Altered auditory feedback with a biophysical model for sound production

A. AMADOR, J. D. SITT, F. GOLLER AND G.B. MINDLIN.

Chicago

Simposio; Tucker-Davis Symposium on Advances and Perspectives in Auditory Neurophysiology.; 2009

Songbirds have been widely studied as animal models for exploring sensorimotor plasticity in a well defined neural circuit critical for learning and memory. The parallels to human vocal development and sound production mechanisms are remarkable. Male zebra finches (Taeniopygia guttata) learn song during development, resulting in a stereotyped adult song that, in normal conditions, does not subsequently change. Altered auditory feedback experiments have shown that song learning and adult maintenance is an active process, but little is known about the mechanisms of auditory feedback evaluation.In the last years, a number of mathematical models have been proposed in order to integrate a large body of experimental work with the expected mechanical processes involved in birdsong production.However, the mechanisms of how the avian sound source might contribute to spectral richness are largely unknown. In order to address this question, we analyze the zebra finch song, that presents syllables with low fundamental frequencies that are spectrally rich, and syllables with high fundamental frequency that are nearly tonal. In this work we present the study of a nonlinearly enhanced flapping model for birdsong production. The characteristics found in the zebra finch song can be synthesized using as input for the model physiological measurements recorded while the bird is singing. Using this model we can generate realistic synthetic song while the bird is singing. In this way, altered auditory feedback experiments can be performed as the acoustic properties of the synthetic song are controlled by parameters in the model.