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

Re: vibrato

Laurent Demany wrote:
> Seashore's observation has been verified for very small, and rather rapid,
> vibratos: It is impossible to discriminate a just-detectable sinusoidal FM
> from a just-detectable sinusoidal AM if the modulation rate exceeds about 5
> Hz. But the discrimination is possible is the modulation rate is about 2 Hz.
> See:
>         Laurent Demany

Back in the old days, following the first measurements of the pure tone
JND by Shower and Biddulph (JASA 1931), Fletcher integrated the JND to
recover the place-frequency map. He found the same map as obtained by
the integration of the critical ratio data, except the frequency JND data
was shifted by about 1/2 octave. He also showed that the frequency JND
was equal to 1/20 of the critical band. He concluded that you do the freq. JND
based on the high-freqency slope, as amplitude discrimination detection.
This is clearly the story above a few kHz, as if you rove the intensity,
the freq. JND seriously degrades (Moore). At low frequencies it would be
possible to use the phase-locked signal to estimate the pitch. I think
the data suggested that this might be the case below a few hundred Hz.
I am not sure about these details however. Perhaps someone could fill us
in on these. This is described on page 169-175, and page 150-152 of his
1950 book. (You can buy it from ASA for < $50, which is a bargain. Sorry
for the small ad.)

1) How do people view this result today?

2) This story seems to be in agreement with the above observation described by

Jont B. Allen, Room E161
AT&T Labs-Research 180 Park AV.
Florham Park NJ 07932
973/360-8545voice, x8092fax