Re: voice tracking

[beauchamp james w <j-beauch@xxxxxxxxxxxxxxxx>]

Dear Piet Vos,

Thanks for the interesting question. It is true that we hear the
higher voices best, but it may be that we hear best in the range
1000 - 3000 Hz -- this certainly shows up on equal-loudness curves.
Male tenor voices certainly do very well against the orchestra,
and analysis shows that they project a lot of energy in a range
around 2000 Hz. Cellos and trombones also due very well.

Finally, I note three phenomena that favor the lower notes:
First we tend to associate pitch with the fundamental rather than
the harmonics, but high energy harmonics can enhance the voice's
ability to stand out. Second, although soprano notes are easiest to
hear, bass notes may be the second easiest to hear, whereas inner
voices, with their increased masking, tend to blend together. Third,
if a given voice is performing a familiar melody, that will greatly
its chance of being heard out.

On a different topic, how does one join Psymus? I hope the traffic
is not too great.

Jim Beauchamp
University of Illinois at Urbana-Champaign

On July 2, 2002, you wrote:
> Hi,
>
> Could anybody tell me about the availability of empirical/experimental
> data about the following phenomenon. When listening to plural voiced
> music, say a simple choral or hymn, it is quite easy, if not natural /
> default, to track the soprano voice, whereas it requires attention
> (effort, training..) to track lower voices. Similarly so, there exist
> much more concerto's for a high pitched solo instrument (eg violin) and
> orchestra than for combinations with low-pitched instruments. It's
> natural to suppose that the reasons are eventually found in
> psycho-acoustic constraints such as the anisotropic aural sensitivity
> for auditory frequencies and masking effects. This question was recently
> addressed also to the Psymus list, dedicated to music perception and
> performance issues, and received a few relevant answers on the possible
> mechanisms behind the phenomena, but is was suggested that no empirical
> research data are available.
>
> Thank you in advance for your help,
>
> Piet Vos
> --
> Piet G. Vos
> section Perception NICI, U. Nijmegen
> P.O.Box 9104
> 6500 HE Nijmegen NL
> tel: +31 24 36126 31/20; fax: +31 24 361 60 66; vos@nici.kun.nl
> home-page: http://www.nici.kun.nl/~vos
>
>