Just to clarify,A "chroma feature" is just an spectrogram warped into the 12-semitone ET pitch class space
http://labrosa.ee.columbia.edu/matlab/chroma-ansyn/ So for each audio frame you get a vector of 12 values. That's useful for chord and tonality recognition. By stream based I just meant realtime. Not file-based analysis.Josep > Date: Mon, 3 Jun 2013 17:00:25 +0200 > From: zmoel...@iem.at > To: pd-list@iem.at > Subject: Re: [PD] chroma > > -----BEGIN PGP SIGNED MESSAGE----- > Hash: SHA1 > > On 2013-06-03 16:26, Jeppi Jeppi wrote: > > Hi again people,just wonder whether has anyone attempted to > > implement a stream-based chroma descriptor in Pd. > > it would be great if you could clarify your expressions "chroma" and > "stream-based". > > (in Pd we are dealing with loads of totally different "streamed" media > that have "chroma") > > fgmsdr > IOhannes > -----BEGIN PGP SIGNATURE----- > Version: GnuPG v1.4.12 (GNU/Linux) > Comment: Using GnuPG with Icedove - http://www.enigmail.net/ > > iEYEARECAAYFAlGsr4YACgkQkX2Xpv6ydvSu+ACgt/RzObVbCsBC10lCCqs2i45c > GGsAoMRYmT227cnhML7p+Aadcwf0vSR9 > =iXF5 > -----END PGP SIGNATURE----- > > _______________________________________________ > Pd-list@iem.at mailing list > UNSUBSCRIBE and account-management -> > http://lists.puredata.info/listinfo/pd-list
_______________________________________________ Pd-list@iem.at mailing list UNSUBSCRIBE and account-management -> http://lists.puredata.info/listinfo/pd-list
