19 Jan
2010
19 Jan
'10
5:57 p.m.
Simon Fielding wrote:
My sister is a specialist teacher for hearing-impaired
primary school
children. As part of her curriculum she includes music and in
particular, nursery rhymes etc for the younger children. She would like
them to be able to sing these at home with and/or for their parents. For
those children with non-hearing-impaired parents, this is not a problem
but many of the children also have hearing-impaired parents. Therefore
she would like to produce a CD of her singing for the children to take
home and use. I would really appreciate any help members can give me as
I am not a professional musician or recording engineer.
1) She will be singing unaccompanied (she is a trained singer and is
perfectly competent to do this) in an alto register to avoid any
distraction for the children from accompaniments etc. Does anyone have
any suggestions about this? (eg effects for recording, effects for the
headphone mix etc)
For most hearing impaired listeners, less reverberation is usually
better (if understanding the words is desirable). The effect of
compression very much depends on the accompaniment and/or distracting
noises; in a clean situation compression usually makes things worse (see
e.g., R. Plomp: The negative effect of amplitude compression in
multichannel hearing aids in the light of the modulation-transfer
function. J. Acoust. Soc. Am., vol 83(6), 1988). Try to avoid pops
during recording.
2) This one's a bit more specialist so you may not be able to help -
hearing impairment often starts with loss of high frequency response.
The obvious thing would seem to be to boost these but I don't know if
that would be correct. Does anyone know?
If they will listen via loudspeaker and wear their hearing aids
(assuming that they have one), all individualized frequency shaping is
already done by the hearing aid. Listening via headphone is not really
working with hearing aids, thus an individual frequency shaping might be
beneficial. However, for that you need to know a bit about the
individual hearing (e.g., hearing threshold and whether it is
sensorineural or conductive loss, as a bare minimum). A very simple rule
of thumb is to apply a frequency dependent gain which is 40% of hearing
threshold for sensorineural loss and 100% of threshold for conductive loss.
Linux audio tools can be used to do batched pre-filtering of audio
material based on individual hearing loss (e.g., the command line tool
'applyplugin' which applies LADSPA-Plugins to files, can load equalizers
and set their gains).
Best regards,
Giso
Any comments on any other aspects of this project would also be more
than welcome.
Hopeful regards,
Simon
_______________________________________________
Linux-audio-user mailing list
Linux-audio-user(a)lists.linuxaudio.org
http://lists.linuxaudio.org/listinfo/linux-audio-user