Re: pitch neurons (Eckard Blumschein )

Subject: Re: pitch neurons
From:    Eckard Blumschein  <Eckard.Blumschein(at)E-TECHNIK.UNI-MAGDEBURG.DE>
Date:    Tue, 8 Oct 2002 08:59:13 +0200

At 11:29 07.10.2002 -0400, Allen G Lindgren wrote: >Eckard >Can you elaborate on "Merely below about 1kHz, >> synchrony can be very high.". Yes I can: I wrote: >> On average, ten auditory nerve fibers cooperate statistically with just >> one inner hair cell. Each fiber is bound to refractory time. So roughly >> speaking, all ten together cannot continuously phase lock with intervals >> smaller than a tenth of refractory time. Merely below about 1kHz, >> synchrony can be very high. >> This limitation due to statistic time sharing does, however, not exist >> immediately after nearly all fibers reached a state of readiness for >> firing. >> Given frequency of modulation or rate of clicks is low enough for that, >> then the stationary upper frequency limitation is much less relevant >> for the whole temporal structure, including interaural envelope delay >> (IED), during a split millisecond of adaptation. I refer to Eric D. Young who wrote chapter 4 of 'Synaptical Organization of the Brain'. Fig. 4.11 shows synchrony in the ANF not much deviating from the value 0.8 up to 1 kHz but then increasingly dropping and reaching zero at 4.5 kHz. Merely synchrony of so called primarylike VCN neurons can approach the value one. We realize these limitations like the optimal ones if we understand how the random cooperation of the ten fibers works. Albert Einstein said: 'God does not play dice'. With respect to causality he was probably correct. However, we have to learn that stochastic cooperation of neurons must not be ignored. Forget the illusion that Gammatone filters adequately describe what happens. Just some food for thought: How to correctly include adaptation in the models? Why is too much synchrony a problem for cochlear implants? How to explain why 'available descriptions of musical instrument tones must be considered inadequate, because they fail to pass the foolproof synthesis test' (quoted from p. 117 in 2nd ed. of D. Deutsch's 'the Psychology of Music')? How to answer Yost's three questions on p. 81 of the same book? In reply to Matt, I would like to add a new aspect: Matt reminds of dual perception. It's well known that apparent frequency doubles with transition from tonotopic half way rectification to full wave rectification of the envelope. To what extent does this effect contribute to octave ambiguity of pitch? Matt, you believe that efferent return signals phase out afferent ones via the auditory nerve. I consider that unrealistic. Also, I cannot entirely agree with Jean-Sylvain Lienard. Periphery definitely performs more than just 'F0 measurement'. Eckard

This message came from the mail archive
maintained by:
DAn Ellis <>
Electrical Engineering Dept., Columbia University