[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

carrell & opie, summerfield & culling, carlyon,darwin,&russell

Al Bregman replies to my comment that Summerfield and Culling found no effect
of differential AM rate on the identification of vowel pairs:
" If the vowels used by Summerfield and Culling were either natural
or synthesized in the normal way, they already had an AM due to the
glottal pulse, even before the addition of AM.  This may have
interfered with the effects of whatever AM was added
(presumably low-frequency AM).  That's why the sine-wave speech
signal is so useful.  It has no inherent AM"

Summerfield and Culling's vowels were synthesised in the normal way,
so there was indeed some AM at a rate corresponding to the F0 (100 Hz).
However, their finding (of no effect of differences in low-freq AM rate)
was obtained both when the vowels were on the same and on a different F0.
Also, note that there is some evidence that the auditory system is tuned for
AM rate (e.g. Houtgast, jasa, 1989). Although this "tuning" is fairly broad,
it should be sufficient to preclude interference between, to cite one of
Summerfiel and Culling's conditions, a 4-Hz AM rate and a 100-Hz glottal pulse
rate. Note also that
summerfield and assmann have shown that the synchrony of glottal
pulses has no effect on performance on this task, for F0s of 100 Hz and above.
(I have shown, too, (in a article soon to be submitted to JASA) that listeners
are not very sensitive to pitch-pulse asynchrony between pairs of formants,
for F0s above 100 Hz).

Generally, the point of my comment was to alert potential experimenters to
the existence of a relevant study, rather than to stifle an interesting line
of research. I suggest that those planning experiments in this area,
or are just interested in this topic read the article, or, preferably, buy the
book (details of which Iwill make available later this month when published!)

bob carlyon