Re: Pitch learning (Stewart Hulse )


Subject: Re: Pitch learning
From:    Stewart Hulse  <hulse@xxxxxxxx>
Date:    Fri, 23 Feb 2007 15:51:01 -0500
List-Archive:<http://lists.mcgill.ca/scripts/wa.exe?LIST=AUDITORY>

Annabel -- It was fun to read your comments about my article with Tony Wright et al. It would have been nice to do other transpositions, as you suggest, but you are right that there is only so much that one can do in a given project. Most importantly, although we could have used sine tones for stimuli, that would probably not eliminate completely the possibility of an harmonic-based transposition because of the difficulty (if not the impossibility) of generating harmonic-free stimuli even with sine tones. Even more to the point, it seems to me that it is critical to focus not only on the stimulus properties of the tonal melodies, but also on those of the atonal melodies, because the same concerns would apply to the atonal melodies for which transposition did not occur. The fact that tonal -- but not atonal -- melodies produced transposition is the key point, and that suggests very strongly that something other than transposition based on some physical characteristic of the stimulus was at work in the tonal melodies. Otherwise, we would have expected transposition in the atonal melodies as well. Best regards, Stew Hulse -----Original Message----- From: AUDITORY - Research in Auditory Perception [mailto:AUDITORY@xxxxxxxx On Behalf Of Annabel Cohen Sent: Wednesday, February 07, 2007 9:22 PM To: AUDITORY@xxxxxxxx Subject: Re: Pitch learning On this discussion of the representation of pitch in music, Martin Braun's reference to http://web.telia.com/~u57011259/eng7.htm provides a further reference to: http://web.telia.com/~u57011259/Wright.htm the Wright, Rivera, Hulse, Shyan & Neiworth (2000; Journal of Experimental Psychology:General) article on octave generalization in rhesus monkeys. The evidence in this paper for octave generalization for tonal melodies by rhesus monkeys is impressive, however, whether this reflects something special about sensitivity to the octave (chroma) rather than sensitivity to the overtone series or periodicity is still not clear from this study. The data are interpreted in terms of sensitivity to the octave. The possibility of the sensitivity to periodicity is not mentioned (from my quick recent reading of the article). Transposition to the tritone (6 semitones) was the only non-octave transposition examined. I expect for reasons of expediency (i.e., a lot more testing required), the authors did not test rhesus monkeys on a condition of transposition to the perfect fifth (ratio 3/2 - 7 semitones up or 5 semitones down) in addition to the 1.5 octave ( 6 semitones -- the tritone). It would have been nice to know if rhesus monkeys could be shown to recognize a transposed melody better if transposed to the (harmonically near) perfect fifth as compared to the tritone. The monkey's sensitivity to tonality also shown in the Wright et al. (2000) study is consistent with this possible outcome, i.e., that the perfect fifth would lead to higher performance than the tritone transposition. It can be pointed out also that the timbres used in the Wright et al. study were synthesized instrument tones, which physically represent the octave. A more conservative approach would be to use sine tones such that the octave was not present in the original melodic stimulus (or even synthesized timbres lacking small integer multiples of the fundamental). Annabel Cohen On 7 Feb 2007 at 20:02, Martin Braun wrote: > Ole Kühl wrote: > > "We could say that we conceptualize a pitch through the integration of > information from two different domains: a perceived timbre domain and > a learned schema for partials." > > It does not seem to be a cognitive process, rather a sensory one. > Musicians make use of it without being aware of much. > > Integration across different auditory domains, yes. I would call it > the timbre domain and the chroma domain (for the given case of extreme > low pitch from musical instruments). > > It may not be necessary to assume "a learned schema for partials", > though. The anatomy of the auditory thalamus suggests a chroma > filtering that is as precognitive as the f0 filtering in the midbrain. > For a very short summary, see: > > http://web.telia.com/~u57011259/eng7.htm > > > Martin > > --------------------------------------------------------------------- > Martin Braun Neuroscience of Music S-671 95 Klässbol Sweden web site: > http://w1.570.telia.com/~u57011259/index.htm Annabel J. Cohen, Ph. D. Department of Psychology University of Prince Edward Island Charlottetown, P.E.I. C1A 4P3 CANADA e:mail acohen@xxxxxxxx phone: (902) 628-4325 office; (902) 628-4331 lab fax: (902) 628-4359 www.upei.ca/~musicog www.upei.ca/~cmtc


This message came from the mail archive
http://www.auditory.org/postings/2007/
maintained by:
DAn Ellis <dpwe@ee.columbia.edu>
Electrical Engineering Dept., Columbia University