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

*To*: AUDITORY@xxxxxxxxxxxxxxx*Subject*: Re: Robust method of fundamental frequency estimation.*From*: "Richard F. Lyon" <DickLyon@xxxxxxx>*Date*: Wed, 28 Feb 2007 07:54:31 -0800*Comments*: To: Dan Ellis <dpwe@EE.COLUMBIA.EDU>*Delivery-date*: Wed Feb 28 11:11:33 2007*In-reply-to*: <2340a7630702272102h48f879f7s4ee91dff3c5d8dc5@mail.gmail.com>*List-archive*: <http://lists.mcgill.ca/scripts/wa.exe?LIST=AUDITORY>*List-help*: <http://lists.mcgill.ca/scripts/wa.exe?LIST=AUDITORY>, <mailto:LISTSERV@LISTS.MCGILL.CA?body=INFO AUDITORY>*List-owner*: <mailto:AUDITORY-request@LISTS.MCGILL.CA>*List-subscribe*: <mailto:AUDITORY-subscribe-request@LISTS.MCGILL.CA>*List-unsubscribe*: <mailto:AUDITORY-unsubscribe-request@LISTS.MCGILL.CA>*References*: <20070131171152.74585.qmail@web26301.mail.ukl.yahoo.com> <2a17c00a0701311310i538f446ana9c8532934e57529@mail.gmail.com> <2728.209.251.130.178.1170460061.squirrel@webmail.cise.ufl.edu> <p06230900c206338eee40@192.168.23.150> <3043.209.251.130.178.1172552172.squirrel@webmail.cise.ufl.edu> <1172561822.45e3df9e069a2@webmail.uni-magdeburg.de> <3836.209.251.130.178.1172586425.squirrel@webmail.cise.ufl.edu> <p06230901c20a46476243@192.168.1.100> <1257.209.251.130.178.1172633241.squirrel@webmail.cise.ufl.edu> <2340a7630702272102h48f879f7s4ee91dff3c5d8dc5@mail.gmail.com>*Reply-to*: "Richard F. Lyon" <DickLyon@xxxxxxx>*Sender*: AUDITORY - Research in Auditory Perception <AUDITORY@xxxxxxxxxxxxxxx>

Arturo -

However, what I said about the equivalence of applying a summary autocorrelation to the output of a orthonormal linear filterbank and applying autocorrelation to the input signal is still true, it just does not apply to your model, because of the HWR.

That's true, but I wouldn't assume that people doing filterbank -> autocorrelation -> summary are simply misguided. Very often in addition to nonlinearity there is per-channel normalization (and/or weighting to emphasize particular subbands/sources), which has a profound effect. I used gammatones plus nonlinearity normalization in the pitch tracker in my CASA system, but I felt like I understood it better when I saw the way that Tolonen & Karjalainen simply whitened the spectrum (on a suitably coarse scale) then autocorrelated a low band (with HWR?) and a high band (with envelope extraction?) to avoid the need for a large number of parallel channels. I think their paper gets at a lot of the essence of many such pitch models.

True. You don't get envelope pitch that way (AM modulated noise, for example), but you do get many other effedts. I agree that a filterbank can be a useful part of a whitening system in front of an ACF.

Dick

DAn.

Tolonen, T. and M. Karjalainen (2000). A computationally efficient multipitch analysis model. IEEE Transactions on Speech and Audio Processing 8(6), 708716.

**References**:**Robust method of fundamental frequency estimation.***From:*Roisin Loughran

**Re: Robust method of fundamental frequency estimation.***From:*Yi-Wen Liu

**Re: Robust method of fundamental frequency estimation.***From:*Arturo Camacho

**Re: Robust method of fundamental frequency estimation.***From:*Arturo Camacho

**Re: Robust method of fundamental frequency estimation.***From:*Eckard Blumschein

**Re: Robust method of fundamental frequency estimation.***From:*Arturo Camacho

**Re: Robust method of fundamental frequency estimation.***From:*Arturo Camacho

**Re: Robust method of fundamental frequency estimation.***From:*Dan Ellis

- Prev by Date:
**Re: Pitch learning** - Next by Date:
**Re: Pitch learning** - Previous by thread:
**Re: Robust method of fundamental frequency estimation.** - Next by thread:
**Re: Robust method of fundamental frequency estimation.** - Index(es):