Subject: Re: Intermediate representation for music analysis From: Dan Ellis <dpwe@xxxxxxxx> Date: Wed, 19 Jul 2006 15:35:14 -0400>> Could you give me some references to people/articles that make use of this >> technique ? In fact, this is exactly the technique I use to improve the >> frequency resolution of the constant-Q filterbank. I have an implementation in Matlab, "ifgram.m", which calculates a spectrogram including an instantaneous frequency estimate for each bin. You can download it from http://www.ee.columbia.edu/~dpwe/resources/matlab/sinemodel/ .. search for ifgram. This idea recurs quite a lot. It's in Flanagan & Golden's original 1966 Phase Vocoder paper: http://www.ee.columbia.edu/~dpwe/e6820/papers/FlanG66.pdf also F. J. Charpentier from ICASSP 86: http://www.ee.columbia.edu/~dpwe/papers/Charp86-pitchphase.pdf and developed more recently by Abe and co-workers: http://citeseer.ist.psu.edu/context/1301499/0 DAn.