in Milliseconds 87 u *H a, ro •H o E-i -P P o I — I o U O Figure 24. Comparison of photo-glottographic and photographic frame-by-frame analysis for the vowel /i/ produced in falsetto register for Subject 4

PAGE 101

Glottographic 88 Figure 25. Comparison of photo-glottographic and photographic frame-by-frame analysis for the vowel /i/ produced in falsetto register for Subject 5

PAGE 102

Milliseconds 89 o 00 c •H o e •H O •H b a nJ U On o -p +J O I — I o o -rH a) cn M >i 2 rH tn (0 •rH C fo ra 26. Comparison of photo-glottographic and photographic frame-by-frame s for the vowel /i/ produced in vocal fry register for Subject 3

PAGE 103

Time in Milliseconds 100 90 o rq m •H 1 3 -H & ro •H G fci rO . Comparison of photo-glottographic and photographic frame-by-frame for the vowel /i/ produced in vocal fry register for Subject 4

PAGE 104

APPENDIX E FUNDAMENTAL FREQUENCIES

PAGE 105

Table 5. Fundamental frequencies produced by five subjects used in the present investigation. The measures are reported in Hz for the sustained vowel /i/ produced in vocal fry, modal, and falsetto registers. 92 o fa fa 0 0 o o O o o o b CO CP 00 O 00 CO •p fa I — 1 p* cO oo in b (0 Cn CO a fa li g -H P o •H fa p 0 fa On co P 0 O G no H fa p o tn O 0 p •H b £ CL a a. a) >i m 0 b 0 CD i — 1 b b b 0 -p p 0 rH p > b fa P -p p o co P CL p 0 m fa 0 b i — 1 o c E p o o O o o CD P b i — i p00 CO 0 £ fa fa O (N CM 1 — 1 i — i oo p CD G P S 0 b O P £ t? 1 0 CO 0 b fa CD CO fa o > G 0 •P O 0 •n CO 0 & b P P P p mi *• CO fa G P G p 0 CD •H > CO CO 0 i b CO * p CD •P rH fa fa p tP P CO * * * o 0 a 0 I — 1 O O o o o b P O P P OO i — 1 1 — 1 CN CN fa > 0 o i — i >i fa o p P 0 G > fa b -p fa fa p i — 1 0 b P fa H -p 0 O b 0 P b > 0 P 0 p 0 fa g P b 0 p CO -P c b o co -p CL G 0 •p X O b & 0 O ms g b fa i — i 0 0 O P fa -p o p G o b p 0 • O CO G > £ •n i — 1 CN 00 P 1 m 0 0 H b fa b P *P p H a a fa cn * phototransducer which seem to indicate that glottal area was defined by change of light quantity penetrating the area between the vocal folds.

PAGE 106

APPENDIX F RELATIVE INTENSITY

PAGE 107

94 c ' •H >1 U 'V 44 o <13 •n rQ 3 03 <0 o o > 03 rG 4-> 34 O 03 0-1 > h pq 04 fO 03 G ,G -H -P T3 3h 03 O +J 04 p O g a. O 03 r4 34 P 3 03 a 34 o 03 03 J>i 4-3 •H 03 G 03 4-3 G •H G O •H 4-1 3 Co H O 03 4-> 4-3 03 4-* 03 0) 03 34 03 43 (0 r4 Co 0) 34 > •H 4-3 3 i — ! 03 (4 4-1 03 I — I 3 04 G G 03 G • 03 rtj IQ 03 34 * 0) an < — I 05 ,Q 03 G <0 o E4 44 £ o 44 44 0) O in (43 O 00 03 sf i 34 &4 1 — 1 m oo 3 1 o m rH oo oo oo ro o o > 44 C3 03 -ro h cn oo m ,Q 3 c/q 03 3 >4 rH 4J a •H 03 CO G G 03 •H 44 g G 34 •H O O 03 03 > 34 H 44 03 3 rG rH 44 03 • 34 £ > O £ 0) 34 03 04 o 03 i — i rG m 44 g 3 3h CO G 1 G •H 3h H 3 c 03 44 •H 3 3 3 i-4 £ 44 3 O rQ > *H o 44 G G t • 3 rH 3 44 i — I O 3 Or 03 £ 03 34 44 3 O 3 44 Qi 3 G i — 1 03 H 3 •H £ rG 3 34 44 rG O 44 04 >1 rQ >1 i — 1 rQ 3 G G 3 G O A 3 •rH •rH 44 44 3h 3 3 C3 CJ 44 3 •H 3 3 G a. to G £ •H o 3 a rQ 3 3 03 >4 rH ,G 3 3 tH £ > 3 pq • • i — 1 G 03 3 44 > 3 O 3 rG *3 i — 1 44

