Re: Stimulus properties of sound streams

[Al Bregman <BREGMAN@xxxxxxxxxxxxxxxxxxxx>]

Dear Jan,

I have inserted my replies between your questions.  They aren't
based on specific pieces of research but on my own expectations
(and sheer guesses) after listening to many types of laboratory
sounds.

>   1.. what properties of the sound streams make it easy or
difficult for the
> listener to pre-attentively perceive a change in the display,

An abrupt increase in loudness should facilitate the registration
of the change; however, it might also attract attention to
itself, so that it didn't remain pre-attentive.  Of course, the
greater the magnitude of the change, the easier it would be to
pre-attentively register the change.  Also changes in louder
streams should be most easily registered.  Finally, the greater
the magnitude of the difference that segregates one stream from
the others, the easier it should be to register a difference in
that stream.


>   2.. what properties facilitate the voluntary direction of
attention to the
> changing auditory dimensions within a stream, for eg. pitch
and/or tempo, so
> that information about the variable of interest can be
extracted.

If there are several streams, each with a different pitch range
or frequency range, changes in the highest should be more
noticeable, followed by changes in the lowest, then by changes in
the steams in the intermediate range (such as the inner voices in
music (you may want to ask Caroline Palmer about this question:
cpalmer@magnus.acs.ohio-state.edu).

> I am wondering whether anyone has some thoughts about whether
auditory
> stimulus dimensions can be integral or separable, along the
lines of
> Garner's (1970, 1974)classification of visual stimuli. Integral
dimensions
> facilitate tasks where both dimensions are attended to, but
interfere with
> tasks where one dimension is attended to. Separable dimensions
have the
> opposite effect. While there are some reports in the literature
of integral
> auditory dimensions such as pitch and loudness, I haven't seen
any reference
> to separable dimensions, and in fact I'm wondering whether this
is a concept
> that could be applied to sound stimuli.

I would imagine that the length of a sound (where the length is
at least 400ms, to exceed the temporal integration window for
loudness) would be separable from any steady-state property of
the sound, such as pitch, loudness, or timbre.

> The other concept I'm wondering
> about is auditory emergent features. Visual emergent features
have a strong
> facilitative effect on monitoring performance - what would an
auditory
> emergent feature within a sound stream be like?

A phoneme would be an emergent feature arising from the pattern
of intensities of frequency components, whether or not the
components are harmonic, inharmonic, or random (as in whisper).
An example of an emergent feature in music is whether the chord
is major or minor, or, indeed, any chord quality.  A stream is
also an emergent entity ("entity" rather than "feature").
Corresponding with visual shapes, which are emergent, are
melodies.  Timbral "brightness" can also be considered to be an
emergent feature, arising from the relations among partials, with
no partial essential in producing the brighness.

Al
-------------------------------------------------
Albert S. Bregman, Emeritus Professor
Dept of Psychology, McGill University
1205 Docteur Penfield Avenue
Montreal, QC, Canada H3A 1B1

Office:
     Phone:  +1 (514) 398-6103
     Fax: +1 (514) 398-4896
Home:
     Phone & Fax: +1 (514) 484-2592
Email:
     bregman@hebb.psych.mcgill.ca
-------------------------------------------------