Abstract Calculating Tonal Fusion by The Generalizes Coincidence Function.pdf

1 Calculating Tonal Fusion by The Generalizes Coincidence Function Martin Ebeling mar.ebeling@arcor.de Abstract 1. Background Based on the frequency analysis of the cochlea and the volley principle, as Wever (1949) described it, periodic sound is represented in the auditory system by periodic firing patterns. Cariani Delgutte (1996) have shown that the distribution of interspike intervals (ISI) between all spikes in the auditory nerve provide a code of acoustical information. A time series analysis of neural spike trains leads to autocorrelation histograms that show maxima for periods corresponding to the perceived pitch. Tramo, Cariani, Delgutte Braida (2001) found out that consonant intervals evoke periodic firing patterns in the auditory nerve, whereas dissonant intervals show irregular spike trains without periodicity. Since Licklider (1951), a neuronal autocorrelation mechanism has been regarded as a possible basis of pitch perception. Licklider’s model still was highly speculative, but Langer Schreiner (1988) have really found a bank of neural circuits to perform a neuronal periodicity analysis which corresponds to the mathematics of autocorrelation. This neuronal periodicity detection is performed in the time domain by a neuronal delay mechanism and a coincidence detection in the colliculus inferior. 2. Aims In a neuronal periodicity analysis mechanism common periods of the two tones of an interval are detected as coinciding pulses. The mathematical calculation of spike trains by autocorrelation functions shall help understanding the logical structure of coinciding neuronal firing patterns that represent intervals in the auditory systems. A measure value for the degree of coincidence shall be defined to answer the question whether neuronal coincidence may lead to consonance. 3. The mathematical model of periodicity detection of intervals To form a mathematical model of neural coding and autocorrelation for perio