You are on page 1of 9

Pitch Discrimination and Pitch Matching Abilities

with Vocal and Nonvocal Stimuli


*Robert E. Moore, *Julie Estis, *Susan Gordon-Hickey, and Christopher Watts
*Mobile, Alabama and Harrisburg, Virginia

Summary: Various stimulus types have been investigated in pitch discrimination and pitch matching tasks. However, previous studies have not explored
the use of recorded samples of an individuals own voice in performing these
two tasks. The purpose of this study was to investigate pitch discrimination
and pitch matching abilities using three stimuli conditions (participants
own voice, a neutral female voice, and nonvocal complex tones) to determine
if pitch discrimination and/or pitch matching abilities are influenced by the
type of stimuli presented. Results of the pitch discrimination tasks yielded
no significant difference in discrimination ability for the three stimuli. For
the pitch matching tasks, a significant difference was found for the participants voice versus neutral female voice and the participants voice versus
tonal stimuli. There was no significant difference in pitch matching ability between the neutral female voice and the tonal stimuli. There was no significant
correlation between pitch discrimination and pitch matching abilities for any
of the three stimuli types. These results suggest that it is easier to match the
pitch of ones own voice than to match the pitch of a neutral female voice and
nonvocal complex tones, although no difference was found for pitch discrimination abilities. One possible implication of this study is that differences in
matching the pitch of ones own voice compared to matching other stimuli
types may help to differentiate the source of singing inaccuracy (motor vs discrimination skills).
Key Words: Pitch matchingPitch discriminationVocal stimuliNonvocal stimuliAuditory feedback.

INTRODUCTION
The ability to accurately discriminate and match
pitches varies across individuals. Singers and trained
musicians must exhibit accurate pitch discrimination and pitch matching skills to adequately perform
musical tasks. Some individuals are unable to sing in
tune or to hear differences among pitches. The nature of these differences in pitch discrimination
and pitch matching abilities across individuals is
uncertain. Such differences may result from inherent
ability, from structural or functional differences, or
from exposure to music and musical training.1,2

Accepted for publication October 27, 2006.


From the *Department of Speech Pathology & Audiology,
University of South Alabama, Mobile, Alabama; and the
Department of Communication Sciences & Disorders, James
Madison University, Harrisburg, Virginia.
Address correspondence and reprint requests to Robert E.
Moore, Department of Speech Pathology & Audiology, University of South Alabama, 2000 UCOM, Mobile, AL 366880002. E-mail: rmoore@usouthal.edu
Journal of Voice, Vol. 22, No. 4, pp. 399407
0892-1997/$34.00
2008 The Voice Foundation
doi:10.1016/j.jvoice.2006.10.013

399

400

ROBERT E. MOORE ET AL

To accurately match pitch with the voice, individuals must use accurate pitch discrimination
and precise control of the vocal mechanism. Accurate pitch discrimination first involves hearing the
presented stimuli; therefore, normal auditory function is required. Then, the pitch must be held in
pitch memory while being compared.3 Higher-level
processes that may be involved include classification of stimuli types, identification of stimuli relationships, and allocation of attention resources.
Similarly, accurate pitch matching involves hearing
the target stimulus, forming an internal representation of that stimulus in pitch memory, plus the planning and coordination of the vocal mechanism to
produce a sound that matches the target stimulus.
One of the processes likely to be involved in
pitch matching is auditory feedback. Auditory feedback is a primary means of sensory feedback for
speech.46 Speakers appear to monitor their auditory feedback to produce speech that is intelligible
and in agreement with the linguistic and suprasegmental intent of the message. Researchers have
shown that individuals attempt to modify the pitch
or loudness of their voice if it does not match the
auditory feedback.5,7,8 For example, when the vocal
fundamental frequency (F0) being fed back to an
individual is altered, the individual responds by altering the pitch of their voice in either a positive or negative direction.9 While producing a pitch to match
a target stimulus, the individual monitors the pitch
of their voice and compares it to the perceived pitch
of the target. Adjustments to vocal production are
made based on that feedback. It is likely that individuals who are accurate pitch matchers possess perceptual systems that are more finely tuned, or use
different strategies for monitoring auditory feedback
of their own vocal production, compared to inaccurate pitch matchers. This occurs although perception
of our own voice is different when heard during production as compared to when heard after being recorded and played through an external transducer.
The perception of our voice during production is impacted by the fact that the sound is heard via bone
conduction and air conduction, whereas when heard
through an external transducer, the sound is heard
almost exclusively via air conduction.
Individuals with good pitch matching skills tend
to have good pitch discrimination skills. This
Journal of Voice, Vol. 22, No. 4, 2008

