Re: front to rear reversals

[William L Martens <wlm@xxxxxxxxxxxx>]

With regard to Tim Cox's observation on front to rear reversals
in binaural recordings, I have made this distinction:

Discussion of the problem of front/rear reversals is clarified if
the distinction is made between the subjective experience of the
observer and the overt behavioral responses that the observer makes
in listening experiments.  When the observer is forced to make a choice
between two response categories for free-field source incidence angle,
front versus rear, then it is possible to have front/rear reversals.
In studies of this sort (e.g., Blauert, 1969/70) there is a correct answer,
as the sound source is actually located to the front or the rear of the
observer.  When, however, observers are asked to examine the subjective
experience they have during earphone listening, it is no longer appropriate
to use the term front/rear reversal, as the true report on the observer's
auditory spatial image is the only correct answer.  Of course, if a frontally
positioned sound source is recorded binaurally, we can ask earphone listeners
to make a choice about where they think the sound source was during recording;
but this is a different question than that regarding where the auditory spatial
image is actually located in the observers perceptual space.

For many observers a dry (non-reverberant) binaural recording of
a frontally-positioned sound source will result in a spatial image that
is located intracranially (inside the head).  This phenomenon underscores
the importance of clearly stating the nature of the judgment that the
observer must make in a spatial listening task.  Much of the
spatial hearing literature employing earphone presentation obscures this
issue by calling judgments about auditory images localization judgments.
The distinction must be made between the observer's localization performance
and the observer's ability to report the direction and distance
of an auditory image experienced during earphone listening.
Even when broadband stimuli are reproduced via precisely calibrated
earphones, there is no guarantee that veridical perception will result.
One reason is that veridical perception of sound-source location often is
not observed in free-field listening.  This observation and others led Shaw
(1982) to propose that earphone studies of spatial imagery be termed either
``space perception'' or ``sound localization.''


--
William L. Martens, Ph.D.             Associate Professor
Multimedia Systems Lab                URL: http://www.u-aizu.ac.jp/~wlm
University of Aizu                    TEL: [+81](242)37-2791
Aizu-Wakamatsu  965-8580, Japan       FAX: [+81](242)37-2731