PAGE 108

BIBLIOGRAPHY Allen, E. L., and Hollien, H. A laminagraphic study of pulse register (vocal fry) phonation. Folia Phoniatrica , in press. Arnold, G. E. Morphology and physiology of the speech organs, in Manual of Phonetics , edited by Kaiser, L. Amsterdam: North-Holland Publishing Co., 1957, pp. 31-64. Black, J. W. Speech science. The Quarterly Journal of Speech , 1951, 37., 493-497. Brackett, I. An analysis of the vibratory action of the vocal folds during the production of tones at selected frequencies. Ph.D. dissertation. Northwestern University, 1947. Coleman, R. F. , and Wendahl, R. W. On the validity of laryngeal photosensor monitoring. The Journal of the Acoustical Society of America . 1968, 44, 1733-1735 Colton, R. H. Some acoustic and perceptual correlates of the modal and falsetto registers. Ph.D. dissertation. University of Florida, 1969. Decroix, G., and Dujardin, J. Etude des accolements glottiques au cours de la phonation par la glottographie de daute frequence. The Journal of Francois Otorhinolaryngolica . 1958, 493-499. Dolansky, L., and Tjernlund, P. On certain irregularities of voiced-speech waveforms . Speech Transmission Laboratory Quarterly Progress and Status Report , 1967, 2-3 , 58-65. 95

PAGE 109

96 Fabre, P. Un procede electrique percutane d ^enscription de l'accolement glottique au cours de la phonation: glottographie de haute frequence. Premiers resultats. Bulletin of the Acadamie National Medicine . 1957, 141 , 66-69. Fabre, P., and Frei, A. Analyse harmonique des glottogrames et des phonogrammes de la voix chantee. Annals of Oto -Laryngology , 1959, 7j5, 459-463 . Fant, G., Ondrackova, J., Lindqvist, J., and Sonesson, B. Electrical glottography . Speech Transmission Labora tory Quarterly Progress and Status Report . 196/, 4, 15-21. 4 Farnsworth, D. W. High-speed motion pictures of the human vocal cords. Bell Telephone Laboratories Record , 1940, 18., 203-206. Fink, B. R., and Kirschner, F. Observations on the acoustical and mechanical properties of the vocal folds. Folia Phoniatrica , 1959, l^L, 168-177. Fletcher, S. G. Analysis of cinema films in diagnosis and research. Journal of Biological Photographic Associ ation , 1958, 26., 29-33. Fourcin, A. J., and Abberton, E. First applications of a new laryngograph . Medical and Biological Illustration . 1971, 21 ^, 172-182. FrokjaerJensen, B. Manual for Photo-Electric Glottograph . Holte : Engineering Firm of Acoustical and Medical Electronics, 1970. Griesman, Bruno L. Mechanism of phonation demonstrated by planigraphy of the larynx. Archives of Otolaryngology , 1943, 38 _, 17-26. Hartmann, W. , and Wullstein, H. Untersuchunger uber den Bewegungsvorgang an den schwingenden Stimmlippen von Kehlkopf praparaten mit verberserter Photozellenmethode . Archives Ohren— ITasen-und Kehlkopfheilkunde , 1938, 144, 348-360.