relationship between pitch discrimination and pitch


matching has been investigated in a variety of populations. These include school-aged children, nonvocal musicians, trained singers, and individuals
with no formal music training.1013 Trained musicians and singers often excel at pitch matching
and pitch discrimination tasks. In addition, some individuals with no formal music training exhibit accurate pitch discrimination and pitch matching
abilities similar to those of trained vocal musicians.
Watts et al13,14 reported that these individuals were
judged as being accurate singers, whereas those
with poor pitch discrimination and pitch matching
skills tended to be judged as inaccurate singers.
Studying inaccurate adult singers, Bradshaw and
McHenry15 found that all participants performed
poorly on a pitch matching task. They also reported
no correlation between pitch discrimination and
pitch matching abilities with this population.
For pitch discrimination and pitch matching tasks,
various stimulus types have been explored. These
include pure tones, computer-generated complex
tones, complex tones generated by musical instruments, and vocal samples.1619 Previous research
involving pitch matching tasks have not used recorded samples of participants own voices. In particular, pitch matching accuracy for ones own
voice has not been compared to pitch matching accuracy with other stimuli types, such as neutral voice
samples and nonvocal complex tones. The accuracy
of matching the pitch of ones own voice may hold
clinical relevance, as a traditional means of therapy
has been to improve a patients perceptual capacity,
via exercises or alterations in vocal F0 used to
restore a normal balance of laryngeal muscular
activity.2022 If stimulus type influences discrimination and reproduction accuracy relative to vocal pitch
targets, then stimulus type could be used to enhance
therapeutic management. Additionally, singers must
monitor their own voice to achieve accurate pitch
production. During performances singers listen to
their own voice through monitors and adjust the pitch
accordingly. The purpose of this study was to examine pitch matching abilities with three types of stimuli (viz, participants own voice, a neutral female
voice, and nonvocal complex tones) to determine if
pitch matching abilities are influenced by the type
of stimuli presented. The research questions were

PITCH MATCHING/DISCRIMINATION VOCAL/NONVOCAL STIMULI


1. Does stimulus type influence pitch discrimination and pitch matching accuracy in untrained individuals?
2. Are there differences in pitch discrimination
and pitch matching accuracy in untrained individuals who are accurate and inaccurate
pitch matchers, when presented with different
stimuli types?
It was expected that pitch discrimination accuracy and pitch matching accuracy would be influenced by stimulus type. Specifically, it was
expected that discrimination and pitch matching
would be more accurate when presented with an individuals own voice as the target for discrimination
and reproduction. Additionally, it was expected that
individuals who display accurate pitch matching
abilities to nonvocal complex tones would be able
to both discriminate and match pitches of varying
stimulus types consistently more accurately compared to inaccurate pitch matchers.

METHODS
Participants
Participants included 20 females recruited from
the University of South Alabama and community.
Because there appears to be an effect of age on frequency discrimination, only individuals between 20
and 30 years of age participated.23 All participants
had hearing thresholds of 25 dB hearing level or
better at 500, 1000, 2000, and 4000 Hz.24 Each participant completed a questionnaire to ensure that
there was no history of voice, speech, language,
and/or hearing disorders. Only potential participants who indicated on the questionnaire that they
had not received formal vocal musical training
were included in the study.
Equipment and stimuli
All preliminary procedures, pitch discrimination,
and pitch matching tasks were performed in
a sound-attenuated room, meeting specifications
for permissible ambient noise levels.25 The audiometric screening was done using an audiometer,
calibrated to ANSI standards.26 The pure tone stimuli were presented through TDH-50 headphones
(Teledyne, New York, NY).

