Speechmap Technical Details; Wideband Measurements In Speechmap; Speechmap Stimuli - audioscan verifit User Manual

9 Speechmap Technical Details

The following sections outline the principles underlying Speechmap stimuli and measurements, especially for
wideband tests.

Wideband measurements in Speechmap

The availability of hearing instruments claiming to provide amplification above 8 kHz, and the reported benefit
of those wider bandwidths for young children has motivated Audioscan to extend the bandwidth of the Verifit to
16 kHz which encompasses the 12.5 kHz third octave band. This necessitated several significant advances:
1. Microphones with a flat response to 16 kHz and a low noise floor and development of production calibration
processes.
2. A 0.4cc coupler to overcome standing wave problems in the 2cc coupler and increase the signal level.
3. Digitized speech stimuli with measurable energy to 16 kHz.
4. A Wideband RECD (WRECD) measurement protocol based on the 0.4cc coupler and improved to reduce
errors associated with the effects of earmold and coupler tubing.
5. Provisions for threshold and WRECD entry to 12.5 kHz.
6. Conversion of entered and normative RECD values to WRECD values for frequencies below 8 kHz.
7. Derivation of Microphone Location Effects (MLE) to 12.5 kHz.

Speechmap stimuli

Speechmap provides a variety of stimuli for the testing of non-linear hearing aids with a range of digital
processing features.
The Speechmap fitting environment provides a number of well-controlled, distinctly different stimuli for both
On-ear and Test box measurement modes. Several different digitized speech passages are provided as well as
the International Speech Test Signal (ISTS) and the single-talker International Collegium of Rehabilitative
Audiology (ICRA) distorted speech signal. Two speech passages have been filtered to provide the same long-
term average speech spectrum (LTASS) as the ISTS for average vocal effort. In addition, one of these passages
has been filtered to suppress the 1/3 octave bands above 1 kHz by 30 dB with the exception of a 1/3 octave band
at 3.15 kHz (Speech 3150), a band at 4 kHz (Speech4000), a band at 5kHz (Speech5000) and a band at 6.3 kHz
(Speech6300). These latter signals may be used to determine the amount of frequency shift provided by
frequency-lowering hearing aids and the sensation level of the lowered components of the speech signal. Pink
noise at levels of 50 to 80 dB SPL in 5 dB steps and a tone burst at 85 or 90 dB SPL are also available. The
ICRA noise, female and child speech signals are presented as recorded with no spectral shaping.
The ISTS consists of 500 ms segments from recordings of 6 female talkers reading the same passage in
American English, Arabic, Chinese, French, German and Spanish. These segments have been spliced together
with appropriate pauses and filtered to match the average female spectrum from Byrne et al, An international
comparison of long-term average speech spectra. J. Acoust. Soc. Am. 96 (1994), 2108-2120. The 15 second
version of this signal has been provided.
Live speech may also be used.
The speech, ISTS and ICRA signals are presented unfiltered for levels from 50 to 70 dB SPL. At 75 and 80 dB
SPL, the spectrum is filtered to represent a 'loud' vocal effort. An 'own voice' filter is also available at the 75 dB
SPL level. This is intended to produce a spectrum and level at a BTE microphone location similar to that
produced by the voice of an average wearer.
Verifit ® User's Guide Version 4.8 © November 2016
69