PAGE 110

97 Hays, W. L. Statistics . New York: Holt, Rinehart, and Winston, Inc., 1963. Hertz, G. H., Lindstom, K. , and Sonesson, B. Ultrasonic recording of the vibrating vocal folds. Acta Otolarvngoloqica , 1970, 69., 223-230. Hollien, Ii., and Moore, G. P. Measurements of the vocal cords during changes in pitch. Journal of Speech and Hearing Research , 1960, _3, 159-165. Hollien, H. The relationship of vocal fold length to vocal pitch for female subjects. Proceedings of the Xllth International Speech and Voice Therapy Conference , Padua, 1962, pp. 38-43. Hollien, H., and Curtis, J. F. Elevation and tilting of vocal folds as a function of vocal pitch. Folia Phoniatrica , 1962, 14 , 23-36. Hollien, H. , Moore, P., Wendahl, R. W., and Michel, J. F. On the nature of vocal fry. Journal of Speech and Hearing Research , 1966, 9., 245-247. Hollien, H., and Michel, J. F. Vocal fry as a phonational register. Journal of Speech and Hearing Research , 1968, 11, 600-604. / Hollien, H., Damste, H. , and Murry, T. Vocal fold length during vocal fry phonation. Folia Phoniatrica , 1969, 21 , 257-265. Hollien, H. Vocal registers: A new look at an old problem. Paper presented at the Collequim Medicorum Theatri, Buenos Aires, Argentina, 1971. Hollien, H., Brown, W. S., and Hollien, K. Vocal fold length associated with modal, falsetto and varying intensity phonations . Folia Phoniatrica , 1971, 23 , 66-78. Koster, J. P., and Smith, S. Zur interpretation elektrischer und photoelektrischer glottogramme . Folia Phoniatrica , 1970, 22, 92-99.

PAGE 111

98 Lebrun, Y., and Hasquin-Deleval, J. On the so-called Â’dissociations' between electroglottogram and phonogram. Folia Phoniatriea , 1971, 23_, 225-227. Lindqvist, J. Laryngeal mechanism in speech. Speech Transmission Laboratory Quarterly Progress and Status Report , 1969, 2-3 , 26-32. Lindqvist, J. The voice source studied by means of inverse filtering. Speech Transmission Laboratory Quarterly Progress and Status Report , 1970, 1_, 3-9. Lindqvist, J., and Lubker, J. Mechanisms of stop consonant production. Speech Transmission Laboratory Quarterly Progress and Status Report , 1970, _1 , 1-9. Lisker, L., Abramson, A. S., Cooper, F. S., and Schvey, M. H. Transillumination of the larynx in running speech. The Journal of the Acoustical Society of America , 1969, 45 , 1544-1546. Luchsinger, R. , and Arnold, G. Voice-Speech-Language : Clinical Communicology, Its Physiology and Pathology . Belmont, California: Wadsworth Publishing Co., 1965. Metzger, W. The mode of vibration of the vocal cords. Psychological Monographs , 1928, 38., 82-159. Miller, R. L. Nature of the vocal cord wave. The Journal of the Acoustical Society of America , 1959, _31, 667677. Minifie, F., Kelsey, C. A., and Hixon, T. Measurements of vocal fold motion using an ultrasonic Doppler velocity monitor. The Journal of the Acoustical Society of America , 1967, J5, 1165. Moore, P. A short history of laryngeal investigation. The Quarterly Journal of Speech , 1937a, 2_3, 531-564. Moore, P. Vocal fold movement during vocalization. Speech Monographs , 1937b, 4_, 44-55. Moore, P., and von Leden, H. Dynamic variations of the vibratory pattern in the normal larynx. Folia Phoniatriea , 1958, 10, 205-238.