401

Stimuli for pitch discrimination and pitch matching tasks were created using Adobe Audition (Version
1.5; Adobe Systems Inc., San Jose, CA) sound editing software and the Computerized Speech Laboratory (CSL 4500; Kay Elemetrics Corporation,
Lincoln Park, NJ). Three types of stimuli were created, namely, nonvocal complex tones, neutral
female voice samples, and samples of each participants voice. Five nonvocal complex tones were
generated with an F0 of 212, 218, 224, 206, and
200 Hz. With 212 Hz being the reference frequency, the other frequencies represent 50 cents,
100 cents,  50 cents, and  100 cents, respectively. These complex tones were composed of four
equal amplitude harmonics, which were added in
phase. Each tone had a total duration of 350 milliseconds, and was gated on and off with 10 milliseconds
linear amplitude ramps. All nonvocal complex tones
were recorded at equal amplitudes.
For the neutral female voice condition, a female
volunteer, who did not participate as a listener, sustained the vowel ah at a comfortable pitch. This
sample was recorded using the CSL 4500 at a sampling rate of 44.1 kHz. Then, this sample was digitally manipulated using Adobe Audition to have an
F0 of 212 Hz and duration of 350 milliseconds. Using the same software, four additional voice samples were created having the same frequencies
and durations as the nonvocal complex tones.
In a similar manner, each participants voice was
recorded as they sustained production of the vowel
ah at a comfortable pitch. Acoustic analysis was
conducted via the CSL 4500 and Multidimensional
Voice Range Profile-Advanced (MDVP-A2.7.0; Kay
PENTAX, Lincoln Park, NJ) software to ensure that
cycle-to-cycle variations in frequency and amplitude (jitter and shimmer, respectively), and noiseto-harmonic ratios were within normal limits.
This recording was manipulated to have the same
duration as the nonvocal complex tones and female
voice samples. For this condition, the 50 cent,
100 cent, 50 cent, and 100 cent samples
were referenced to each participants individual
F0, thus varying for each individual. The individuals F0 and additional tones are shown in Table 1.
For the pitch discrimination task, the stimuli
were grouped into sequential sets of three within
each condition. For each set, there were four
Journal of Voice, Vol. 22, No. 4, 2008

402

ROBERT E. MOORE ET AL

TABLE 1. Fundamental Frequencies for Each


Participants Own Voice Sample and Digitally
Derived Samples Used for Pitch Discrimination
and Pitch Matching Tasks
Subject F0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20

216.72
232.96
230.75
288.45
258.00
231.00
266.00
246.57
239.77
235.97
265.80
221.00
276.60
239.00
240.00
203.17
296.11
207.90
219.06
207.92

F0 L 100
cents

F0 L 50
cents

F0 D 50
cents

F0 D 100
cents

203.87
219.34
218.30
272.51
243.76
218.85
251.14
233.22
225.80
239.70
251.37
207.84
261.14
225.39
227.27
191.57
279.00
196.10
206.77
196.20

210.15
225.53
224.26
280.81
249.70
224.70
258.00
239.80
233.09
246.10
258.26
213.84
269.20
231.77
233.68
197.23
287.39
201.86
212.79
201.87

222.35
239.77
237.53
297.39
266.39
237.22
273.67
254.42
246.75
261.25
272.96
227.56
285.16
246.30
248.29
209.34
305.20
213.92
225.58
214.31

228.27
246.54
245.54
305.98
273.50
245.00
282.48
261.00
254.00
269.37
281.72
234.45
294.30
253.94
255.62
215.80
313.54
221.71
231.89
220.72

possible combinations: (1) all three stimuli having


the same F0, (2) the F0 of the first stimulus differing
from the second and third stimuli, (3) the F0 of the
second stimulus differing from the first and third
stimuli, and (4) the F0 of the third stimulus differing from the first and second stimuli. A total of
65 stimuli sets were generated for each of the three
stimuli conditions (Appendix A).
The stimuli for the pitch matching task consisted
of nonvocal complex tones, neutral female voice
samples, and samples of the participants voice at
the five generated pitches, for a total of 15 stimuli.
For this task, each stimulus was presented in sound
field via a Tucker-Davis System 3 psychoacoustic
workstation.
Procedures
For the pitch discrimination task, participants
were seated at a desk inside the sound-attenuated
room, with a computer monitor and mouse. The investigator monitored the participants responses
and progress on a networked computer from
Journal of Voice, Vol. 22, No. 4, 2008

