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

Re: Pitch orientation-discriminating feature detectors?

Thanks for these sweeping comments... I'll try to provide a couple of
precisions where I can.

Al's comment:

> On what scale were the frequency excursions?  If it was a
> log-frequency scale,

A justification for the scale, as Laurent pointed out, would be that
you cover approximately the same distance on the basilar membrane with
a log scale. EEG studies I've seen nevertheless used a linear scale
and they still found different responses to upward and downward

Pierre's comment:

> Question re/ your new fMRI experiments on frequency change: What
> cortical structures are you looking at?

We are doing MEG at the moment and we only tried a 2 dipole model
--you make the simplifying hypothesis that the magnetic field observed
comes from 2 focal sources. The sources were where you would expect
Heschl's gyrus, but we'll need the anatomical MRI for each subject to
say something more interesting about the structures involved.

About the relatively obscure reference, the work by Ruhm has actually
been repeated and expanded by others. For instance, see:

  author =       {Kohn, M. and Lifshitz, K. and Litchfield, D.},
  title =        {Averaged evoked potentials and frequency modulation},
  journal =      {Electroenceph. clin. Neurophysiol.},
  year =         1978,
  volume =       45,
  pages =        {236-243}

Eckard's comment:

I'd be more than happy with a peripheral explanation for the
phenomenon, but I still don't feel there is a satisfactory one
around. It cannot be masking by basilar membrane spread of excitation,
because it would go in the wrong direction: you would expect more
masking for peaks than for troughs. The travelling wave delay is
rather small in comparison to the rate of FM over which the perceptual
asymmetry can be observed. I'm still not sure how neurotransmitter
release might do the trick.

Martin's comment:

In another unprobable reference, Horii (1989) observed that singers
usually start their vibrato with an upward and not a downward
glide. Possibly this has to do with muscular tension.

  author =       {Horii, Y.},
  title = {Frequency modulation characteristics of sustained /a/ sung
  in vocal vibrato},
  journal =      {Journal of Speech and Hearing Research},
  year =         1989,
  volume =       32,
  pages =        {829-836}

Wolfgang's comments:

Thanks for pointing out this paper. If I'm not mistaken, the effect
you found was only present for small frequency deviations. In
contrast, the peak/trough asymmetry can be seen for deviations larger
than one octave. This does not mean that the kind of across-channel
combination of information that you put forward cannot occur in such a
case (I believe it can) but this still needs to be formalised.