PAGE 112

99 Moore, G. P., White, F. D., and von Leden . H. Ultra high speed photography in laryngeal physiology. Journal of Speech and Hearing Disorders , 1962, 27_, 165-171. Moore, P. Discussions of the proceeding paper. Annals of the New York Academy of Sciences , 1968a, 155 , 39-41. Moore, P. Otolaryngology and speech pathology. The Laryngoscope , 1968b, 78., 1500-1509. Morner, J., Franeson, P., and Fant, G. Vocal register terminology and standard pitch. Speech Transmission Laboratory Quarterly Progress and Status Report , 1963, 4. Musehold, A. Stroboskopische und photographische Studien uber die Stellung der Stimlippen im Brust-und FalsettRegister. Archives of Laryngology und Rhinology , 1897, 7 , 1-21. Ohala, J. A. new photo-electric glottograph. University of Los Angeles, Working Papers in Phonetics , 1966, 4, 40-55. Ohala, J. Studies of variations in glottal aperture using photo-electric glottography . The Journal of the Acoustical Society of America , 1967, 4JL, 1613(A). Peterson, G. E. Speech and hearing research. The Journal of Speech and Hearing Research , 1958, 1 , 3-11. Pronovost, W. Experimental study of the habitual and natural pitch levels of superior speakers. Speech Monographs , 1942, 9 _, 111-123. Rubin, J. J., and Hirt, C. C. The falsetto. A high speed cinematographic study. The Laryngoscope , 1960, 70 , 1305-1324. Russell, G. 0. Speech and Voice . New York: The MacMillan Co., 1931. Sawqshima , M. , and Hirose, H. New Laryngoscopic technique by use of fiber optics. The Journal of the Acoustical S ociety of America , 1968, 43, 168-169.

PAGE 113

100 Sawashima, M. , Hirose, H., Kiritani, S., and Fujimura, 0. Articulatory movements of the larynx. The 6th Inter national Congress on Acoustics , Tokyo, 1968, pp. 1-4. Sawashima, M. Devoiced syllables in Japanese A preliminary study by photoelectric glottography . Annual Bulletin Research Institute of Logopedics and Phoniatrics, Uni versify of Tokyo , 1969, 3_, 35-41. Sawashima, M. , Abramson, A., Cooper, F., and Lisker, L. Observing laryngeal adjustments during running speech by use of a fiberoptics system. Status Report on Speech Research, Haskins Laboratories , 1969, 19-20 , 201 210 . / u / / Sedlackova, E. Sonnees stroboscopiques en relation avec le developpement de la voix des enfants. Folia Phoni atrica , 1961, 131 , 81-92. Smith, S. Remarks on the physiology of the vibrations on the vocal cords. Folia Phoniatrica , 1954, 6 , 166-178. Snedicor, J. S. The pitch and duration characteristics of superior female speakers during oral reading. The Journal of Speech and Hearing Disorders , 1951, 16 , 44-52 . Sonesson, B. A method for studying the vibratory movements of the vocal cords. Journal of Laryngology and Otology , 1959, 73, 732-737. Sonesson, B. On the anatomy and vibratory pattern of the human vocal folds. Acta Oto-Laryngologica Supplement , 1960, 156., 1-80. Sonesson, B. The functional anatomy of the speech organs, in Manual of Phonetics edited by Malmberg, B. Amsterdam: North-Holland Publishing Co., 1968, pp. 45-75. Sovak, M. , Courtois, J., Haas, C., and Smith, S. Observations on the mechanism of phonation investigated by ultraspeed cinef luorography . Folia Phoniatrica , 1971, 23, 277-287.

PAGE 114

101 Spencer, F. R. Transillumination of the larynx and upper trachea. Annals of Otology Rhinology, and Laryngology , 1917, 2j5 , 530-536. Tarnoczy, T. H. The opening time and opening-quotient of the vocal cords during phonation. The Journal of the Acoustical Society of America , 1951, 23_, 42-44. Timcke, R., von Leden, H., and Moore, P. Laryngeal vibrations: Measurements of the glottic wave. Part 1, The normal viberatory cycle. Archives of Otolaryngol ogy , 1958, 6j3, 1-9. Timcke, R., von Leden, H., and Moore, P. Laryngeal vibrations: Measurements of the glottic wave. Part 2, The physiological variations. Archives of Otolaryngology , 1959, 69, 438-44. Vallancien, B., and Faulhaber, J. What to think of glottography . Folia Phoniatrica , 1967, 19, 39-44. van den Berg, Jw., Zantema, J. T., and Doornenbal, P. On the air resistance and the Bernoulli effect of the human larynx. The Journal of the Acoustical Society of America , 1957, 2j3, 626-631. van den Berg, Jw. Myoelastic-aerodynamic theory. Journal of Speech and Hearing Research , 1958, 1 , 227-244. van den Berg, Jw. Vocal ligaments versus registers. Current Problems in Phoniatrics and Logopedics , 1960, 1 , 19-34. van den Berg, Jw. Mechanism of the larynx and the laryngeal vibrations, in Manual of Phonetics edited by Malmberg, B. Amsterdam: North-Holland Publishing Co., 1968a, pp. 278-308. van den Berg, Jw. Register Problems. Annals of the New York Academy of Sciences , 1968b, 155 , 129-134. Van Michel, P. C. MGUvements glottiques phonatoires sans emission sonore. Etude electroglottographique . Folia Phoniatrica , 1966, 2J3, 1-8.