outside the room. The investigator was also able


to see the participant using a closed-circuit camera
and monitor system. Presentation of each series of
stimuli was controlled by ECos/Win software
(AVAAZ Innovations, Ontario, Canada). The stimuli were presented at 75 dB Sound pressure level
(SPL) in sound field through the same speaker
and equipment used in the pitch matching task.
Participants were instructed to listen to each
three-tone set and indicate if the stimuli were the
same pitch or if one of the stimuli was different
in pitch. If one of the stimuli was different in pitch,
the participant was instructed to identify that stimulus as the first, second, or third stimulus. Participants recorded their judgment of the three tones
by clicking on the appropriate icon displayed on
the computer monitor with a computer mouse. A
stimulus set was not presented until the participant
had made a judgment regarding the preceding set.
Each of the 65 stimuli sets was presented two times
in random order. The order of the presentations was
randomized by the ECos/Win software (AVAAZ Innovations). All participant responses were saved to
the computer hard drive for further analysis.
For the pitch matching task, each of the 15 targets was randomly presented two times via sound
field through a Fostex 7301B3E amplified speaker
(Fostex Corporation, Tokyo, Japan) at 75 dB SPL
with the participant seated 1.5 m from the speaker.
A total of 30 targets were presented in two trials.
Participants were instructed to listen to the target
presented, and then to vocally match the pitch of
each target with the vowel ah for 6 seconds.
The investigator timed the 6-second interval. Participants pitch matching responses were recorded using a head-mounted microphone routed to the CSL
4500, digitized at a sampling rate of 44.1 kHz. Each
participants response was saved in an individual
file to the hard drive of the computer for future
analysis.
Analysis
After data collection, participants were assigned
groups based on pitch matching accuracy to the
nonvocal complex tones. Participants who had an
average semitone difference score in the nonvocal
stimulus condition less than 1.0 semitones were
considered accurate pitch matchers. All other

PITCH MATCHING/DISCRIMINATION VOCAL/NONVOCAL STIMULI


participants were considered inaccurate pitch
matchers, because a 1.0 semitone difference or
greater constitutes a half tone or more difference
on the musical scale.
Two dependent variables were measured in this
study: pitch discrimination accuracy and pitch
matching accuracy. Performance accuracy for the
pitch discrimination task was measured by calculating the percentage of correct score for each participant from a total of 130 stimulus presentations
for each of the three stimuli types. Pitch matching
accuracy was calculated by analyzing the middle
2 seconds of each participants vocal match. This
time frame was chosen to avoid including onset
and offset of vocal production, which could lead
to variability in the measurement. Short-term variation in frequency, which could also impact overall
F0 measurement, was within normal limits for all
participants. The F0 of each sample was calculated
by the CSL 4500 using the MDVP-A software. An
average of the two vocal attempts for each target
was calculated. The frequency of the target and average frequency of the pitch matching attempts
were used to calculate the difference in semitones.
An average semitone difference was calculated in
this manner for each of the three stimuli conditions. Absolute values were used for all measures
to accurately reflect F0 differences. The use of actual differences with positive numbers indicating
productions higher in frequency than the target
stimuli and negative numbers indicating productions lower in frequency than the target stimuli
could have resulted in misrepresentation of an inaccurate pitch matcher as an accurate pitch matcher.
A smaller score indicated more precise pitch
matching accuracy.
A 2 (group)  3 (condition) repeated measures
ANOVA was performed to compare the pitch discrimination score means of accurate and inaccurate
pitch matchers in the three stimulus conditions. A
second 2 (group)  3 (condition) repeated measures
ANOVA was performed to compare the pitch
matching response means of accurate and inaccurate pitch matchers in the three stimulus conditions.
The presence and strength of the relationship between pitch discrimination and pitch matching for
the three conditions was also evaluated using Pearsons product-moment correlations.

