Group delay variations

[Pierre Divenyi <pdivenyi@xxxxxxxxxxxxxxxxxxxxxx>]

This is getting interesting. Jont says that monaural and binaural phase
sensitivity are essentially the same, provided we look at (1) group delay
and (2) the variations therein.
Roy says in his reply that in his 1987 paper he showed that the
(monaural) group delay perception is proportional to bandwidth. Thus, it
appears that there is a Gabor-type uncertainty principle operating here:
the product of delay and bandwidth look roughly constant. 

Interestingly, Chandler and Grantham, in their 1992 paper, showed just
that: the just-noticeable angle of the (azimuthal) motion of a sound
source is inversely proportional to the bandwidth of the source. Note
that this was observed for MOVING sound sources, i.e., for changing group
delays only. In the proceedings of the Carcans ISH meeting (1991), Pierre
Zakarauskas and I have a paper in which a similar azimuth/bandwidth
inverse relation is derived.

Pierre Divenyi

Reference:
Divenyi, P. L. and P. Zakarauskas (1992). The effect of bandwidth on
auditory localization: An estimation theory model. Proceedings of the
12th International Symposium on Hearing. L. D. a. K. H. Y. Cazals.
London, Pergamon Press: 563-570.