Transcript Alternative Assessments for Musicians

Alternative Assessments
for Musicians
Marshall Chasin,
AuD., M.Sc., Reg. CASLPO, Aud(C)
Musicians’ Clinics of Canada
Why do any tests?
 Hearing
loss prevention
 Assessment
of hearing aids for music
 (Assessment
of musicians)
Hearing Loss Prevention
 Audiometric

High frequency testing
• Actual bandwidth of hearing
• Notches in hearing
• Up to 12kHz vs. higher…
 Non-audiometric


OAE (TTS analog)
Speech in noise tests
Audiometric
 High




frequency artifacts
Stinson (1988, JASA): minimal artifacts up to
12kHz but large variance above this.
Minimized with proper earphones that have
both a smooth frequency response and wide
bandwidth.
Monitor at TM with probe microphone to verify
calibration.
Good test-retest reliability WITHIN subjects…
From: Nancy E. Vaughan, et al. 2002, Journal of Res. Rehab. Development
(within subjects)
The right earphone for the right
frequency…
The 4000 Hz notch
 Results
from a 28 mm long ear canal that
functions as a quarter wavelength
resonator…
F
= (2k-1) v/4L
 Odd

number multiples of 4000 Hz…
12kHz, 20kHz,…
The 4000 Hz notch
 Even
though the ear canal is on average,
28 mm…
F = 340,000 mm/sec/4x28 mm = 3035 Hz
But… F = 2700 Hz
The eardrum adds several mm of acoustic
length because of its compliance. (3-4 mm)
The 4000 Hz notch
 BUT,
for higher frequencies, the eardrum
does not supply the extra several mm of
acoustic length, so that odd number
multiples of the 4000 Hz notch is really
higher than 12kHz (… more like 1415kHz).
High frequency audiometry


So,…. With the proper earphones and within
subjects, one can reliably test in the higher
frequency range.
(1) higher frequency notches (14k-15kHz) may
occur prior to the 4000-6000 Hz notches.
 (2) 10k-12k thresholds may help identify
etiology. (eg. Violins have an 8kHz notch).
 (3) good reliability within subjects (longitudinal).
Ear Q software
 Ear
Q is one of many software products
that can allow your lap top to become a
(high frequency) audiometer…
High frequency audiometry
 Interesting
trivia…
 Ashihara
(2007) JASA: humans can hear
up to 28kHz.

As long as presentation level is 100 dB SPL.
Non-audiometric tests
 OAE

Poor test-retest reliability between subjects.

But, good test-retest reliability within subjects.
• Like high frequency audiometry, can be used for
longitudinal testing of differences within subjects.
OAEs
 Can
be used to assess temporary
changes (eg. TTS).
 Can
be used to assess deterioration over
the years.
 Can
be used to test non-audiometric
frequencies.
Otoacoustic Emissions
(Drummer - symmetrical “hearing”)
Disortion Product OAEs for Mr. WJ.
30
dB SPL
20
10
Left ear
0
-10 0
Right ear
2000
4000
6000
-20
Frequency (Hz)
8000
Otoacoustic Emissions
(Lead singer moved away from
drummer)
Mr. GJ initially and after 6 month
recheck
Left ear initially
dB SPL
20
10
0
-10 0
5000
-20
Frequency (Hz)
10000
Right ear initially
Left ear - after
6 month
recheck
OAE “notch” at 5000 Hz and not
measured on audiometry...
OAE Loss at a Non-Audiometric Frequency
15
10
dB SPL
5
4 weeks after "TTS"
0
0
1000
2000
3000
4000
-5
5000
6000
7000
2 weeks after "TTS"
Day after "loud gig"
-10
-15
-20
Frequency (Hz)
What about assessment of hearing
aids for music?
 “Traditionally”






one looks at pre-selection…
Wide bandwidth
WDRC
Appropriate compression
High input limiting level or equivalent
Disabled feedback and noise management
6 dB lower gain and output for same level
but,… how do you evaluate aids?
 (1)
Ensure that electro-acoustic settings
and parameters are correct.
 (2) An
assessment scale (like the AI)….
Development of an assessment
scale…
 Gabrielsson
 Cox
et al. (1974, 1991)
and Alexander (1983)
 Chasin
and Russo (2004)
Development of an assessment
scale…
 Loudness-
music is sufficiently loud
 Fullness- music is full (vs. thin)
 Crispness-music is clear and distinct (vs.
blurred or diffuse)
 Naturalness- as I remember it (as if there
was no aid there)
 Overall fidelity- dynamics not constrained
Development of an assessment
scale…

Cannot develop an AI (or SII) for music since
there is no long term spectrum as there is for
speech.
 Can develop correlations (and regressions)
between these five outcomes measures and
various bands of music.
Successive low pass filtering and the five
perceptional parameters will be assessed at
each of the cutoff frequencies of the LP filters.
Development of an assessment
scale…
 Goal:
determine which perceptual
parameters (or combination of
parameters) correlates with which
frequency band, similar to an SII measure.