403

RESULTS
Based on pitch matching accuracy to nonvocal
stimulus tones, 16 participants were accurate pitch
matchers (average semitone differences ranging
from 0.07 to 0.61 semitones) and four were inaccurate pitch matchers (average semitone differences
ranging from 1.04 to 5.05 semitones). For the pitch
discrimination task, both groups were most accurate for their own voice. The neutral female voice
samples were most difficult to discriminate for
the accurate pitch matchers, whereas the nonvocal
complex tones were most difficult to discriminate
for the inaccurate pitch matchers.
Individual means and standard deviations for
pitch matching are presented in Table 2. In both
groups, pitch matching of their own voice was
most accurate for most participants. The neutral female voice samples were more difficult to match
than the nonvocal complex tones. Group means
and standard deviations for pitch matching tasks
are presented in Figure 1.
TABLE 2. Individual Means and Standard
Deviations of Absolute Semitone Differences for
the Three Conditions of the Pitch Matching Task
Subject
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20

Own
Voice
0.18
0.89
0.11
0.18
0.04
0.75
0.65
0.09
0.07
0.15
3.01
0.31
0.44
0.09
0.10
0.14
0.22
0.51
0.27
1.53

(0.20)
(0.65)
(0.09)
(0.04)
(0.02)
(0.50)
(0.37)
(0.06)
(0.04)
(0.11)
(1.02)
(0.25)
(0.28)
(0.07)
(0.08)
(0.15)
(0.15)
(0.44)
(0.07)
(1.04)

Female
Voice
0.07
2.05
0.19
0.09
0.13
0.67
2.21
0.38
0.14
0.26
5.90
0.72
2.31
0.10
0.21
0.28
0.40
0.29
0.16
0.70

(0.07)
(0.61)
(0.18)
(0.09)
(0.07)
(0.31)
(0.67)
(0.08)
(0.07)
(0.13)
(2.64)
(0.35)
(0.81)
(0.09)
(0.11)
(0.15)
(0.12)
(0.18)
(0.10)
(0.68)

Nonvocal
Tone
0.09
2.00
0.61
0.07
0.09
0.31
1.04
0.40
0.09
0.24
5.05
0.47
1.85
0.14
0.18
0.20
0.37
0.31
0.21
0.41

(0.06)
(0.70)
(0.39)
(0.02)
(0.05)
(0.18)
(0.45)
(0.26)
(0.11)
(0.09)
(3.09)
(0.46)
(0.50)
(0.14)
(0.08)
(0.10)
(0.16)
(0.31)
(0.18)
(0.34)

Average
0.15
1.83
0.49
0.10
0.09
0.51
1.38
0.33
0.15
0.23
5.61
0.56
1.64
0.12
0.17
0.21
0.48
0.34
0.18
1.13

Note: Standard deviations are in parenthesis. Boldface type


indicates the inaccurate pitch matching group.
Journal of Voice, Vol. 22, No. 4, 2008

404

ROBERT E. MOORE ET AL
two stimuli conditions. There was no significant
difference between the neutral female voice condition and the nonvocal complex tone condition. Statistical summary of paired comparisons across
groups and stimuli for pitch matching is presented
in Table 3.
DISCUSSION AND CONCLUSION

FIGURE 1. Mean semitone difference scores for individuals


with accurate and inaccurate pitch matching abilities in four
stimuli conditions.

For the pitch discrimination task, a repeated


measures ANOVA indicated no significant main effects for stimulus type and no interaction between
stimulus type and pitch matching accuracy. Multiple bivariate correlation analyses using Pearson
product-moment correlations were performed to investigate the relationship between pitch discrimination and pitch matching for each stimulus
condition. For each of the three stimuli conditions,
correlations were nonsignificant.
For the pitch matching task, a repeated measures
ANOVA was used to compare pitch matching performance across groups and among conditions.
Mauchlys Test of Sphericity was significant
(W 5 0.481); therefore, sphericity was not assumed, and Huynh-Feldt corrected tests were used
to measure significance. Results of the ANOVA indicated a significant main effect for stimulus type,
F 5 31.626, P ! 0.05, and a significant interaction
between stimulus type and pitch matching accuracy, F 5 29.217, P ! 0.05. The significant difference for stimulus type was further studied by
three paired comparisons: own voice versus neutral
female voice, own voice versus nonvocal complex
tone, and neutral female voice versus nonvocal
complex tone. To control for type I errors, the
Holms sequential Bonferroni procedure was used
to adjust the alpha level. Pairwise comparisons indicated a significant difference between the participants matching of their own voice and the other
Journal of Voice, Vol. 22, No. 4, 2008

