I am looking for some help, or a pointer to a source of help, for a problem
that seems not at all straightforward. Sorry to all those who hate
I am trying to improve the intelligibility of speech for my mother who is
now 88. She has had a gradually worsening high frequency loss for many
years but only in the last 3 years has this been a problem for her.
Interestingly the problem came into focus by accident when she told me that
the top notes on an electric piano I had loaned her seemed not to be
Without detailing her audiogram at this stage, she has hearing in the lower
third of 'normal range' below 650Hz in her right ear and then a sharp 30dB
fall-off followed by a progressive hf loss from around 900Hz. Her left ear
is similarly 'normal' below 1350Hz and then follows the same pattern as the
right - a sharp 30dB fall-off followed by progressive hf loss.
She has two NHS bte analogue aids which have been adjusted to an optimal
setting [albeit this is still only a crude approximation to the correction
her audiogram indicates] They do not help her at all with speech
recognition and are a constant frustration to her in consequence of which
they are little used. Last year I purchased for her a good pair of Beyer
circumnaural open headphones which she uses with both the tv and radio.
These delight her because without any frequency correction at all and with
a very moderate volume level they
allow her to understand all but certain poor movie soundtracks. They
improve her speech intelligibility far more than they ought to and far more
than I can account for in any conventional mechanistic sense.
Now comes the problem I am wrestling with: having tested her extensively
with recorded speech from radio and tv I have been unable to determine the
common factors making an individual either easy or difficult to understand.
Low pitched, slow speaking, clearly articulating males would be expected to
be the easiest and in many cases are but not consistently so. Similarly,
high pitched fast talking females would be expected to be difficult and
usually are but again not always.
I am an EE by training and had an interest in all matters audio all my
life. I have tested my mothers hearing extensively and have a very
detailed audiogram for her. I have corrected a number of 'difficult'
speech samples with FFT on my computer to exactly correct for her hearing
loss. This does not seem to help and I am left wondering why when speech
is corrected such that the frequency response matches 'normal' across the
speech spectrum [below 6000Hz] it is not perceived as normal by the person
with the hearing loss.
Could it be that a gradual hearing loss, particularly in older people, is
accompanied by a loss of ability in the brain to interpret the incoming
stimulus? Clearly my mother can hear the high frequencies in the corrected
speech samples because she asks me 'what are all those funny clicks' to
which I reply those are all the consonant 't', 'c' and 'p' sounds that you
cannot usually hear.
Any advice on how to pursue this would be much appreciated.