PAGE 115

102 Van Michel, C., Pfister, K. A., and Luchsinger, R. Electroglottographie et cinematographic laryngee ultra-rapide . Folia Phoniatrica , 1970, 2_2, 81-91. van Oordt, J. W. A., and Drost, H. A. Development of the frequency range of the voice in children. Folia Phoniatrica , 1963, 15 , 289-298. Vennard, W. The Bernoulli effect in singing. The National Association of Teachers of Singing Bulletin , 1960, 17 . Walker, H. M. , and Lev. J. Statistical Inference . New York: Henry Holt and Co., 1953. Wendahl, R. W., and Coleman, R. F. Vocal-cord spectra derived from glottal area waveforms and subglottal photocell monitoring. The Journal of the Acoustical Society of America , 1967, 41, 1613(A). West, R. The nature of vocal sounds. Quarterly Journal of Speech Education , 1926, 12 , 244-295. Zemlin, W. R. A comparison of high speed cinematography and a transilluminat ion-photoconductive method in the study of the glottis during voice production. M.S. thesis. University of Minnesota, 1959.

PAGE 116

BIOGRAPHICAL SKETCH R. Joyce Redus Harden was born June 27, 1929, at McCaulley, Texas. In June, 1945, she was graduated from Hamlin High School, Hamlin, Texas. She received a Bachelor of Arts degree with a major in Speech in June, 1949, from ' Texas Technological College, Lubbock, Texas. She taught in the Brown Schools in Austin, Texas, as a teacher of Exceptional Children in 1950. In 1953 she was employed as a classroom teacher of the deaf in the public schools of Fort Worth, Texas. She taught at the Ridglea Presbyterian Day School, Fort Worth, Texas, as a kindergarten teacher in 1964. In June, 1965, she enrolled in the Graduate School of Texas Christian University where she was granted a teaching assistantship . In August, 1967, she was graduated from Texas Christian University with a Master of Arts degree with a major in Speech Pathology. She was appointed to the faculty as an Instructor in 1967. In 1970, she was granted a leave of absence from Texas Christian University and enrolled in the Graduate School of the University of Florida in September of that year. 103

PAGE 117

104 She held an assistantship there until September, 1971, followed by a teaching associate appointment until June, 1972 . She is a member of Alpha Psi Omega, Sigma Alpha Eta, the Texas Speech and Hearing Association, and The American Speech and Hearing Association. She holds professional certification in speech and hearing therapy in the state of Texas, and the Certificate of Clinical Competence in Speech Pathology in the American Speech and Hearing Association. She is married to Robert Wayne Harden and is the mother of three sons .

PAGE 118

I certify that I have read this study and that in ray opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. G. Paul. Moore, Chairman Professor of Speech I certify that I .have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. Edward C. Hutchinson Associate Professor of Speech I certify that I have read this study and that in ray opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. 7 > Norman N. Mark el Associate Professor of Speech, Psychology, and Anthropology

PAGE 119

I certify that I have read this study and that in ray opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. This dissertation was submitted to the Department of Speech in the College of Arts and Sciences and to the Graduate Council, and was accepted as partial fulfillment of the requirements for the degree of Doctor of Philosophy. June, 1972 J Assistant Professor of Ps^bhology Dean, Graduate School