The purpose of this study was to investigate


whether stimulus type influences the accuracy of
pitch discrimination and vocal pitch matching.
Three stimulus types were compared: (1) nonvocal
complex tones, (2) a neutral female voice, and (3)
ones own voice. In addition, this study sought to
determine if pitch discrimination and pitch matching abilities were differentially influenced by stimulus type in individuals who were accurate pitch
matchers and those who were inaccurate pitch
matchers.
The ability to perceptually discriminate the pitch
of comparison tones was not significantly influenced by stimulus type. However, the results indicated that individuals are more accurate at
matching the pitch of their own voice compared
to another female voice or a nonvocal complex
tone. One reason for this result may have to do
with the frequency of the stimuli. The complex
tone and the female voice were at set frequencies,
whereas the frequencies for the participants own
voice varied across participants. The stimuli created
from the participants own voice were produced by
having the participant sustain ah at their most
natural pitch. The F0 of the sample was used as
the base frequency for each participant. From
TABLE 3. Statistical Summary of Paired
Comparisons Across Stimuli Conditions
for Pitch Matching
Paired
Comparisons
Own voice
vs female voice
Own voice
vs complex tone
Female voice
vs complex tone

Mean
Difference

Standard
Error

Significance

0.957

0.115

0.000*

0.746

0.164

0.001*

0.210

0.088

0.082

Note: * indicates significant difference between conditions.

PITCH MATCHING/DISCRIMINATION VOCAL/NONVOCAL STIMULI


each sample obtained, five stimuli were created.
These were at F0, F0 100 cents, F0 50 cents,
F0  50 cents, and F0  100 cents. Thus, the F0
of the samples with the participants own voice differed from the frequency of the complex tone and
female voice stimuli. The fact that the own voice
stimuli were at or near the natural pitch of the participant may have made pitch matching easier. This
is emphasized by comparing the accurate and inaccurate pitch matching groups. The accurate pitch
matchers were generally accurate for all three stimuli. However, the inaccurate pitch matchers were
most accurate for their own voice.
Based on listening to ones own voice from a
recording device, it might seem surprising that an
individual is able to match their own voice more
accurately. The reason ones own voice sounds different from a recording device is that when heard
naturally, the voice is heard by both bone conduction and air conduction. When heard through a recording device it is heard through air conduction
only. It appears from the findings of this study that
F0 is maintained when heard through both bone conduction and air conduction. The difference we hear
is due to changes in the spectrum of the sound.
If individuals show good pitch discrimination
and matching abilities, their mechanism for completing these tasks appears to be more flexible as they are
able to accurately match pitch regardless of the stimulus type. However, for individuals who are less accurate and less flexible, they may have difficulty with
stimuli that differ from their natural range of production. In other words, when presented with a stimulus
having a pitch different from their own natural pitch,
they will tend to produce a sound near their natural
pitch. Continued exploration of the impact of ones
own voice as the stimulus for pitch matching is warranted. Specifically, comparing the individuals performance at or near their natural F0 with other types
of vocal and tonal stimuli will provide insight regarding the nature of the improved performance seen in
the present study.
In the present study, no correlation was found
between pitch discrimination ability and pitch
matching ability. This is not in agreement with previous work in this area, which has shown a significant correlation between pitch discrimination
ability and pitch matching ability.13 One possible

