Jensen Jesper Hojvang author Christensen Mads Gaesboll author Ellis Daniel P. W. author Columbia University. Electrical Engineering Jensen Soren Holdt author Columbia University. Electrical Engineering originator text Articles 2009 English For music information retrieval tasks, a nearest neighbor classifier using the Kullback-Leibler divergence between Gaussian mixture models of songs' melfrequency cepstral coefficients is commonly used to match songs by timbre. In this paper, we analyze this distance measure analytically and experimentally by the use of synthesized MIDI files, and we find that it is highly sensitive to different instrument realizations. Despite the lack of theoretical foundation, it handles the multipitch case quite well when all pitches originate from the same instrument, but it has some weaknesses when different instruments play simultaneously. As a proof of concept, we demonstrate that a source separation frontend can improve performance. Furthermore, we have evaluated the robustness to changes in key, sample rate, and bitrate. Electrical engineering Acoustics 17 4 693 703 2009-05 http://dx.doi.org/10.1109/TASL.2008.2012314 http://hdl.handle.net/10022/AC:P:11816 NNC NNC 2011-11-18 12:33:10 -0500 2012-12-18 00:48:51 -0500 5799 eng