Chroma perception; was: Robust method of fundamental frequency estimation.

[Martin Braun <nombraun@xxxxxxxxx>]

Dear Pierre, Arturo, Rob, and others,

on Friday, 2 Feb, Pierre Divenyi raised this question:

"What is the most remarkable, and still begging for explanation (which
should be difficult to obtain experimentally because it tackles a subjective
dimension), is the subjective smoothness of a descending scale played on any
of the low-pitched instruments, despite the lack of smoothness when you play
only a sinusoid."

At least there is now a good candidate for an "explanation":

chroma perception (also called pitch class perception).

Anatomical and physiological evidence indicates that the mammalian auditory
brain carries out chroma filtering in the ventral division of the medial
geniculate nucleus (MGN) of the thalamus.

For this reason, tones from musical instruments around C1 (32.7 Hz) can be
well identified, even when a periodicity at this frequency is absent.
Information from octave spaced harmonics C2, C3, etc. is always present. And
most importantly, it usually is the strongest component of all octave-spaced
sound energy of the tone.

Octave spaced signals are assumed to be integrated and extracted in the
thalamus, thus forming the basis of chroma perception. So, if the f0 at C1
on the piano is fuzzy or absent, our brain still perceives the chroma of C.
In combination with the available timbre information it then can easily
attribute the pitch of C1 to the sound.

Musician, and presumably most non-musicians as well, have no problems
hearing the difference between C1 and C#1 from natural sounds of musical
instruments.

The relevant anatomical and physiological evidence is listed here:

http://web.telia.com/~u57011259/Wright.htm


Martin

---------------------------------------------------------------------
Martin Braun
Neuroscience of Music
S-671 95 Klässbol
Sweden
web site: http://w1.570.telia.com/~u57011259/index.htm