Subject: Re: Pitch learning From: Martin Braun <nombraun@xxxxxxxx> Date: Wed, 28 Feb 2007 16:37:09 +0100 List-Archive:<http://lists.mcgill.ca/scripts/wa.exe?LIST=AUDITORY>Dear Annabel, Stew, and others, Annabel Cohen wrote: "I am willing to concede that sensitivity to overlapping harmonics may not be the basis of the musical and octave sensitivity of monkeys; what remains unclear to me is whether there is an "octave circular pitch processor" or rather than a "small-integer / periodicity- sensitive processor"." A periodicity detector processes the tones A3 and A4 differently. That's why we hear a pitch difference. Either the units for A3 or those for A4 respond more strongly. For some sounds the output may be uncertain, and octave confusion can occur. But if there is a pitch, there is only one. So it is not possible that a monkey can generalize from A3 to all A, i.e. to chroma A. Annabel then asked: "How also does one explain the monkey's superior performance on tonal as opposed to atonal melodies, when tonal melodies are characterized by tones related by small integer ratios (though typically not octaves) as compared to tone relations in atonal melodies." The authors of the study suggested that "tonal melodies [that] are characterized by tones related by small integer ratios" may be more "memorable". This makes a lot of sense, and the same has been observed in humans. In the auditory system of mammals we have a division of labor. The auditory midbrain extracts periodicity (thus extracting more information from "tones related by small integer ratios"), whereas the auditory thalamus presumably extracts chroma. The anatomy of these two complexes is quite clear on this matter. The midbrain neuron layers are laid out to integrate best over the distance of ca 260 to ca 520 Cent, thus being most sensitive for the ratios 6/5, 5/4, and 4/3. The thalamus neuron layers are laid out to integrate best over the distance of ca 1200 Cent, i.e. the octave. More details are easily available on the internet. Martin --------------------------------------------------------------------- Martin Braun Neuroscience of Music S-671 95 Klässbol Sweden web site: http://w1.570.telia.com/~u57011259/index.htm