Re: Discussion about "place pitch" concept (Pierre Divenyi )


Subject: Re: Discussion about "place pitch" concept
From:    Pierre Divenyi  <pdivenyi(at)EBIRE.ORG>
Date:    Fri, 19 Mar 2004 08:11:33 -0800

--=====================_78702312==_.ALT Content-Type: text/plain; charset="us-ascii"; format=flowed Concerning Andy Oxenham's great paper, I would venture to propose that Steve Greenberg and I presented a similar phenomenon some nine years ago. This teaches us procrastinating with publishing... From Abstracts from the 1995 ARO meeting: The residue reconsidered: Isolating temporal cues for the low pitch of complex tones. S. Greenberg, University of California, Berkeley, and International Computer Science Institute, P.L. Divenyi, Veterans Affairs Medical Center, Martinez, CA A.P. Algazi, Veterans Affairs Medical Center, Martinez, CA For well over a century and a half, auditory theorists have debated the relative contribution of spectral and temporal cues towards the formation of low pitch. The issue remains unresolved, in part, because temporal cues (the "residue") are difficult to isolate, due to their inherent association with the spectrum and their correlation with tonotopic location. In order to gauge the contribution of purely temporal cues in the low pitch of complex signals, we use signals in which a 150-ms sinuoidal carrier (of frequency fc) is amplitude-modulated (AM) by a sinusoid of frequency fmod1 which has been amplitude-modulated by a sinusoid of frequency fmod2. The ratio fmod1/fmod2 is similar to the carrier/modulator ratio of AM signals known to produce a distinct low pitch sensation. Measuring psychophysically, in an adaptive 2AFC paradigm, the discriminability of the pitch generated by fmod2, allows us to project the fine temporal structure associated with pitch-salient low-frequency AM signals to the mid- and high-frequency regions of the tonotopic gradient (at 4 kHz and above) where neural phase-locking to the carrier frequency is largely absent. Because the two modulating frequencies are within the phase-locking limit, temporal characteristics of the resulting excitation pattern are similar to those associated with low-frequency AM signals. The low pitch discriminability of these dual-modulation signals is comparable to that of conventional AM and sinusoidal signals over a wide range of modulator ratios and carrier frequencies, suggesting that temporal cues, based on neural phase-locking, play an important role in the coding of low pitch. --=====================_78702312==_.ALT Content-Type: text/html; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable <html> Concerning Andy Oxenham's great paper, I would venture to propose that Steve Greenberg and I presented a similar phenomenon some nine years ago. This teaches us procrastinating with publishing...<br><br> From Abstracts from the 1995 ARO meeting:<br><br> The residue reconsidered: Isolating temporal cues for the low <br> pitch of complex tones. <br> S. Greenberg, University of California, Berkeley, and International Computer Science Institute, P.L. Divenyi, Veterans Affairs Medical Center, Martinez, CA<br> A.P. Algazi, Veterans Affairs Medical Center, Martinez, CA<br><br> For well over a century and a half, auditory theorists have debated the relative contribution of spectral and temporal cues towards the formation of low pitch. The issue remains <br> unresolved, in part, because temporal cues (the &quot;residue&quot;) are difficult to isolate, due to their inherent association with the spectrum and their correlation with tonotopic location. <br> In order to gauge the contribution of purely temporal cues in the low pitch of complex signals, we use signals in which a 150-ms sinuoidal carrier (of frequency fc) is amplitude-modulated (AM) by a sinusoid of frequency fmod1 which has been amplitude-modulated by a sinusoid of frequency fmod2. The ratio fmod1/fmod2 is similar to the carrier/modulator ratio of AM signals known to produce a distinct low pitch sensation.&nbsp; Measuring psychophysically, in an adaptive 2AFC paradigm, the discriminability of the pitch generated by fmod2, allows us to project the fine temporal structure associated with&nbsp; pitch-salient low-frequency AM signals to the mid- and high-frequency regions of the tonotopic gradient (at 4 kHz and above) where neural phase-locking to the carrier frequency is largely absent. Because the two modulating frequencies are within the phase-locking limit, temporal characteristics of the resulting excitation pattern are similar to those associated with low-frequency AM signals. The low pitch discriminability of these dual-modulation <br> signals is comparable to that of conventional AM and sinusoidal signals over a wide range of modulator ratios and carrier frequencies, suggesting that temporal cues, based on neural phase-locking, play an important role in the coding of low pitch. <br> </html> --=====================_78702312==_.ALT--


This message came from the mail archive
http://www.auditory.org/postings/2004/
maintained by:
DAn Ellis <dpwe@ee.columbia.edu>
Electrical Engineering Dept., Columbia University