405

reason for this difference is that in previous research participants were asked to determine if two
stimuli presented were the same or different in
pitch. In this study, participants were asked to
make pitch discrimination judgments between three
stimuli. Participants indicated if the three stimuli
were the same in pitch or identified the stimuli of
the three with the odd pitch. In reviewing the data
from this study and previous studies, it would appear that the three stimuli paradigm was easier
than the two stimuli paradigm even though frequency differences between stimuli were similar
(eg, within one semitone or less). In general, participants in this study were more accurate pitch
discriminators than in previous studies. All participants in this study achieved more than 80% accuracy on the pitch discrimination task. It may be
that hearing two stimuli of the same frequency solidified the representation of the pitch in working
memory, thereby making it easier to correctly discriminate the pitch of the differing stimulus. Future
research should investigate differences in pitch discrimination performance by comparing a two stimuli paradigm to a three stimuli paradigm.
Another possible factor for the lack of agreement
in correlation of pitch discrimination and pitch
matching findings between this study and previous
research relates to different or multiple profiles of inaccurate singers. Bradshaw and McHenry15 studied
adults who sing inaccurately and did not find a significant correlation between pitch matching and pitch
discrimination. Their results indicated two profiles
of inaccurate singers. The two profiles were those
who were accurate pitch discriminators and poor
pitch matchers and those who performed poorly at
both tasks. They suggested that those who were
accurate pitch discriminators and inaccurate pitch
matchers lacked coordination and flexibility of the
vocal mechanism. Because pitch discrimination
underlies pitch matching, the limited pitch discrimination abilities of those who were inaccurate for
both tasks hindered their pitch matching ability. Of
the four participants who were inaccurate pitch
matchers, two participants were accurate pitch discriminators (less than 13 errors in 390 trials). The
focus of future research should include a greater
number of participants having inaccurate pitch
matching and discrimination abilities.
Journal of Voice, Vol. 22, No. 4, 2008

406

ROBERT E. MOORE ET AL

One implication of this study is that using samples of an individuals own voice for pitch matching
tasks may help differentiate the source of singing
inaccuracy (motor skills vs discrimination skills).
If an individual possessing accurate pitch discrimination ability performs better on pitch matching
tasks with their own voice than with tonal stimulus,
it may indicate a lack of flexibility or coordination
of the vocal mechanism required for accurate singing. For these individuals, teachers of singing might
focus on improving the motor aspects of singing
rather than improving pitch discrimination abilities.
Also, these individuals may have more potential for
singing improvement than individuals who perform
poorly with both types of stimuli. For those individuals with poor discrimination, regardless of stimuli
type, teachers of singing might first focus on
improving discrimination abilities because pitch
discrimination is a foundation of pitch matching
ability.
REFERENCES
1. Brashears S, Morlet T, Berlin C, Hood L. Olivocochlear
efferent suppression in classical musicians. J Am Acad
Audiol. 2003;14:314324.
2. Gaab N, Gaser C, Zaehle T, Jancke L, Schlaug G. Functional
anatomy of pitch memoryan fMRI study with sparse temporal sampling. Neuroimage. 2003;19:14171426.
3. Deutsch D. The deterioration of pitch information in memory [unpublished doctoral dissertation]. San Diego, CA:
University of California at San Diego; 1970.
4. Fairbanks G. Selective vocal effects on delayed auditory
feedback. J Speech Hear Disord. 1955;20:333346.
5. Siegle G, Pick H. Auditory feedback in the regulation of
voice. J Acoust Soc Am. 1974;56:16181624.
6. Siegle G, Kennard K. Lombard and sidetone amplification
effects in normal and misarticulating children. J Speech
Hear Res. 1984;27:5662.
7. Fairbanks G. Systematic research in experimental phonetics. I. A theory of the speech mechanism as a servosystem.
J Speech Hear Disord. 1954;19:133140.
8. Howell P, El-Yaniv N, Powell D. Factors affecting fluency
in stutters. In: Peters HFM, Hulstijin W, eds. Speech Motor
Dynamics in Stuttering. New York: Springer-Verlag; 1987:
361369.
9. Elman J. Effects of frequency-shifted feedback on the pitch
of vocal productions. J Acoust Soc Am. 1981;70:4550.

