audioscan RM500SL User Manual page 102

VERIFIT and RM500SL Test Signals and Analysis
1 Introduction
The Verifit and RM500SL are full-duplex
dual-channel audio measurement systems
designed for the testing and fitting of all
types of hearing instruments and many
assistive listening devices. They provide
an acoustic test chamber for the testing of
devices coupled to standard 2 cm
couplers as well as sound-field speaker(s)
for real-ear measurement (REM) of hearing
aid performance. The Verifit test chamber
contains two loudspeakers for the
functional testing of directional hearing
instruments. Test signals are delivered to
the device under test via the test chamber
loudspeaker(s), the test chamber telephone
magnetic-field simulator (TMFS), the test
chamber magnetic field test loop (Verifit
only) or the sound-field loudspeaker(s).
Device output signals are measured in the
3
2 cm coupler via the coupler microphone or in a real ear via the probe microphone. Data from the 2 cm
may be viewed as coupler SPL, coupler gain or as estimated SPL in the ear canal (simulated real-ear
measurements, S-REM). Data from the probe microphone may be viewed as ear canal SPL, as ear canal SPL re
normal hearing threshold (i.e. dB HL), as insertion gain or as ear canal SPL in the context of an auditory area
®
(Speechmap ).
2 The test signals
Four types of test signals are currently available. These are tonal signals, pseudo-random noise, digitized real
speech signals and the ICRA distorted speech signal. Test signals are generated in real time by the digital signal
processor (DSP) or played from 16 bit binary audio files stored in the on-board flash memory. In the Verifit, these
audio files are up-loaded from the internal compact disc (CD) drive at power up. In order to provide a repeatable
acoustic signal to the device under test, a reference microphone is used in conjunction with a digital control loop
to maintain the desired band level at each frequency. Live audio (speech, music etc.) may also be used as a test
signal but it is not controlled for spectrum or level.
2.1 Narrow-band signals
2.1.1 Pure tones
Pure-tones are used in the ANSI hearing aid tests, in Multicurve 2 cm
3
in 2 cm and real-ear single frequency (manual) tests, in real-ear insertion gain tests and in tone-burst maximum
output (MPO) tests. Frequencies used are 1/12 octave except for the MPO test which uses 1/3 octave
frequencies. Levels from 40 to 90 dB SPL are available at the reference microphone. Control of the level at each
frequency is maintained by measuring the frequency response of the signal path to the reference microphone 256
ms before each test using a 256 ms, 50 – 55 dB SPL, multi-tone complex or a 256 ms chirp, depending on the
test to follow. Drive levels at each frequency are then set to achieve the desired band levels at the reference
07/05/06 © Etymonic Design Incorporated, 41 Byron Ave., Dorchester, ON, Canada N0L 1G0
USA 800-265-2093 519-268-3313 FAX 519-268-3256 www.audioscan.com
Ø
Overall SPL is the SPL in a band containing all significant frequency
components of the signal.
Ø
Spectrum level is the SPL in a band 1 Hz wide.
Ø
Band SPL is the SPL in a restricted frequency range. If the signal is
uniform in the band, band SPL = Spectrum level + 10*log(bandwidth).
Ø
A spectrum is the band SPL, or spectrum level, in a series of
adjacent bands.
3
Ø
For a broad-band signal, the overall SPL is greater than the band SPL
and the band SPL is greater than the spectrum level. For a pure tone,
the overall SPL, the band SPL and the spectrum level are the same.
Ø
A white noise signal has a spectrum level that is independent of
frequency (i.e. constant SPL per Hertz).
Ø
A pink noise signal has a spectrum level that is inversely proportional
to frequency (i.e. constant SPL per octave).
Ø
A fractional octave band (e.g. 1/3 octave) has a bandwidth that is
proportional to frequency.
Ø
A pink noise signal has a flat spectrum when analyzed in fractional-
octave bands.
Ø
A critical band is a band within which the loudness of a continuously-
distributed signal of constant SPL is independent of bandwidth.
S
ome useful facts about broad-band signals
3
frequency response tests (gain or output),
Bill Cole BASc, PEng
3
coupler
Page 1