[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Rhythm perception

A couple of additional references in this area:

Fabien Gouyon, who is defending his PhD thesis on "Computational Rhythm
Description" very soon, available at:
and Miguel Alonso work, whose system was ranked 1st on the MIREX 2005 Audio
Tempo Extraction contest: 

Further references on this area can be found at the MIREX 2004 and 2005
Audio Tempo Extraction contest webpages:




Luis Gustavo Martins
Researcher / PhD Student

> -----Original Message-----
> From: AUDITORY Research in Auditory Perception
> [mailto:AUDITORY@xxxxxxxxxxxxxxx] On Behalf Of Michael Fulton
> Sent: quinta-feira, 17 de Novembro de 2005 15:36
> To: AUDITORY@xxxxxxxxxxxxxxx
> Subject: Re: [AUDITORY] Rhythm perception
> >
> >We were discussing rhythm patterns the other day and the question came
> >up about how one determined the start of a pattern.  If there is a
> >heavy emphasis on one beat of the sequence then I can understand that
> >that is taken as the first beat.
> >   But if the sequence is unemphasised how does one decide?  Or do
> >people decide differently, or is it cultural?
> >
> >I am not sure where to start to look -- as ever this is outside my
> >general field of study -- but I though this list might be the place
> >to ask for clues.
> >
> >I hope I have explained the question sufficiently!
> >
> >==John ffitch
> As far as I am aware the two most successful systems to date for
> extracting
> information about musical measures are by Klapuri and Goto (Klapuri being
> the better). These both identify the foot tapping rate or tactus as its
> called and also the start of each measure. The system by Goto relied on
> detecting chord changes and Klapuris system looked for rythmic
> fluctatuations in the signals amplitude i.e. low,loud,-,loud.
> However, both these relied on their being some form of emphasis/change and
> were designed for primarily western popular music. If the sequence is
> unemphasisied then I suppose it really is a purely  subjective decision
> based on experience and culture.
> The papers I have mentioned are
> http://www.cs.tut.fi/sgn/arg/klap/sapmeter.pdf
> and
> http://staff.aist.go.jp/m.goto/PROJ/bts.html                   No.9 on
> page
> Hope this helps as this is also a problem that has been troubling me
> recently i.e. given a peice of music how do we reliably regocnise the
> start
> of each measure ?
> Mick