Journal of Voice, Vol. 22, No. 4, 2008

10. Brody VA. The emergence of song. Music Educ J. 1949;


36:2224.
11. Amir O, Amir N, Kishon-Rabin L. The effect of superior
auditory skills on vocal accuracy. J Acoust Soc Am.
2003;113:11021108.
12. Pitt M. Perception of pitch and timbre by musically trained
and untrained listeners. J Exp Psychol. 1994;20:976986.
13. Watts C, Moore R, McCaghren K. The relationship between vocal pitch-matching skills and pitch discrimination
skills in untrained accurate and inaccurate singers. J Voice.
2005;19:534543.
14. Watts C, Murphy J, Barnes-Burroughs K. Pitch-matching
accuracy of trained singers, untrained participants with
talented singing voices, and untrained participants with
non-talented singing voices in conditions of varying
feedback. J Voice. 2003;17:187196.
15. Bradshaw E, McHenry M. Pitch discrimination and pitch
matching abilities of adults who sing inaccurately. J Voice.
2005;19:431439.
16. Moore B, Glasberg B, Proctor G. Accuracy of pitch matching for pure tones and for complex tones with overlapping
and non-overlapping harmonics. J Acoust Soc Am. 1992;
91:34433450.
17. Singh P, Hirsh I. Influence of spectral locus and F0 changes
on the pitch and timbre of complex tones. J Acoust Soc
Am. 1992;92:26502662.
18. Jarvelainen H, Verma T, Valimaki V. Perception and
adjustment of pitch in inharmonic string instrument tones.
J New Music Res. 2002;31:311319.
19. Schueller M, Fucci D, Bond Z. Perceptual judgment of
voice pitch during pitch matching tasks. Percept Mot
Skills. 2002;94:967974.
20. Boone D, McFarlane S, Von Berg S. The Voice and Voice
Therapy. Boston: Allyn & Bacon; 2005. 201206.
21. Stemple J. Clinical Voice Pathology: Theory and Management. San Diego: Singular; 2000. 304310.
22. Mathieson L. The Voice and Its Disorders. London:
Whurr; 2001. 512554.
23. He N, Dubno J, Mills J. Frequency and intensity discrimination measured in a maximum likelihood procedure from
young and aged normal-hearing subjects. J Acoust Soc Am.
1998;103:553565.
24. American Speech-Language-Hearing Association. Guidelines for manual pure-tone audiometry. ASHA. 1978;20:
297301.
25. American National Standards Institute. Maximum Permissible Ambient Noise Levels for Audiometric Test Rooms.
ANSI S3.1-1991. New York: American National Standards
Institute; 1991.
26. American National Standards Institute. Specifications for
audiometers. ANSI S3.6-1996. New York: American
National Standards Institute; 1996.

PITCH MATCHING/DISCRIMINATION VOCAL/NONVOCAL STIMULI

407

APPENDIX A. SIXTY-FIVE COMBINATIONS OF PITCH DISCRIMINATION


TASK STIMULI SETS

F0/F0/F0
F0/50/F0
F0/100/F0
F0/50/F0
F0/100/F0
F0/F0/50
F0/F0/100
F0/F0/50
F0/F0/100
50/F0/F0
100/F0/F0
50/F0/F0
100/F0/F0

50/50/50
50/F0/50
50/100/50
50/50/50
50/100/50
50/50/F0
50/50/100
50/50/50
50/50/100
F0/50/50
100/50/50
50/50/50
100/50/50

100//100/100
100/F0/100
100/50/100
100/50/100
100/100/100
100/100/F0
100/100/50
100/100/50
100/100/100
F0/100/100
50/100/100
50/100/100
100/100/100

50/50/50
50/F0/50
50/50/50
50/100/50
50/100/50
50/50/F0
50/50/50
50/50/100
50/50/100
F0/50/50
50/50/50
100/50/50
100/50/50

100/100/100
100/F0/100
100/50/100
100/100/100
100/50/100
100/100/F0
100/100/50
100/100/100
100/100/50
F0/100/100
50/100/100
100/100/100
50/100/100

Notes: Each of the combinations was presented two times for each of three conditions (nonvocal complex tone, neutral female voice,
and own voice). 50 5 50 cents relative to F0; 100 5 100 cents relative to F0; 50 5 50 cents relative to F0; 100 5 100
cents relative to F0.

Journal of Voice, Vol. 22, No. 4, 2008