Recording questions

[Yu-Hsiang Wu <yu-hsiang-wu@xxxxxxxxx>]

Dear List,

I am trying to record real-world sounds for hearing aid testing and am 
wondering if anyone can help me out. 

Specifically, I used a manikin (KEMAR) and the G.R.A.S. RA0045 ear 
simulator to record street noises. The recorded sound will be presented to 
subjects in the lab via earphones. My goal is to accurately replicate the 
sound (especially in sound level) so that the listener is seemingly placed 
on the street.

These are what I did.
1. Record the street sound (let’s call this sound is Input A) using the 
KEMAR and the RA0045. The KEMAR’s ears are not occluded.  The recorded 
sound is called Output A.
2. Process Output A (e.g., filtering). Let’s call the processed Output A as 
Input B. 
3. Present Input B to the KEMAR through earphones (Sennheiser IE 8 ear-
canal headphones), which are inserted into KEMAR’s ears. Record the sound 
using the same RA0045. The recorded sound is Output B.
4. Compare Output B to Output A. Repeat steps 2 and 3 until a filter bank 
that makes Output A and Output B have the same level and spectrum is found.
5. Use this filter bank to process Output A and present it to subjects.

I think these procedures could make earphones transparent because the 
coupler microphone generates the same output for sounds from the street and 
for sounds from the earphone.

However, when I was listening to the processed sound through the earphones 
(step 5), my perception was that the sound delivered from the earphones was 
louder than the sound that I heard on the street. Any comments?  Thanks in 
advance. 

Yu-Hsiang Wu
University of Iowa